Download PPI Final Report: Main Text

Market investigation into
payment protection insurance
29 January 2009
© Competition Commission 2009 Website: www.competition-commission.org.uk
Members of the Competition Commission who conducted this inquiry
Peter Davis (Chairman of the Group)
Professor John Baillie
Christopher Bright
Professor John Cubbin
Richard Farrant
Chief Executive and Secretary of the Competition Commission
Martin Stanley
The Competition Commission has excluded from this report information which the inquiry group considers should be excluded having regard to the three considerations set out in section 244 of the Enterprise Act 2002 (specified information: considerations
relevant to disclosure). The omissions are indicated by . Some numbers have been replaced by a range. These are shown in square brackets. Non-sensitive wording is also indicated in square brackets. The following typographical corrections have been made to the version of this report published on the CC website on 29 January 2009: • ‘Practical’ corrected to ‘practicable’ in paragraphs 79 and 10.567.
• Appendix 10.1 personal quote for stand-alone and short-term income protection
has been amended to remove the following words which were inadvertently
included ‘If you want to buy PPI from us you may contact us 24 hours after [the
conclusion of the credit sale period] or we may contact you from 7 days after this
date’ and ‘To take out this PPI please ring xxx xxx xxx or visit
www.companyname.com after xx.xx.2009 [this date is at least 24 hours after
the conclusion of the credit sale period]’.
iii
Market investigation into payment protection insurance
Contents
Page
Summary................................................................................................................................. 1 Findings ................................................................................................................................ 17 1.
Introduction ............................................................................................................... 17 The roles of the FSA, the OFT and the CC in the oversight of PPI........................... 17 Events leading up to this investigation ...................................................................... 18 Conduct of the investigation...................................................................................... 19 Publication of evidence and other materials ............................................................. 20 Report overview ........................................................................................................ 21 2.
PPI underwriting and distribution .............................................................................. 22 What is PPI? ............................................................................................................. 22 Short-term income protection.................................................................................... 23 Possible alternative policies ...................................................................................... 24 Sales of PPI and shares of supply ............................................................................ 25 Customers................................................................................................................. 27 Organization of the PPI value chain .......................................................................... 27 Underwriters ......................................................................................................... 28 Distributors ........................................................................................................... 29 Vertically-integrated businesses........................................................................... 30 Intermediaries....................................................................................................... 30 Stand-alone provision........................................................................................... 32 How PPI is sourced and sold .................................................................................... 33 Contractual relationships between underwriters and distributors ......................... 33 The sale of PPI to customers ............................................................................... 34 3.
Market definition for the distribution of PPI ............................................................... 45 Summary................................................................................................................... 45 Introduction ............................................................................................................... 47 Product market.......................................................................................................... 49 The responsiveness of PPI demand to PPI prices ............................................... 49 The substitutability of different levels of PPI cover sold at a point of sale ............ 52 The substitutability of PPI policies offered by different distributors ...................... 53 Substitutability of stand-alone PPI and short-term income protection .................. 54 The substitutability of other insurance products with PPI ..................................... 63 Substitution between combinations of PPI and credit .......................................... 65 The position of non-standard credit consumers ................................................... 72 Geographic market.................................................................................................... 74 Conclusion on market definition ................................................................................ 74 The impact of our market definition on our analysis of competition .......................... 75 4.
Indicators of the extent of competition between PPI providers in the supply of PPI . 78 Summary................................................................................................................... 78 Introduction ............................................................................................................... 79 Variation of PPI prices over time ............................................................................... 79 Indicators of non-price competition ........................................................................... 80 New PPI policies................................................................................................... 80 Innovations within existing products ..................................................................... 81 Reasons for introducing new and amended PPI policies ..................................... 81 Advertising/marketing ........................................................................................... 82 Price dispersion......................................................................................................... 83 Search....................................................................................................................... 85
Switching................................................................................................................... 85 Switching PPI without switching credit product .................................................... 86 Claims ratios ............................................................................................................. 88 iv
5.
6.
Profitability................................................................................................................. 90 How we examined profitability of PPI distribution................................................. 90 Results of profitability analysis ............................................................................. 93 Documentary evidence on profitability.................................................................. 94 Conclusions on profitability................................................................................... 94 Conclusions on the extent of competition between distributors ................................ 97 Coordination in the distribution of PPI....................................................................... 98 Factors affecting the nature and extent of competition in the supply of PPI ............. 99 Competition between PPI providers ........................................................................ 100 Search..................................................................................................................... 100 Benefits to search............................................................................................... 100 Barriers to search ............................................................................................... 103 Conclusions on search ....................................................................................... 110 Switching................................................................................................................. 111 Barriers to switching ........................................................................................... 111 The effect of single-premium policies on switching ............................................ 112 Conclusions on switching ................................................................................... 114 Barriers to entry and expansion .............................................................................. 115 The point-of-sale advantage ................................................................................... 116 What is the point-of-sale advantage ................................................................... 117 Evidence for the existence of a point-of-sale advantage.................................... 117 The causes of the point-of-sale advantage ........................................................ 118 The advantages gained by distributors from selling PPI at the point of sale ...... 119 The advantages gained by customers from PPI being sold at the point of sale. 121 How the point of sale affects competition ........................................................... 121 The possible competitive constraint from new stand-alone products...................... 122 The effects on consumers of the level of competition in the market ....................... 124 Conclusions on factors affecting the nature and extent of competition ................... 125 Retail PPI ................................................................................................................ 127 Introduction ............................................................................................................. 127 Retail PPI ................................................................................................................ 127 What is retail PPI ................................................................................................ 127 Who sells retail PPI? .......................................................................................... 128 Possible alternative policies ............................................................................... 129 Sales of retail PPI ............................................................................................... 129 Organization of the retail PPI value chain .......................................................... 130 How retail PPI is sold.......................................................................................... 131 Market definition...................................................................................................... 132 Substitutability of retail PPI with short-term IP and stand-alone CCPPI............. 133 The substitutability of the retail credit and retail PPI combination with alternative combinations of credit and PPI....................................................... 135 The substitutability of the goods, retail credit and retail PPI bundle with alternative bundles of goods, credit and PPI.................................................... 140 Geographic market for retail PPI ........................................................................ 142 Conclusion on market definition ......................................................................... 143 Indicators of the extent of competition between retail PPI providers in the supply of retail PPI ........................................................................................................... 144 Variation of retail PPI prices over time ............................................................... 144 Indicators of non-price competition..................................................................... 144 Advertising.......................................................................................................... 147 Price dispersion .................................................................................................. 148 Profitability .......................................................................................................... 149 Recent developments in the supply of retail PPI..................................................... 157 Conclusions on the extent of competition between distributors .............................. 159 Factors affecting the nature and extent of competition in the supply of retail PPI... 159 Competition between retail PPI providers .......................................................... 160 v
7.
8.
9.
10.
Search ................................................................................................................ 160 Switching ............................................................................................................ 162 Barriers to entry and expansion.......................................................................... 163 The point-of-sale advantage............................................................................... 164 The effects on consumers of the level of competition in the market................... 166 Conclusions on factors affecting the nature and extent of competition ................... 166 The underwriting of PPI........................................................................................... 168 Market definition...................................................................................................... 168 Product market ................................................................................................... 168 Geographic market ............................................................................................. 170 Conclusion on the market definition for underwriting of PPI............................... 171 Is there evidence of market power being exerted by underwriters?........................ 171 Concentration of the market ............................................................................... 171 Capacity.............................................................................................................. 171 Entry, exit and expansion ................................................................................... 172 Countervailing buyer power................................................................................ 172 Profitability of underwriting ...................................................................................... 173 Documentary evidence from the parties............................................................. 173 Financial performance of underwriting ............................................................... 173 Conclusions on the profitability of PPI underwriting ........................................... 173 Conclusions on the underwriting market ................................................................. 174 Vertical integration .................................................................................................. 175
Vertically-integrated businesses ............................................................................. 175 Ways in which vertical integration might harm consumers...................................... 175 Reducing the number of bidders for non-integrated underwriting business ....... 176 Limiting the size of the underwriting market ....................................................... 176 Removing the competitive pressure on in-house underwriters........................... 176 Creating a greater incentive for (in-house) underwriters to reject PPI claims .... 177 Conclusions on vertical integration ......................................................................... 178 Findings................................................................................................................... 179 The distribution of PPI............................................................................................. 179 The underwriting of PPI........................................................................................... 180 Vertical integration .................................................................................................. 180 R
emedies ................................................................................................................ 181
Introduction ............................................................................................................. 181 Framework for the assessment of remedies and relevant customer benefits ......... 182 General issues ........................................................................................................ 183 Impact of ICOBS regulation................................................................................ 183 The impact of the economic downturn................................................................ 184 Impact on the ‘protection gap’ ............................................................................ 185 Increased risk of adverse selection .................................................................... 186 Comparison with extended warranties and store cards...................................... 187 Recent competitive developments...................................................................... 187 Remedies the CC has decided to implement .......................................................... 188 Prohibition on selling PPI at the credit point of sale and within a fixed time period of the credit sale (the ‘point-of-sale prohibition ’) .................................................. 188 Summary of this element of the remedies package ........................................... 188 How the point-of-sale prohibition contributes to addressing the AEC ................ 190 Risks of a point-of-sale prohibition ..................................................................... 190 Effectiveness of the remedy and alternatives to a point-of-sale prohibition ....... 199 Conclusion on the need for a point-of-sale prohibition ....................................... 201 The design of the point-of-sale prohibition ......................................................... 203 Obligation to provide a personal PPI quote............................................................. 221 Summary of this element of the remedies package ........................................... 221 How the obligation to provide a personal PPI quote addresses the AEC........... 222 The content of the personal PPI quote ............................................................... 222 vi
Obligation to provide information about the cost of PPI and ‘key messages’ in marketing material ................................................................................................ 226 Summary of this element of the remedies package ........................................... 226 How the provision of information in marketing materials will address the AEC.. 227 Issues raised ...................................................................................................... 227 The design of a common price metric ................................................................ 232 An obligation to provide information to OFT and FSA for monitoring and publication and to provide information about claims ratios to third parties ........... 237 Summary of this element of the remedies package ........................................... 237 How the requirement to provide information to third parties will address the AEC.................................................................................................................. 237 Prohibition on the selling of single-premium PPI policies........................................ 243 Summary of this element of the remedies package ........................................... 243 How the prohibition of single-premium PPI policies will address the AEC ......... 243 Issues raised ...................................................................................................... 243 Obligation to offer retail PPI separately from merchandise cover when both are offered as a bundled product ................................................................................ 250 Summary of this element of the remedies package ........................................... 250 How the requirement to unbundle retail PPI from merchandise cover will address the AEC .............................................................................................. 250 Issues raised ...................................................................................................... 251 An obligation to provide annual statement of PPI cost and a reminder of the consumer’s right to cancel .................................................................................... 256 Summary of this element of the remedies package ........................................... 256 How the requirement to provide an annual statement will address the AEC...... 256 Issues raised ...................................................................................................... 257 The design and implementation of this measure................................................ 258 Implications of remedies package for providers of stand-alone PPI and short-term IP ......................................................................................................... 264 Options we decided not to pursue........................................................................... 265 Further standardization of PPI information given to the consumer at the point of sale............................................................................................................... 265 All PPI policies to be renewed annually.............................................................. 266 Alternative remedies to a prohibition of single-premium PPI policies................. 266 Minimum standards for elements of PPI policies that act as a barrier to switching .......................................................................................................... 269 Obligation to share information about customer claims...................................... 270 Obligation to share information about consumers’ credit card and retail credit balance with a nominated underwriter ............................................................. 271 Price caps........................................................................................................... 272 Relevant customer benefits..................................................................................... 273 Potential customer benefits of the point-of-sale advantage................................ 274 Knowledge by credit providers that a consumer will take out PPI ...................... 285 Lower credit prices ............................................................................................. 287 The effectiveness and proportionality of the package of remedies ......................... 291 The rationale for implementation of all elements of the remedies package ....... 291 Benefits and synergies of the remedies package............................................... 293 Modification of the remedies package for relevant customer benefits................ 294 Extent of consumer detriment............................................................................. 296 The cost of the package of remedies ................................................................. 297 Conclusion on effectiveness and proportionality ................................................ 300 Implementation of remedies.................................................................................... 301 How the remedies should be implemented ........................................................ 301 The timescale for implementation....................................................................... 302 Monitoring compliance and effectiveness........................................................... 304 Summary of our decisions on remedies .................................................................. 315 vii
Summary
1. We found that each distributor and intermediary faces little competition for the sale of
Payment Protection Insurance (PPI) when it is sold in combination with the credit it
insures. We found that there were features of relevant markets which led to an
adverse effect on competition (AEC) in these markets and in turn resulted in
consumers facing higher prices and less choice than they would if there was effective
competition between PPI providers. As a result of this lack of competition we found
that it is highly profitable for distributors to sell PPI, though we found that some of the
resultant profit is used to subsidize credit prices. We concluded that there were
serious deficiencies in the competitive process for selling PPI policies, and, in order
to remedy the adverse effects identified, a package of remedies would be required
which includes some significant restrictions on what parties selling both PPI and
credit can do (and also impose some burden on parties that offer only PPI to consumers). We concluded that such an intervention in these markets would enhance
overall consumer welfare, and that the scale of the problem identified warranted a
significant intervention.
2. We concluded that we should impose: a prohibition on distributors and intermediaries
from selling PPI to their credit customers within seven days of a credit sale, unless
the customer has proactively returned to the seller at least 24 hours after the credit
sale; a prohibition on selling single-premium PPI policies (where the premium is paid
in one upfront payment, generally by adding the premium to the credit borrowed); a
requirement on retail PPI distributors to offer retail PPI separately when they also
offer retail PPI bundled with merchandise cover; and several requirements to provide
specified information in marketing materials, at the points of sale of credit and PPI,
and each year after the PPI policy has entered into force.
Background
3. On 7 February 2007 the Office of Fair Trading (OFT) referred the supply of all PPI
(except store card PPI) 1 to non-business customers in the UK to the Competition
Commission (CC) for investigation under section 131 of the Enterprise Act 2002 (the
Act). The reference followed the receipt of a super-complaint about PPI on
13 September 2005 from Citizens Advice. This document, together with its
appendices, constitutes our final report.
4. PPI covers repayments on the insured credit product if the borrower suffers an
insured event—usually accident (A), sickness (S), unemployment (U) or death
(referred to as life (L) cover). PPI is sold to cover a variety of credit products, but
nearly 95 per cent of PPI sold in the UK in 2007 was either personal loan PPI
(PLPPI), credit card PPI (CCPPI), mortgage PPI (MPPI) or second-charge mortgage
(also known as secured loan) PPI (SMPPI). We looked at two other forms of PPI
during our investigation. The first was motor finance PPI (motor PPI) where the credit
is given for the specific purpose of buying a car. We found that motor PPI was PLPPI
and our conclusions on PLPPI in this report apply to motor PPI unless otherwise
specified. The other form of PPI is retail finance PPI (retail PPI). We found two types
of retail PPI (see paragraph 6.2 for further details): ‘personal loan’ retail finance PPI,
which we concluded was a form of PLPPI; and ‘credit account’ retail finance PPI. We
consider ‘credit account’ retail finance PPI (‘retail PPI’) separately from other forms of
PPI in this summary, and in Sections 6 and 10.
1
The OFT Terms of Reference define store card PPI as services supplied for the purpose of protecting a store card holder’s
ability to maintain repayments due under the store card agreement.
1
PPI distributors and their business models
5. Most PPI policies are sold by distributors 2 at the point of sale of the credit being
insured (referred to as the point of sale in this report). There are a few providers of
PPI policies that do not also supply the credit to be insured (stand-alone PPI), and of
short-term income protection (short-term IP), which provides a pre-agreed amount of
money in the case of an insurable event occurring. We found that short-term IP is a
form of PPI and that short-term IP policies sold as a result of a referral during a credit
sale are effectively sales of PPI at the point of sale. PPI is also sold via intermediaries. This is particularly evident with MPPI where intermediaries are significant players, but they tend not to sell the distributor’s MPPI policy with its mortgage. In this
report we use the term ‘PPI providers’ where we refer to distributors, intermediaries
and stand-alone providers selling PPI.
6. The PPI policies sold are underwritten by insurers (usually referred to as underwriters). Four of the largest distributors of PPI also have insurance businesses within
their organizations which underwrite some or all of the policies they sell (these are
referred to as vertically-integrated businesses). Other distributors rely on external
underwriters.
7. In 2007 customers in the UK paid £3.8 billion in PPI premiums. Distributors are
contractually entitled to a percentage of this gross written premium (GWP)—the
amount of money paid by customers, net of insurance premium tax—as commission,
to cover expenses and contribute to profits. Typical commission rates are 50 to
80 per cent for PLPPI and CCPPI and 40 to 65 per cent for MPPI. The remaining
GWP is passed to the underwriter to cover expenses, including claims. We found that
between 11 and 28 per cent of GWP is paid out in claims (net of rebates), depending
on the product. In the event that claims levels are less than expected, the resulting
profit is generally split between the underwriter and distributor according to an agreed
profit share percentage; typically 90 to 100 per cent in favour of the distributor. A
separate profit share arrangement will typically apply to any investment income
earned by the underwriter on premium income and may also apply to tax benefits on
life business.
8. PPI is predominantly sold through three distribution channels: face–to-face contact in
branches, over the telephone, and over the Internet. Over half of all policies are sold
through face-to-face contact with a sales adviser. PPI is sold either on an advised
basis, where a personalized recommendation is provided as to the suitability of the
product for a customer, or a non-advised basis, where no personalized recommendation is made.
9. PPI policies are paid for either by a regular monthly premium or a single premium.
The single premium is a lump-sum payment to cover the cost of PPI for the term of
the policy, and is almost always paid for by the credit provider advancing the cost of
PPI, adding to the amount borrowed when the loan is agreed. This arrangement
means that the customer pays interest on the whole amount borrowed, including that
borrowed to pay for the single-premium PPI. Most PLPPI and about two-thirds of
SMPPI policies are paid for by a single premium. MPPI, CCPPI and retail PPI are
paid for using regular monthly premiums.
10. The amount paid for a PPI policy varies significantly between PPI providers, and can
be significant compared with the interest payable on the credit. For most PLPPI
policies we looked at, the cost of PLPPI over the term of the loan was greater than
2
See paragraphs 2.35–2.38 for details.
2
the interest payable on the loan. However, whilst prices vary between PPI providers,
we saw little price movement over time—we saw many PPI policies (especially
SMPPI, MPPI and CCPPI policies) whose price did not change at all over the period
2002 to 2006, whilst those that did change generally saw infrequent price increases
(there were very few examples of price decreases). While prices did not change, the
terms and conditions did change from time to time (for example, through removing an
exclusion that had prevented customers claiming for a particular event, such as
sickness due to a stress-related condition).
Customers and customer behaviour
11. PPI consumers are more likely to earn less than the national average income or
come from socio-economic groups C and D. Various surveys show that the most
frequently cited reasons given by consumers for taking out PPI relate to ‘peace of
mind’.
12. Not many consumers compare different policies before purchasing PPI. By far the
greatest level of search was among consumers looking at MPPI, where 21 per cent
of consumers compared policies. Most people who did compare policies did so by
looking at different combinations of credit and PPI.
13. Once consumers have a PPI policy they tend not to switch PPI without also switching
the underlying credit product. When consumers switch the PPI and credit combination, the primary driver does not appear to be a better PPI policy or a better deal on
the PPI; drivers for change include the end of a time-limited mortgage deal, 0 per
cent annual percentage rate (APR) on balance transfer offers for credit cards, and
better APR deals on loans.
Market definition
14. We looked at market definition, in order to use it as a framework for considering the
extent of competition between PPI providers. We considered the market definition for
retail PPI separately from other types of PPI policy (see paragraphs 21 to 26).
15. We looked at the product market from two perspectives. First, we looked at the
competitive constraints imposed on a PPI provider by other PPI providers and from
alternative types of insurance. Second, we looked at the extent to which distributors
and intermediaries supplying combinations of credit and PPI constrained one
another.
16. We found that PPI distributors and intermediaries were not constrained by other PPI
providers or by alternative types of insurance:
(a) The internal documents and testimony from the distributors of PPI indicated that
demand for their own PPI policies is unresponsive to changes in price.
(b) Our analysis of the evidence on consumer search and switching patterns
indicated that relatively few customers shop around for PPI. A survey we
commissioned of PPI consumers also indicated that limited numbers of PPI
consumers compared two or more PPI policies before their purchase. Our
analysis of these results indicated limited responsiveness of an individual
distributor’s PPI demand to changes in the price of its PPI policies.
(c) We assessed each potential source of competitive constraint in turn and found
that there was no strong evidence that any of these was likely to constrain prices.
3
(d) The high margins earned on PPI are indicative of a low responsiveness of
demand for PPI policies to changes in their prices.
17. We considered the potential competitive constraint imposed on PPI distributors or
intermediaries by the complementary relationship between PPI and the credit
products that they insure. We found the constraint that distributors or intermediaries
supplying combinations of credit and PPI imposed on one another to be low. The
evidence leading to this conclusion included:
(a) We found that the responsiveness of consumers in general to changes in prices
of PPI policies was low. The responsiveness of credit sales to changes in the
price of PPI was lower still.
(b) Internal documents received from the largest distributors showed that they were
not deliberately or consciously constraining their PPI prices because of the
prospect that high PPI prices might damage their credit sales.
(c) The CC GfK NOP 2008 survey indicated that the majority of search involved
searching ‘on the bundle’; however, the incidence of search was low. CC analysis
indicated that the level of search was insufficient to drive PPI prices down from
the currently high prices to competitive levels (see paragraph 16(d)).
(d) Our analysis of the main distributors’ sales data did not show clear evidence that
increases in PPI prices in the past had resulted in a significant reduction in credit
sales.
(e) The profit margins that PPI distributors earn on the underlying credit product are
far lower than those earned on PPI. Credit sales therefore have a proportionately
lower impact on profits than the loss of a PPI sale.
18. We found that individual distributors and intermediaries supplying PPI combined with
credit were not competitively constrained by other PPI providers, or by alternative
types of insurance policies. We also found that distributors and intermediaries
supplying PPI combined with credit did not constrain one another.
19. We therefore concluded that the product market was an individual distributor’s, or
intermediary’s, sales of a particular type of PPI policy. However, we found that whilst
PPI sold by distributors and intermediaries is not competitively constrained by standalone PPI, asymmetric constraints mean that stand-alone PPI is constrained by PPI
policies sold by distributors and intermediaries.
20. We found that the geographic market was the area within the UK in which a
distributor or intermediary sold PPI.
21. In considering the market definition for retail PPI, we identified three ways in which
retail PPI customers could potentially react to a small but significant non-transitory
increase in price (SSNIP) in retail PPI and which could result in that SSNIP being
unprofitable. These were that customers could substitute to other insurance policies,
customers could substitute to other forms of insured credit and customers could
substitute to other retailers that offered a credit account and retail PPI.
22. We found that a lack of suitable alternatives to retail PPI, the limited sales of possible
alternatives, their limited effectiveness as substitutes and the limited impact that they
appeared to have on retail PPI sales following a price rise indicated that the prospect
of customer substitution to alternative types of credit insurance was a weak constraint on retail PPI distributors.
4
23. With regard to substitution between bundles of credit and PPI, we found that it was
more difficult (and certainly no easier) for retail PPI consumers to consider the whole
cost of a bundle including retail PPI when compared with other types of PPI. We saw
no clear evidence to suggest that retail PPI customers were intrinsically more likely to
substitute between bundles of credit and PPI than customers of other types of PPI.
We also received conflicting views from the retailers regarding the extent to which
their customers substitute between bundles in response to changes in retail PPI
prices and terms. We concluded that the possibility that retail PPI customers may
substitute to alternative bundles of credit and PPI, in response to an increase in retail
PPI prices, was a weak constraint on retail PPI distributors.
24. We also found that it was likely to be difficult for retail PPI customers to substitute
between retailers offering retail credit accounts and retail PPI. We received
conflicting views from the retailers regarding the extent to which their customers
would substitute in this way in practice and none of the parties was able to provide
primary evidence that customers substitute between retailers in response to changes
in retail PPI prices. We concluded that the possibility that retail PPI customers may
substitute to other retailers offering home shopping, retail credit accounts and retail
PPI, in response to an increase in retail PPI prices, was a weak constraint on retail
PPI distributors.
25. Because the constraint on retail PPI distributors from each of the three potential
sources of substitution was weak, we concluded that a narrow market definition was
appropriate, and that it would be inappropriate to include other types of credit
insurance, or other bundles of credit and PPI, or other bundles of merchandise, credit
and PPI within the relevant market. We therefore defined the relevant product market
as the supply of retail PPI by a distributor to its own retail customers.
26. We defined the relevant geographic market as the supply of retail PPI by distributors
to their customers within the UK.
Indicators of the extent of competition between PPI providers
27. We assessed a series of indicators of the level of competition between PPI
distributors and intermediaries. The purpose of this was twofold. First, it acted as a
check on whether our market definition was correct. Second, it allowed us to
understand the extent to which PPI providers actually competed with each other. We
considered retail PPI separately from other forms of PPI (see paragraphs 64 to 66).
28. We found that there is little variation in PPI prices over time, and there is little
evidence of PPI distributors seeking to win sales from each other by competing on
PPI prices.
29. There is little evidence of PPI distributors seeking to win sales from each other by
competing on non-price factors such as quality, innovation or choice. We consider
that it is unlikely that there is vigorous competition on non-price factors because
these factors are particularly difficult for consumers to observe and compare.
30. There is very limited advertising of PPI. Advertising spending is on the credit product
and where PPI is included in advertising literature, the focus of advertising is predominantly on the underlying credit product.
31. PPI prices and the prices of PPI combined with credit are widely dispersed even
when differences in quality are taken into account. We found no evidence of a clear
relationship between PPI price and quality in either PLPPI or CCPPI and a relatively
weak relationship between price and the quality of MPPI.
5
32. Despite this dispersion in prices, and the apparent benefits to search and switching
that this would seem to confer, the incidence of substitution between PPI policies, or
combinations of PPI and credit, is low.
33. The claims ratios (the claims incurred and provided for in the year as a percentage of
net premiums earned) on PPI policies are low and, to the extent that consumers are
aware of this, consumers who would rationally buy this insurance are likely to be
those: who place a high value upon it and have a high aversion to risk, risk high
consequential losses in the event of being unable to make payments, and/or have a
belief that they are of above-average risk.
34. When analysed separately from credit, PPI distribution is highly profitable. We
calculated that the 12 largest distributors made profits after tax in excess of the cost
of capital on PPI of £1.4 billion in 2006, representing a return on equity (RoE) of
490 per cent. Of the total economic profits of £1.4 billion, PLPPI contributed
£645 million, CCPPI contributed £336 million, MPPI contributed £112 million, and
SMPPI contributed £126 million. Other forms of PPI made up the remaining
£147 million. The RoEs in 2006 for PLPPI, CCPPI, MPPI and SMPPI were similar to
the overall level of 490 per cent. We found that the results for 2007 were similar to
2006.
35. The evidence on how intermediaries sell PPI included: that an intermediary or
intermediary network typically sells one MPPI product or a small selection of MPPI
products (sometimes referred to as a panel of products); that the MPPI offering for
each intermediary does not change very often; rather intermediaries tend to sell the
same PPI products for a considerable period; and that the level of commission is
used as a marketing tool for MPPI distributors or underwriters and intermediary
networks in attracting intermediaries to sell their PPI.
36. We considered that these findings were consistent with our findings that there is
limited substitutability between PPI policies provided by different PPI providers and
that PPI distributors and intermediaries are not competitively constrained by the
potential for customers to substitute between their PPI policies. The evidence
therefore supported our market definition.
37. These findings indicated to us there was less competition on prices and other
competitive variables between PPI providers than would be the case in a wellfunctioning market with vigorous and effective rivalry between market players.
Factors affecting the nature and extent of competition among distributors and
intermediaries
38. Having established that there is little competition among distributors and intermediaries in relation to the supply of any type of PPI policy sold at the point of sale, we
looked at why this is the case and what factors are involved. The purpose of this was
to see if there are any features or combination of features of relevant markets which
are preventing, restricting or distorting competition in the supply of PPI to consumers.
We considered retail PPI separately from other forms of PPI (see paragraph 67).
39. We focused on five factors which might contribute to the lack of competition:
(a) the extent to which distributors or intermediaries actively seek to compete with
each other;
(b) whether comparing PPI policies is beneficial and whether there are barriers to
search;
6
(c) having taken out PPI, whether consumers switch between PPI policies, and
whether there are barriers to switching;
(d) whether there are barriers to entry into the supply of PPI by new distributors and
intermediaries, or barriers to expansion by PPI providers already in the market;
and
(e) whether there is an advantage for distributors or intermediaries to selling
combinations of PPI and credit at the point of sale (the point-of-sale advantage)
and whether this contributes to the lack of competition we found.
40. We then looked at whether some stand-alone PPI and short-term IP products which
have recently been introduced to the market might impose a competitive constraint in
the short to medium term.
The extent of competition between providers
41. The evidence we saw indicated that PPI distributors and intermediaries do not
compete with one another based on the price of PPI. Nor did we find compelling
evidence of competition on non-price factors. In addition, we saw little evidence of
advertising of PPI to customers (as opposed to details of the policies being available
on distributors’ websites)—the evidence suggested to us that the point of sale of
credit was the time PPI was advertised to customers.
Benefits of, and barriers to, search
42. Next we looked at whether there are benefits to consumers in comparing PPI policies
and whether, given that we found that there is only limited search going on, there are
barriers to search. We found that there are benefits to comparing policies—the terms
and conditions and prices of policies vary considerably. It is therefore worthwhile
searching for the best-value product; however, search on price alone will not
guarantee that the best value product is purchased.
43. In terms of consumer search for PPI policies, we identified five significant barriers to
effective search, which impact consumers to different extents depending on the type
of PPI policy involved.
44. The first barrier we identified was that it is time-consuming to obtain accurate quotes,
with some firms only providing accurate price illustrations for PPI by going through a
full credit application. The making of a full credit application will also have an impact
on the credit ‘footprint’ of the customer.
45. The second barrier we identified was that it is difficult to make comparisons with the
information currently available. Price comparison websites, whilst useful, are limited
in their offer of comparisons. The websites focus on price, and there are several
combinations (such as bundles of PPI and credit) on which it remains difficult to
search.
46. The third barrier we found was the complexity of PPI policies. Variations in pricing
structure (particularly, but not only, with respect to single-premium policies), policy
terms and conditions, and the manner in which information is provided by firms,
mean that the cost of PPI is not presented in such a way that it is easy to make
comparisons, and this has a detrimental effect on consumers’ ability to understand
that information.
7
47. The fourth barrier was that significant numbers of consumers perceive either that PPI
take-up will have a positive influence on their credit application process or that it is a
condition for taking the credit. Whilst they think that this is the case, they incorrectly
perceive that they have a very limited incentive to search around for alternatives to
the combination of credit and PPI offered by the distributor, such as sourcing the
credit product from one supplier and searching separately for the best PPI policy.
48. The final barrier we identified was the relatively low level of stand-alone provision.
This restricts consumers’ ability to search for PPI on a stand-alone basis (although it
is still possible to search for combinations of PPI and credit).
49. We found that all these barriers were present when consumers search for PLPPI,
MPPI and SMPPI. We concluded that all of these barriers, except for the first barrier
(that obtaining quotes was time-consuming), were present when consumers search
for CCPPI. The effect of these barriers to search is to impede the ability of
consumers to make comparisons, and therefore effective choices between PPI
policies. They also, therefore, act as barriers to expansion for other PPI providers, in
particular providers of stand-alone PPI.
Barriers to switching
50. We identified four significant barriers to switching in addition to the barriers to search
that we identified. We found that access to consumers’ credit information was
restricted to the credit provider and that, for CCPPI, this meant that other PPI
providers could not offer insurance which would track the outstanding balance. It is
therefore difficult for stand-alone PPI providers to offer as high quality a product as
the point-of-sale distributor, because of this inability to track the balance.
51. The second barrier we identified was that the initial exclusion period at the commencement of PPI policies acts as a deterrent to switching. When a policy
commences there is usually an initial period during which claims cannot be made. A
consumer with a PPI policy who considers switching to another PPI provider has to
be willing to, in effect, not be insured for the duration of the exclusion period.
52. The third barrier we identified was the existence of exclusions for pre-existing
conditions. This could be a barrier to switching for a consumer who has developed a
condition whilst already having a PPI policy. Whilst the consumer may be covered for
that condition under his existing policy, moving to a new policy may result in the
condition being considered a pre-existing condition and the new insurance policy
excluding claims for that condition.
53. The final barrier to switching we identified was the rebate policy on single-premium
policies. Rebates are not given on a pro-rata basis; they take account of the higher
risks we were told underwriters face earlier in the life of a PPI policy. As a result, if a
consumer cancels a PPI policy, the rebate given is not enough to take out an
identical policy to cover the remaining balance for the rest of the term of the loan. We
calculated that the shortfall could be significant, amounting to between 50 and nearly
80 per cent of the cost of an identical replacement policy.
54. We concluded that two of these barriers—initial exclusion periods and exclusions for
existing conditions—applied to all types of PPI policies. In addition, the lack of access
to customers’ balance information was a barrier to switching for CCPPI policies, and
rebate profiles for single-premium policies was a barrier to switching for consumers
of PLPPI and SMPPI. These barriers to switching limit consumer choice. They also,
therefore, act as barriers to expansion for other PPI providers, in particular providers
of stand-alone PPI.
8
Barriers to entry and expansion
55. We considered whether there are barriers to entry or expansion, either for providers
selling stand-alone PPI or short-term IP, or distributors selling PPI combined with
credit. We concluded that in addition to the barriers to search and switching that we
have identified, there are three significant barriers to entry and expansion for
providers of stand-alone PPI.
56. The first barrier we identified was adverse selection. Providers of stand-alone PPI are
at greater risk of being asked for cover by consumers who think they are likely to
experience an event for which they could make a claim than is the case with
distributors and intermediaries supplying PPI at the point of sale of credit. This is
because at the point of sale distributors and intermediaries can try and sell PPI to
everyone taking out credit. These will include some consumers that may think a
claimable event is likely to happen soon, as well as consumers who want the
insurance for events they do not currently expect to happen. We received evidence
from a party which had entered the stand-alone market, and subsequently exited it,
showing that PPI claims in its stand-alone business were significantly higher than
those in its point-of-sale business.
57. The second barrier we found was poor consumer awareness of PPI. We found that a
significant number of consumers did not consider PPI before approaching their
lender for credit; moreover, a significant number of customers did not know that they
could take out PPI from someone other than their credit provider. This low consumer
awareness and poor understanding of options restricts the ability of providers of
stand-alone PPI successfully to enter or expand into PPI markets.
58. The third barrier to entry we identified was that, in order to attract customers in
sufficient numbers to purchase stand-alone PPI, there are high marketing costs
relative to those of linked credit and PPI sales by credit providers. The evidence we
received suggested that marketing campaigns by providers of stand-alone PPI have
had only limited success.
The point of sale
59. We considered next whether the ability of credit providers to sell PPI at the point of
sale gave them an advantage and whether this impeded competition. We concluded
that there was a point-of-sale advantage associated with selling PPI combined with a
credit product. We found that distributors focused their PPI marketing at the point of
sale, and at current prices and the current market structure this appears to be the
only effective way of marketing PPI. This means that providers of stand-alone PPI
without access to consumers at the point of sale are at a competitive disadvantage.
60. We found that there are economies of scope for distributors and intermediaries
associated with selling both credit and PPI at the point of sale, although this appears
to lead to distorted spending patterns by consumers for both PLPPI and credit.
Further, some distributors offer PPI only at the point of sale, and others do not
promote its availability after the credit product has been sold. This means that
consumers who have not researched PPI before approaching their lender can feel it
is ‘now or never’ for this PPI policy, and may feel it is safest to buy it.
61. We concluded that there are barriers to effective competition associated with the sale
of PPI at the credit point of sale, and that these affected: the ability of both other
distributors and intermediaries, and providers of stand-alone PPI to compete for
customers; and consumers’ choice.
9
New entry
62. Having found that there are some significant barriers to effective competition, we
considered whether this situation might change in the foreseeable future. New shortterm IP policies have been launched as stand-alone insurance policies during the
course of this investigation, and we considered whether they might be expected to
develop so as to offer a competitive constraint on distributors and intermediaries that
supply PPI at the point of sale.
63. However, we concluded that we did not expect them to do so. Evidence of the
difficulties experienced by new entrants in the past indicated that, in the current
market structure, there are significant difficulties associated with building sufficient
scale to achieve a sustainable and profitable business without access to customers
at the point of sale.
Indicators of the extent of competition between retail PPI providers
64. We assessed a series of indicators of the level of competition between retail PPI
distributors:
(a) The level of price variation over time that we saw was consistent with there being
few significant competitive pressures on retail PPI providers. In particular, retail
PPI prices have only been changed three times in six years by the four largest
distributors of retail PPI; two of these distributors did not change their prices at
all.
(b) We found that distributors do not compete to any significant degree on the nonprice aspects of retail PPI. We found that many of the changes to non-price
factors that we identified were due to regulatory or compliance issues and that
there was little evidence that changes in non-price factors were driven by, or
reactions to, competitors’ retail PPI offerings.
(c) We found that retail PPI distribution is highly profitable. Commission levels are
high in relation to the costs incurred in selling PPI and the capital requirements
are low, reflecting the low-risk nature of the activity.
65. These indicators suggested that there was a lack of competition in retail PPI
distribution.
66. In terms of competition between retail PPI providers, these findings indicated that
prices, and other competitive variables, are less favourable than would be the case in
a well-functioning market with vigorous and effective rivalry between market players.
We considered that these findings were consistent with our finding that there is
limited substitutability between retail PPI policies provided by different providers and
that retail PPI providers are not competitively constrained by the potential for
customers to substitute between their PPI policies. The evidence on the extent of
competition between retail PPI providers therefore supported our market definitions.
Factors affecting the nature and extent of competition among retail PPI
distributors
67. We considered the factors affecting the nature and extent of competition among retail
PPI distributors. We found that, as with other forms of PPI, most distributors fail
actively to seek to win retail PPI customers by using the price or quality of their retail
PPI product as a competitive variable. We found that there are benefits for
10
consumers in searching, but that customers face significant barriers in comparing
products. If the benefits of policies were clearer to consumers, and if retail PPI were
available separately from merchandise cover, distributors would have a stronger
incentive to compete with each other to show consumers that their retail PPI product
offer was the best value. In addition to the costs of search already found, we
concluded that there were barriers to switching associated with: the inability of standalone providers to offer insurance which tracks the outstanding balance on a credit
account, or to offer merchandise cover; initial exclusion periods on retail PPI policies;
and exclusions for pre-existing conditions. We concluded that there are significant
barriers to entry and expansion for stand-alone PPI providers seeking to sell retail
PPI products without access to customers at the credit point of sale and to customer
account information. We found that the sale of retail PPI at the initial point of sale and
continued exclusive access to customer accounts restricts the extent to which other
PPI providers can compete effectively.
The underwriting of PPI
68. We concluded that the underwriting market is relatively concentrated, and that some
distributors think that not all underwriters are sufficiently large to be able to underwrite their business. There do not appear to be significant barriers to entry or
expansion. However, switching costs may be significant and incumbent providers of
underwriting services do appear to have an advantage over other underwriters when
re-tendering for a contract.
69. On the other hand, the tender process leads to underwriters competing vigorously
with each other in order to win contracts that will give them access to a significant
number of customers at the point of sale of a credit product. The evidence relating to
profitability that we have seen indicated that underwriters appear to be making
reasonable, but not excessive, rates of return on PPI business. This suggested to us
that underwriters were unable to exert a significant degree of market power.
70. We conclude, therefore, that there is a significant level of competition between
underwriters of PPI, and as a result there are no features of the relevant markets
which prevent, restrict or distort competition for the supply of PPI underwriting
services.
Vertical integration
71. Our analysis on vertical integration of PPI underwriters and distributors did not
identify issues which would prevent, restrict or distort competition in the underwriting
market to the detriment of consumers. We did not identify any issues with respect to
the distribution of PPI policies which are unique to vertically-integrated distributors.
Features
72. We conclude that there are features of the markets for PPI which, alone or in
combination, prevent, restrict or distort competition in the supply of PPI to nonbusiness customers in the UK. These are described in Sections 5 and 6 of the report
and may be summarized as:
(a) Distributors and intermediaries fail actively to seek to win customers by using the
price or quality of their PPI policies as a competitive variable.
(b) Consumers who want to compare PPI policies (including PPI combined with
credit), stand-alone PPI or short-tem IP policies are hindered in doing so. Product
11
complexity (the variations in pricing structures (in particular in relation to singlepremium policies) and in terms and conditions, the way information on PPI is
presented to customers); the perception that taking PPI would increase their
chances of being given credit; the bundling of PPI with credit; and the limited
scale of stand-alone provision act as barriers to search for all types of PPI
policies. The bundling of retail PPI with credit accounts and with merchandise
cover (also known as purchase protection insurance) acts as a barrier to search
for retail PPI. In addition, the time taken to obtain accurate price information is a
barrier in relation to the provision of PLPPI, MPPI and SMPPI. These barriers to
search impede the ability of consumers to make comparisons, and therefore
effective choices between PPI policies. They also, therefore, act as barriers to
expansion for other PPI providers, in particular providers of stand-alone PPI.
(c) Consumers who want to switch PPI policies to alternative PPI providers or to
alternative insurance products are hindered in doing so. Terms which make
switching expensive (in the case of single-premium policies) act as barriers to
switching for PLPPI and SMPPI policies. Terms which risk leaving consumers
uninsured (for a short period of time or in case they suffer a recurrence of a
condition) act as barriers to switching for all types of PPI policies. In addition, the
lack of access to consumers’ balance information acts as a barrier for switching
for CCPPI and retail PPI, and the bundling of retail PPI with merchandise cover
acts as a barrier to switching for retail PPI. These barriers to switching limit
consumer choice. They also, therefore, act as barriers to expansion for other PPI
providers, in particular providers of stand-alone PPI.
(d) The sale of PPI at the point of sale further restricts the extent to which other PPI
providers can compete effectively.
73. We find, pursuant to section 134(1) of the Act, that there are features of relevant
markets which alone or in combination, prevent, restrict or distort competition in the
supply of PPI and accordingly, there is an AEC within the meaning of section 134(2).
The features are those that are discussed in Sections 5 and 6 of this report and
which are summarized in paragraph 72.
Detriments
74. We concluded that the detrimental effects on consumers of these features in relation
to all types of PPI (except retail PPI) were higher prices for, and less choice in, PPI
policies than would be expected in a well-functioning market. We also concluded that
demand for PPI was distorted. We also concluded that it was possible that there was
less innovation than would be expected in a well-functioning market. For retail PPI
we concluded that the detrimental effects on consumers of these features were
higher prices for retail PPI policies than would be expected in a well-functioning
market. We also concluded that it was possible that there was less innovation than
would be expected in a well-functioning market.
75. We considered the scale of the customer detriment associated with the AEC we
found. First, there is a large category of ‘dynamic’ benefits to consumers that we
would expect to arise from increased competition in the provision of PPI. Such
benefits will arise, for example, from arresting the decline in the size of the PPI sector
that results from the current lack of competition (for example, negative publicity
associated with high prices). Indeed, we would expect greater competition to bring
about increased advertising and far more interest in (and awareness of) the sector,
such that the demand for PPI should increase, once it is sold at competitive prices. A
further example of the benefits we would expect to accrue is from selection pressure,
encouraging companies that develop products which benefit consumers and
12
punishing those that develop poor products. Given the considerable size of the PPI
sector even at the current high prices, we would expect these dynamic benefits of
competition to be on a large scale, but we have not been able to put a value upon
them.
76. Further, there are ‘static’ welfare implications of the current high PPI prices—
inefficiencies associated with high PPI prices and low credit prices (ie the ‘deadweight losses’ that stem from people not buying PPI at high prices who would buy it
at competitive prices and, similarly, people being offered credit at lower prices than
would be the case if PPI profits were not being used to fund the sale of credit). We
noted that a distortion in credit prices is not intrinsically beneficial. Even if we
assumed that all PPI profits are used to fund lower credit prices, we found that these
considerations implied an annual net deadweight loss in PLPPI, MPPI and SMPPI in
excess of £200 million a year, on the basis of our analysis of 2006 figures. We noted
that a lower degree of pass-through of PPI profits than the full 100 per cent that
underpins the figure above would imply greater potential ‘static’ gains from
competition for consumers—for example, if one-fifth of the profits from PPI were not
passed through in the form of lower credit prices, our lower-bound reasonable
estimate for these static gains in PLPPI, MPPI and SMPPI would rise to £440 million
a year. We were unable to make any estimate of the static consumer detriment for
CCPPI, but note that the excess profits we identified in the sector (£336 million in
2006) is consistent with substantial further static consumer detriment.
77. Further to this, we also identified a further static detriment that we are unable to
quantify. This is that high PPI prices are likely to have resulted in adverse selection in
the markets for PPI, resulting in increased claims costs on PPI policies and increased
impairment costs on credit sold to PPI customers, compared with the levels that
would arise given the lower PPI price levels that we would expect in a wellfunctioning market. A further detriment to consumers as a result of high PPI prices is
therefore the increased costs of supplying PPI at high PPI prices due to adverse
selection. We were unable to quantify the scale of the adverse selection problem in
the supply of PPI. Our calculation of the static detriment to consumers is therefore an
underestimate, as it does not include the effects of adverse selection.
78. We concluded, therefore, that the total consumer detriment (both static and dynamic)
to be addressed would be significantly more than the £200 million per year of static
detriment a year calculated assuming that all PPI profits were passed through to
consumers as lower credit prices.
Remedies
79. We decided that the following package of remedies would form as comprehensive a ,
solution as is reasonable and practicable to the AEC and detrimental effects on
customers resulting from the AEC that we had identified:
(a) A prohibition on selling PPI at the credit point of sale. PPI cannot be sold by the
distributor or intermediary arranging the credit (or any business covered by the
prohibition, see paragraph 10.127) at the same time as the credit product, nor
within seven days of the conclusion of the credit sale period, or the provision of a
personal PPI quote, if one were not provided during the credit sale period. As a
limited exception to this point-of-sale prohibition, the distributor or intermediary
arranging the credit (or any business covered by the prohibition) may sell PPI to
the consumer over the Internet or telephone 24 hours after conclusion of the
credit sale period provided that the consumer has initiated the transaction and the
consumer has confirmed that they have seen the personal PPI quote (paragraphs
10.34 to 10.156).
13
(b) Provision of a personal quote. We decided that all distributors and intermediaries
who arrange credit for consumers must provide a personal PPI quote to the
consumer in a durable medium (if the distributor or intermediary provides
information about PPI to the consumer during the credit sale period). If the
distributor or intermediary arranging the credit does not provide a personal PPI
quote during the credit sale period, but subsequently contacts the consumer to
offer PPI, a personal PPI quote must be provided at that time. Stand-alone
providers and providers of short-term IP are required to provide a personal PPI
quote to the consumer in a durable medium if the consumer asks the provider
about the cost and/or features of a stand-alone PPI and/or short-term IP policy
sold by that provider (paragraphs 10.157 to 10.181).
(c) Information provision in marketing materials. We decided to impose a requirement on all PPI providers prominently to disclose the following information in any
PPI marketing materials that include pricing claims or cost information, any
indication of the benefits of the PPI product or its main characteristics: the
monthly cost of PPI per £100 of monthly benefit 3 (CCPPI and retail PPI providers
must also show the cost of PPI per £100 of outstanding balance); that PPI is
optional (stand-alone providers do not have to include this statement) and
available from other providers (without specifying those other providers); and that
information on PPI, alternative providers and other forms of protection can be
found on the Financial Services Authority’s (FSA) moneymadeclear website
(paragraphs 10.182 to 10.222).
(d) Provision of information to third parties. We decided to require that all PPI
providers must provide comparative data to the FSA, as specified by, and in the
format requested by, the FSA. In addition to the information that the OFT may
request from time to time for the purposes of monitoring and reviewing the
operation of the remedies package, all PPI providers that meet a specified
threshold (see paragraph 10.538) must provide the following information to the
OFT on an annual basis: annual GWP, split by product type; distributor
penetration rates, split by product type; and aggregate claims ratios for each
provider, split by product type. In addition, all PPI providers must provide to any
person on request, aggregate claims ratios, split by product type, for the previous
year. These can be provided in the form of a range to be specified by the CC
(paragraphs 10.223 to 10.242).
(e) Recommendation to use information for price comparison tables. We decided to
make a recommendation to the FSA that it use the information provided to it
pursuant to this remedies package to populate its PPI price comparison tables
with data on all PPI and short-term IP products (paragraph 10.517).
(f) A prohibition on the selling of single-premium PPI policies. We decided that PPI
cannot be charged on a single-premium basis. Subject to the prohibition on
charging PPI on a single-premium basis, premiums can be charged monthly or
annually. Where an annual premium is paid by a consumer, then a rebate must
be paid to consumers on a pro-rata basis if the consumer terminates the policy
during the year. No separate charges can be levied on a customer for
administration or for the set-up or early termination of a PPI policy (paragraphs
10.243 to 10.277).
3
If the benefit pays out for less than 12 months, notice of this fact must also be clearly disclosed to consumers alongside the
cost of the policy.
14
(g) A requirement to unbundle retail PPI from merchandise cover. Where distributors
of retail PPI offer an insurance package containing PPI and merchandise cover,
they must also offer, as a separate item, PPI cover alone (paragraphs 10.278 to
10.301).
(h) Annual statements. We decided to place a requirement on distributors, intermediaries and stand-alone providers to provide an annual statement for PPI
customers (paragraph 10.302 to 10.332). Provision of this statement will be the
responsibility of the company that sold the PPI policy to the consumer, other than
for sales made by intermediaries where provision of this statement will be the
responsibility of the underwriter or distributor or stand-alone provider with whom
the consumer has an ongoing relationship.
80. We decided that all of the elements of the remedies package, except for the recommendation to the FSA, should apply to all distributors and intermediaries, and where
relevant underwriters. Providers of stand-alone PPI and short-term IP would be
bound by parts of the remedy package, in particular: the requirements in to provide a
personal quote; provision of most of the information required in marketing materials;
provision of information to third parties; and provision of annual statements. Whilst
there are no current providers of stand-alone PPI and short-term IP policies that offer
these policies on a single-premium basis, stand-alone PPI or short-term IP providers
are also restricted to selling only monthly or annual premium policies, and are
prohibited from making separate charges for administration or for the set up or early
termination of a regular-premium PPI policy.
81. We decided that our remedies package should be implemented by means of an
Order, given the large number of PPI distributors active in the UK, except in relation
to the use by the FSA of information to populate its comparison tables, which we
decided should take the form of a recommendation. We concluded that the provision
of information in marketing materials and provision of information to third parties
should be implemented within six months of the CC’s Order and that all other parts of
the remedy package could be implemented for new PPI policies within 12 months of
the entry into force of the Order.
82. The OFT has a statutory role in relation to the monitoring of remedies. The FSA has
responsibility for oversight of much of the financial services sector, and the FSA and
OFT told us that they intended to work together to minimize duplication of effort and
burden on business in carrying out their functions.
83. We considered whether all parts of the remedy package were required, or whether
alternative remedies and remedy packages put forward by parties would be effective
at remedying the AEC identified at lower cost. We concluded that the remedy
package summarized in paragraph 79 was required in full in order effectively to
address the AEC identified. Alternative remedies, and remedy packages, put forward
by parties would not, in our view, be effective in remedying the AEC identified. We
concluded, therefore, that our remedy package was the least-cost, least-intrusive
package in addressing the AEC. Further, we concluded that requiring implementation
of all elements of the package was necessary in light of the detriment identified.
84. We considered whether there were any relevant customer benefits which we should
take into account. Parties put to us that there were several relevant customer
benefits arising from aspects of the current market structure, including that PPI and
credit prices were lower than they otherwise would be and that there were benefits
arising from selling PPI as single-premium policies.
15
85. We concluded that credit prices, and credit cut-off scores, are lower than they
otherwise would be because of PPI income generated at the credit point of sale. For
credit cards the effect is small and for mortgages the effect is very small. We
concluded that these lower prices were a direct result of the distributors’ anticipation
of high profit margins on PPI. Lower credit prices are therefore a direct result of the
features of the sale of PPI that lead to an AEC in the markets for PPI.
86. We concluded, therefore, that there is a relevant customer benefit of lower credit
prices for personal loans (unsecured and secured), mortgages and credit cards. The
only credit products on which we thought that there might be an appreciable
reduction in credit prices were unsecured and secured personal loans, and we could
not be confident that the scale of the relevant customer benefit that we observed in
the period up to December 2006 would persist at that level in the future. We did not
find any other relevant customer benefits.
87. We considered whether we should exercise our discretion with regard to relevant
customer benefits and modify our remedies to preserve the relevant customer benefit
of lower credit prices. We noted that a distortion in credit prices is not intrinsically
beneficial. We noted that, if we imposed remedies, PPI consumers may be expected
to benefit but consumers of both credit and PPI would face higher prices for their
credit products; if we did not intervene, PPI consumers would continue to face high
PPI prices. We therefore considered whether our remedies might be expected to
have a positive or negative impact on total consumer welfare.
88. We found that, for MPPI and SMPPI, any effective intervention would have an overall
positive effect on total consumer welfare. We inferred from our results that we should
reach the same conclusion for CCPPI. For PLPPI, we noted that under some
circumstances (where a number of fairly extreme assumptions in our calculations
combine) when considered in purely static terms intervention might not be welfareenhancing. However, in our judgement the circumstances giving rise to a negative
net consumer detriment were very unlikely to occur, and we were confident that our
remedies would increase both search and the extent to which distributors and
intermediaries actively seek to win customers using price as a competitive variable.
89. We therefore concluded that we should not modify our remedies to preserve the
relevant customer benefit of lower credit prices or cut-off scores.
16
Findings
1.
Introduction
1.1 On 7 February 2007 the OFT referred the supply of all PPI (except store card PPI) to
non-business customers in the UK to the CC for investigation under section 131 of
the Act. The terms of reference for our investigation are set out in Appendix 1.1. This
document sets out our findings from this investigation.
1.2 Our inquiry into the supply of PPI is a market investigation under the Act. Section
134(1) of the Act requires us to consider whether ‘any feature, or combination of
features, of each relevant market 4 prevents, restricts or distorts competition in
connection with the supply or acquisition of any goods or services in the UK or a part
of the UK’. If there is such a feature or combination of features, there is said to be an
‘adverse effect on competition’ pursuant to section 134(2) of the Act.
1.3
Under section 131(2) of the Act, a ‘feature’ of a market may refer to:
(a) the structure of the market concerned or any aspect of that structure;
(b) any conduct (whether or not in the market concerned) of one or more than one
person who supplies or acquires goods or services in the market concerned; or
(c) any conduct relating to the market concerned of customers of any person who
supplies or acquires goods or services.
1.4 Conduct for these purposes includes any failure to act (whether or not intentional)
and any other unintentional conduct 5 .
1.5 If the CC decides that there is an AEC, it is required under section 134(4) of the Act
to decide whether action should be taken by it, or whether it should recommend the
taking of action by others, for the purpose of remedying, mitigating or preventing the
AEC concerned or any detrimental effect on customers 6 so far as it has resulted
from, or may be expected to result from, the AEC; and, if so, what action should be
taken.
1.6 This section sets out the roles of the different regulatory bodies in the oversight of
PPI (paragraphs 1.7 to 1.11), the events leading up to this investigation (paragraphs
1.12 to 1.14), provides an overview of the conduct of the investigation to date as well
as the proposed next steps (paragraphs 1.16 to 1.23), details the publication of
evidence relevant to the investigation (paragraphs 1.24 and 1.25) and finally, sets out
the structure of the remainder of these findings (paragraphs 1.26 to 1.34).
The roles of the FSA, the OFT and the CC in the oversight of PPI
1.7 The FSA is an independent non-governmental body which was given statutory
powers by the Financial Services and Markets Act 2000. It has a wide range of rulemaking, investigatory and enforcement powers to enable it to meet its statutory
4
A ‘relevant market’ is defined in section 134(3) of the Act as a market in the UK for goods or services of a description specified in the reference concerned. 5
Section 131(3) of the Act. 6
Section 134(5) of the Act defines a detrimental effect on customers (or future customers) as taking the form of: (a) higher prices, lower quality or less choice of goods or services in any market in the UK (whether or not the market to which the feature or features concerned relate); or (b) less innovation in relation to such goods or services. 17
objectives, which it summarizes as one overall aim: to promote efficient, orderly and
fair markets and to help retail consumers achieve a fair deal.
1.8 The FSA became responsible for the regulation of non-investment insurance sales
and administration, including PPI, on 14 January 2005. The FSA identified the PPI
market as an early thematic priority in its 2005/06 business plan, due to the concerns
about that market which had been expressed by the media, consumer groups and
other organizations. 7 The FSA’s thematic work was focused on the level of compliance with its rules 8 by firms selling PPI with credit. In response to shortcomings
identified in the ongoing thematic work, the FSA developed the new Insurance
Conduct of Business Sourcebook (ICOBS), which came into force in 2008. The new
ICOBS contains rules specifically designed to address information shortfalls before
and after the sale of PPI (see Appendix 2.5).
1.9 The OFT aims to ensure that markets work well for consumers. The OFT does this in
several ways, including by studying markets and recommending action where
required. The OFT gathers intelligence about markets and trader behaviour from a
wide range of sources and responds to super-complaints about markets from
designated consumer bodies. Where it spots potential problems, it undertakes
market studies and recommends or takes further action where it is needed.
1.10 The OFT may make a reference to the CC where it has reasonable grounds for
suspecting that any feature or combination of features of a market in the UK for
goods or services prevents, restricts or distorts competition in connection with the
supply or acquisition of any goods or services in the UK or a part of the UK. The OFT
made such a reference in relation to PPI.
1.11 The CC’s role in a market investigation is to help ensure healthy competition between
companies in the UK for the benefit of those companies, their customers and the UK
economy. Accordingly, the roles of the FSA and the CC, while different, are complementary. The FSA is concerned with the process by which distributors supply and
sell PPI to consumers, while the CC is concerned with the extent of competition in
the PPI industry.
Events leading up to this investigation
1.12 On 13 September 2005 the OFT received a super-complaint 9 from Citizens Advice,
which is a designated consumer body, about PPI. Further details are provided in
Appendix 1.2.
1.13 The OFT responded to the super-complaint by undertaking to conduct a market
study.
7
FSA, The sale of payment protection insurance—mystery shopping results, prepared for the FSA by GfK NOP, November
2005, p1.
8
The FSA assessed firms’ compliance with its ICOB rules; training and competence rules; rules on systems and controls; rules
set out in the ‘Principles for Businesses’, particularly Principle 6, which requires that ‘a firm must pay due regard to the interests
of its customers and treat them fairly’.
9
The right to submit a super-complaint was created by section 11 of the Act. A super-complaint is defined under section 11(1) of
the Act as a complaint submitted by a designated consumer body that ‘any feature, or combination of features, of a market in
the UK for goods or services is or appears to be significantly harming the interests of consumers’.
18
1.14 On 19 October 2006 the OFT signalled its intention to refer the UK PPI industry to
the CC. It published a document 10 for consultation, and following the end of that
consultation period on 30 November 2006, the OFT made its referral to the CC on
7 February 2007.
Conduct of the investigation
1.15 The CC began the market investigation into the supply of PPI following the referral
from the OFT on 7 February 2007.
1.16 Since commencing this investigation we have been careful to ensure that our
processes have been both thorough and fair. In this respect we have had regard to
the CC’s published guidelines on market investigations 11 and other published
guidance.
1.17 We faced some challenges during this investigation primarily with regard to the
difficulty and/or reluctance of some parties to provide us with evidence and data we
requested. This led to significant delays in the information-gathering process, and of
course delayed our analysis of the evidence we ultimately received.
1.18 Due to these delays it was considered appropriate at various stages during the
investigation to utilize our formal information-gathering powers, pursuant to section
109 of the Act, to require the provision of data within a specified time period. 12 We
issued 12 section 109 notices for the provision of information. 13 There were three
occasions when we notified the recipient of a section 109 notice that its response had
not fully complied with the terms of the notice. However, those parties addressed our
concerns about their non-compliance promptly and completely and we elected not to
take action as a result of the non-compliance on those occasions.
1.19 The following paragraphs set out an overview of the PPI inquiry process and provide
an understanding of how we utilized the evidence, data and information we received.
Further details can be found in Appendix 1.3.
1.20 We published an issues statement on 12 April 2007 and our Emerging Thinking on
6 November 2007, along with nine working papers. These were followed by 12
further working papers. We published our provisional findings on 5 June 2008 and
our provisional findings on retail PPI on 10 October 2008. On 13 October 2008 we
consulted on further technical analysis that we had carried out to aid decision-making
on remedies, excluding retail PPI, along with a further working paper. In the context
of remedies we consulted on the notice of possible remedies (5 June 2008), the CC’s
provisional decision on remedies (13 November 2008); the provisional decision on
retail PPI remedies (21 November 2008); and the application to intermediaries of the
13 November provisional decision on remedies (12 December 2008). Appendix 1.3
10
OFT, Payment protection insurance, Report on the market study and proposed decision to make a market investigation
reference, October 2006.
11
Market Investigation References: Competition Commission Guidelines, CC3, June 2003, available on the CC’s website at
www.competition-commission.org.uk/rep_pub/rules_and_guide/pdf/cc3.pdf.
12
The CC’s ability to fulfil its statutory functions and publish a report within the statutory deadline hinges on the timely receipt of
information from parties. The CC has formal information-gathering powers: section 109 of the Act gives the CC power to issue a
notice requiring the production of information or documents, or requiring specified persons to give evidence to the CC. Failure
to comply with a notice issued under section 109 of the Act within the time specified in that notice can lead to penalties being
imposed.
13
We also issued a further 29 section 109 notices to parties’, at their request. The parties made the request when they were
asked to provide PPI customers’ details to us for surveys, in order to protect their position with regard to the Data Protection
Act. We understand from the Information Commissioner’s office that such protection is not necessary in circumstances where
the CC has made a request for information, and so those notices were issued for the comfort of the requesting parties, not as a
necessity.
19
contains a full list of the working papers, and we discuss our approach to publication
of evidence more generally in paragraphs 1.24 and 1.25. We have conducted a
range of hearings and visited a range of PPI underwriters and distributors.
1.21 We made use of a variety of financial industry publications and data sources during
the investigation. These included Mintel UK Credit Insurance reports, Defaqto
payment protection insurance reports 14 , FSA reports and surveys, OFT reports and
surveys, PPI Insight and Datamonitor reports.
1.22 We undertook qualitative research, two quantitative surveys and a quantitative
experiment among PPI customers during the investigation. The qualitative research
by market research agency BMRB was to explore customers’ perceptions of the PPI
market and their experience in becoming a PPI customer. The first quantitative
survey was undertaken on our behalf by BMRB, and was aimed at understanding
customer motivations and behaviours. This involved interviewing by telephone 2,283
customers who had recently purchased PPI. Our final survey was commissioned
from GfK NOP. Its purpose was to help understand the extent to which PPI
customers search for insurance products for loans and mortgages. GfK NOP
undertook a two-stage process: first, 6,800 telephone interviews were conducted with
customers who had purchased PPI between January 2007 and January 2008, to
establish the extent of searching for stand-alone PPI policies and combinations of
PPI and credit prior to a credit application being made; and second, 258 of those
consumers who said they had compared PPI policies took part in a follow-up
interview which was conducted face to face in the consumer’s home, to investigate in
detail their attitudes and behaviours. This final survey was undertaken in February/
March 2008. Following publication of our provisional findings Synovate conducted
two complementary studies. The first, a qualitative study involved interviewing in
depth 24 customers to gain insight into the extent of consumer understanding of the
costs associated with PPI and explore understanding of, and preference for, different
ways of presenting quotation information. This study was undertaken in August 2008.
The study prepared the ground for a quantitative experiment, interviewing over 300
customers in their homes and asking them to perform a computer-based choice
experiment. Synovate withdrew the results of the quantitative experimental research
following quality assurance of its experimental research fieldwork. We did not draw
upon the experimental results in making the provisional or final decision on remedies.
1.23 The working papers, provisional findings and provisional decisions on remedies we
published reflected our position at the time of publication of the issues relevant to our
investigation. We took the responses we received to these documents into account
as we moved towards this Report. Our working papers, provisional findings and
provisional decisions on remedies were indicative of the work in progress on the
investigation, and are distinct from the conclusions, which we subsequently reached
in this Report.
Publication of evidence and other materials
1.24 During this investigation we published a range of material on the CC’s website. This
included evidence submitted by main and third parties (including non-sensitive
versions of parties’ written submissions and responses to Emerging Thinking),
reports from the survey-based research that we commissioned, our Emerging
Thinking document and associated working papers, and the further working papers
which have been prepared since then.
14
Defaqto is a provider of financial services research; since 2001 it has been producing star ratings for a variety of financial
products.
20
1.25 Our policy in this investigation, as with other recent CC inquiries, has been to publish
our working papers in full to inform discussion and debate, excluding only material
the disclosure of which may have been damaging to a party. The Act requires the CC
to have regard to the need to exclude from disclosure any information the disclosure
of which might significantly harm legitimate commercial or individual interests.
Subject to that condition we sought throughout the investigation to follow a policy of
making as transparent as possible the material we received and published. In doing
so, our aim was to stimulate discussion and debate of the various issues which were
being considered.
Report overview
1.26 Section 2 of this report sets out the background to the supply of PPI, covering both
the underwriting of PPI and its distribution.
1.27 The bulk of the report focuses on the distribution of PPI, which is where the concerns
raised by the OFT leading to the reference were focused.
1.28 Section 3 considers the relevant market definitions for the distribution of PPI for the
purposes of this investigation. This includes analyses of: the responsiveness of PPI
demand to PPI prices; the constraint on a PPI provider by other PPI providers and
other types of insurance policies; and substitution between combinations of PPI and
credit.
1.29 Section 4 looks at various indicators of competition between PPI providers, focusing
on, the variation of PPI prices over time, indicators that competition may be taking
place on factors other than price, price dispersion, the level of search conducted by
consumers, the level of consumer switching, claims ratios, and an analysis of
profitability of the distribution of PPI.
1.30 Section 5 considers the factors which underlie our finding on the extent of competition, and whether there are factors which prevent, restrict or distort competition,
creating an AEC.
1.31 Section 6 considers the state of competition in the supply of retail PPI, considering
the relevant market definition, the various indicators of competition between distributors of retail PPI, the factors underlying our finding on the extent of competition, and
whether there are factors which prevent, restrict or distort competition, creating an
AEC.
1.32 Having considered the distribution of PPI in depth, the report looks at the
underwriting of PPI (Section 7) and whether any AECs arise from a PPI policy
distributed by a company being underwritten by a company within the same group
(vertical integration, Section 8).
1.33 We set out in Section 9 the features of relevant markets which, alone or in
combination, prevent, restrict or distort competition in the supply of PPI to nonbusiness customers in the UK.
1.34 Finally, in Section 10, the report sets out the remedial action we intend to take. This
includes the remedies required to remedy the AEC, our analysis of relevant customer
benefits, and our consideration of whether we should amend our remedy package to
preserve the relevant customer benefit found.
21
2.
PPI underwriting and distribution
What is PPI?
2.1 PPI provides money to a credit provider or a consumer should the consumer suffer
an event specified in the insurance contract (the risks covered by PPI are primarily
accident, sickness, unemployment and death, or a combination thereof). The money
paid to a credit provider is a function of the amount of money due to it (for example,
the monthly repayment on a mortgage or a percentage of the outstanding balance on
a credit card). The money paid to a consumer is either a function of amount of money
due to be paid to a credit provider, or more generally a pre-agreed sum of money (for
example, in the case of stand-alone PPI policies). The main types of PPI are PLPPI,
MPPI, CCPPI, SMPPI, retail PPI, and motor finance PPI (motor PPI). 15 Sales of
PLPPI, MPPI and CCPPI made up nearly 90 per cent of types of PPI policy sold in
2007. 16
2.2 In addition, a further form of PPI not referred to in the OFT’s report 17 was brought to
our attention: personal current account overdraft PPI (overdraft PPI). We estimated
that around £50 million was paid by customers for overdraft PPI in 2006 (see Table
2.1). Given how low sales of this product are, in particular relative to the other types
of PPI policy available, we considered it appropriate to focus our resources, and the
resources of PPI providers on the other types of PPI policies available, which make
up the vast majority of PPI sales. We do not, therefore, reach any conclusions with
regard to overdraft PPI.
2.3 The protection offered by PPI is often referred to as ASU coverage. As stated above,
many policies also include an element L in the event the policyholder dies (though
most MPPI policies do not include a life insurance element). Some PPI providers
offer the ability to choose some rather than all of the ASU and L elements
available. 18 U benefits are usually paid for up to 12 months, and AS benefits anything
up to 60 months (depending on the policy) after an initial exclusion period (which is
usually a specified number of days between the time a policy is purchased and the
time a customer is permitted to make a claim). Further details on the generic
characteristics of PPI policies can be found in Appendix 2.1, and details on individual
PPI policy types can be found in Appendix 2.2.
2.4 The payment method for PPI policies is either ‘single premium’ or ‘regular premium’.
See paragraphs 2.77 to 2.81 for more detail.
2.5 PLPPI is designed to cover personal loans (which are sometimes referred to as
unsecured personal loans and are typically arranged with monthly repayments over
the term of the loan). It offers cover for ASUL, or a combination thereof, and some
PPI providers also offer critical illness (CI) cover. It is usually paid for by way of a
single premium (see paragraph 2.77) which is added to the loan and on which
interest is paid (less than 10 per cent of PLPPI policies are paid for by monthly
premium).
15
PPI is also available to cover store cards, but this is outside the scope of the current investigation. See Table 2.1. 17
OFT, Payment Protection Insurance, Report on the market study and proposed decision to make a market investigation
reference, op cit. 18
For example, Northern Rock offered ‘Gold’, ‘Silver’ or ‘Bronze’ on PLPPI policies (Northern Rock, response to issues statement, paragraph 5.5). 16
22
2.6 MPPI is designed to cover first-charge mortgages. As noted in paragraph 2.3, the
risks usually covered are A, S and U, or a combination of these. It does not usually
offer L. 19 MPPI policies are paid for by way of a monthly premium.
2.7 CCPPI covers a percentage of the outstanding balance on a credit card. It usually
covers L, A, S and U, or a combination of these; some PPI providers also offer the
option of CI cover. The premium is paid monthly and is calculated as a percentage of
the monthly outstanding balance. The benefit payable is typically calculated as a percentage of either the outstanding balance on the credit card at the time of claim or
the monthly statement issued prior to the claim. This percentage can vary between
PPI providers, typically from 3 to 10 per cent.
2.8 SMPPI protects a second-charge mortgage (also known as a secured loan; it is
usually guaranteed on a customer’s home and as such constitutes a ‘second charge’
on that home). The principal risks covered are L, A, S and U or a combination of
these. 72 per cent of policies issued in 2007 were paid for by a single premium which
covers the term of the policy; the remainder were paid for by monthly premiums.
2.9 Retail PPI covers a percentage of the outstanding balance owing on a customer’s
retail finance account or loan; it insures against the risks of A, S, U, L and CI. It is
usually paid for either by monthly instalment or a single premium policy, depending
on the provider. Like CCPPI, retail PPI covers a balance that can change frequently,
as purchases are made using the underlying credit. However, unlike CCPPI, retail
PPI pays off the regular payment instalment due (in a way which is more akin to
PLPPI or MPPI).
2.10 Motor PPI covers repayments on the loan (or hire purchase agreement) for the
purchase of a motor vehicle; it usually covers A, S, U, L and CI, or a combination of
these. It is usually paid for by way of a single premium up front; however, 11 per cent
of policies sold in 2007 were paid for by a regular-premium payment. Motor PPI can
be sold by distributors (see paragraph 2.35) and intermediaries (see paragraph 2.42).
We considered that motor PPI is PLPPI which covers a personal loan taken for the
express purpose of buying a car, and therefore our conclusions on PLPPI apply also
to motor PPI.
2.11 Overdraft PPI protects the repayments a customer is required to make on an
overdraft facility. It usually offers cover against the risks of A, S, U, L and CI, or a
combination of these, and some policies also offer cover against disability and
hospitalization. It is paid for by way of a monthly premium, per person covered.
Short-term income protection
2.12 Short-term IP, sometimes advertised as offering a form of lifestyle cover, typically
offers ASU cover for up to 12 months. The customer will choose the amount for
which they wish to be covered (which could be the amount of a monthly credit
repayment or any other amount), and has the choice of how to use any benefit paid
to them.
2.13 Short-term IP is usually offered on a stand-alone basis—that is, not combined with
credit. However, one provider of short-term IP (HSBC) initially started selling this
product as a replacement for MPPI (and a life insurance product), and now offers to
consumers to whom it previously offered PLPPI and CCPPI the chance to discuss
their broader protection needs with one possible outcome being to take this product
19
The LifeChoices product which HSBC introduced to replace its MPPI policy offers the option of cover for L.
23
(if it is the most suitable product). In addition this product is available on a nonadvised basis through its telephone sales channel.
2.14 We considered whether short-term IP is in fact a form of PPI. Our analysis of this is
shown in Appendix 2.3. We concluded that short-term IP is a form of PPI, though we
recognize that it can be used for alternative purposes. We also concluded that the
sale of a short-term IP policy, such as HSBC’s LifeChoices product, as a result of a
referral made to the sales-person during a credit sale is a sale of that short-term IP
product at the point of sale. We therefore concluded that short-term IP sold in such
circumstances is PPI and enjoys the same advantages, and gives rise to the same
concerns, as PPI sold at the point of sale.
Possible alternative policies
2.15 There are alternative financial insurance policies available to consumers for use to
protect their credit payments, to which we were referred by parties to the
investigation as either being complements to, or substitutes for, PPI. Details of these
are set out in Appendix 2.4. The policies are described briefly below.
2.16 Income protection (IP) (as distinct from short-term IP, which is described in
paragraphs 2.12 to 2.14), sometimes known as permanent health insurance,
provides a regular payment if a consumer loses income because they are unable to
work due to accident or sickness. It is generally a long-term policy, individually
underwritten for the consumer, with benefits payable up to retirement.
2.17 Personal accident insurance provides a lump sum in the event of an accident which
results in loss of limbs or permanent disability within the terms of the policy. Some
policies also include life insurance and hospitalization cover.
2.18 CI insurance provides a lump sum in the event that a consumer develops an illness
specified in the policy.
2.19 Life insurance provides an agreed cash payment if the consumer dies during the
term of the policy. Policies can be ‘whole life’, where the term is effectively the life of
the consumer and so a payout is guaranteed, or ‘term life’, where a payment is made
if death occurs within an agreed policy term. Many PPI policies include a term life
element (see paragraph 2.3).
2.20 Term life insurance gives financial protection if the consumer dies within a specified
period. The benefit is usually fixed and remains the same during the term of the
policy.
2.21 Decreasing term life insurance is the same as term life insurance except that the
benefit decreases over the term of the policy, and is usually designed to tie in with
the outstanding amount on the consumer’s capital and interest borrowing (eg
mortgage).
2.22 The degree of substitutability of alternatives to PPI will be considered as part of our
analysis of the market definition (see paragraphs 3.75 to 3.89). Of relevance to this is
the extent to which money received as the result of a successful insurance claim
reduces the amount of payment of any state benefits. This is discussed in detail in
Appendix 2.5.
24
Sales of PPI and shares of supply
2.23 One of the measures used by both underwriters and distributors to assess scale of
business is GWP—which is the amount of money paid by customers. 20 The GWP
generated for each type of PPI in 2006 and 2007 is set out in Table 2.1, and the
number of active PPI policies at the end of 2006 and 2007 provided by the 12 largest
distributors is set out in Table 2.2.
TABLE 2.1 PPI: GWP and split between types of PPI in 2006 and 2007
GWP
£m
Personal loans†
First-charge mortgages
Second-charge mortgages
Credit cards
Motor loans
Retail credit†
Overdrafts
Other
Total
2,102
628
471
970
77
61
50
48
4,408
2006
Share of total
PPI GWP
%
47.7
14.2 10.7 22.0 1.8 1.4
1.1
1.1
100.0 GWP
£m
2007*
Share of total
PPI GWP
%
1,896
607
251
801
49
73
51
30
3,757
50.5
16.2
6.7
21.3
1.3
1.9
1.4
0.8
100.0
Source: CC, based on data provided by parties.
*Includes premiums for a number of smaller underwriters (making up around 5 per cent of total GWP), which were derived by
extrapolating 2006 figures using the trend in large distributors’ GWPs.
†The 2006 GWP estimates for personal loan and retail credit PPI differ from those reported in our provisional findings. This is
because the previous estimate included some premiums which we have now classed as PLPPI.
TABLE 2.2 Number of active policies provided by the 12 largest distributors, and the proportion paid for using single
and regular premium
PPI policies distributed by the 12 largest
distributors of PPI, end of 2006
Total number of
active policies
singlepremium
%
regularpremium
%
Total number of
active policies
singlepremium
%
regularpremium
%
4,241,235
93.1
6.9
4,127,230
96.4
3.6
Credit type
Personal loans
First-charge
mortgage
Credit card
Second-charge
mortgage
Motor finance
Overdraft
Other
PPI policies distributed by the 12 largest
distributors of PPI, end of 2007
1,174,842
0.0
100.0
1,116,511
0.0
100.0
7,676,777
0.0
100.0
6,773,627
0.0
100.0
225,664
64.6
35.4
197,188
72.3
27.7
195,093
721,862
29,671
80.2
0.0
56.2
19.8
100.0
43.8
132,284
613,720
N/P
88.9
0.0
N/P
11.1
100.0
N/P
Source: CC, based on data submitted by the 12 largest PPI distributors.
N/P = not provided.
2.24 We note that our estimate of the size of the PPI market in 2006 was significantly
lower than that of Mintel, 21 but we are content that our estimate is robust. Table 2.3
shows the GWP of the 12 largest distributors over the period 2002 to 2007, and for
the first half of 2008, based on data we have received from those distributors. This
shows a peak of GWP in 2004, with a slight drop in 2005 and more significant drops
in 2006 and 2007. Mintel has been reviewing the size of the market since 2001, and
20
Net of insurance premium tax (IPT). According to Mintel UK, Creditor Insurance, November 2005, the sector had an estimated GWP of £5.5 billion in 2005. Mintel estimated the 2006 GWP as £5.35 billion—Mintel Creditor Insurance Report (January 2007). 21
25
we noted that its data also suggested that PPI GWP has started to fall, though it
estimated that peak sales of PPI occurred in 2005 (see Table 2.4).
TABLE 2.3 Level and growth rates of PPI policies in the UK of the 12 largest distributors
£ million
GWP*
% growth
2002
2003
2004
2005
2006†
2007
First half 2008
2,977
-
3,792
27.4
3,912
3.2
3,850
–1.6
3,485
–9.5
3,102
–11.0
1,308
-
Source: CC calculations based on data provided by parties.
*Twelve largest participants in PPI market. As noted in Table 2.1, we estimate the total PPI GWP written in 2007 to be
£3.76 billion.
†Our estimate of total GWP in 2006 has increased slightly from that reported in our provisional findings, following further
clarification of data provided by the parties.
TABLE 2.4 Level and growth rates of PPI policies in the UK
£ million
GWP
% growth
2001
2002
2003
2004
2005
2006*
3,199
-
4,086
27.7
4,563
11.7
4,960
8.7
5,500
10.9
5,352
–2.7
Source: Mintel Creditor Insurance Report (January 2007).
*Estimated value.
TABLE 2.5 PPI penetration rates by volume, 2002 to first half 2008*
per cent
Type of PPI
Personal loans
First-charge mortgages
Credit cards
Second-charge mortgages
Motor finance
Overdrafts
2002
2003
2004
2005
2006
2007
First half
2008
62
26
33
83
21
-
58
20
39
83
19
7
51
17
31
80
19
1
48
14
26
74
16
3
44
12
22
70
14
3
40
10
21
56
14
3
38
9
21
46
13
3
Source: CC analysis based on data from the parties.
*Parties who stopped selling PPI during the period were removed from the sample. The implication of this is to improve the
consistency of the estimates across time, but reduce the precision of the penetration rate estimates for earlier years. This
adjustment lies behind the difference in the penetration rates reported here, and those found in our provisional findings.
2.25 Table 2.5 shows the variation of PPI penetration rates (the percentage of credit sales
made with PPI) over time. We noted that penetration rates have declined over the
previous five years for all types of PPI. While the decline in penetration rates for
PLPPI and MPPI has been relatively steady over the period, penetration rates for
SMPPI have dropped sharply since 2006, whilst the decline in CCPPI and motor
finance PPI penetration rates appears to have levelled off somewhat since this date.
Reasons put to us for the decline in penetration rates included the shift by some PPI
26
providers to selling PPI on a non-advised basis when the FSA took over regulation of
PPI in 2005; 22 regulatory scrutiny; and negative publicity in the media. 23,24
2.26 The average penetration rate for sales of PPI made at the point of sale is highest for
second-charge mortgages, with around half of second-charge mortgages being sold
with PPI. Approximately 38 per cent of personal loans are sold with PLPPI. Penetration rates of CCPPI are lower (around 20 per cent). For MPPI, our data indicates that
its penetration rate is around 10 per cent. However, this does not reflect the sales of
MPPI made by intermediaries, for which we do not have data. Our data does indicate
that approximately half of all mortgages sold in 2006 were sold through intermediaries. The Council of Mortgage Lenders (CML) estimated that intermediaries sold
32 per cent of the MPPI policies sold in the second half of 2006. Accordingly, we
assume that the penetration rate for MPPI is higher than the 13 per cent which our
data indicates, but we are of the view that it is still significantly lower than the
penetration rates for SMPPI and PLPPI.
Customers
2.27 We saw evidence that PPI take-up varies with age, gender and socio-economic
group. That evidence also suggests that PPI consumers are more likely to earn less
than the national average income or come from socio-economic groups C and D.
Different types of PPI policies appear to be more popular with different age groups:
CCPPI has its highest sales rate among 35- to 44-year-olds; PLPPI among 35- to 44year-olds and 18- to 24-year-olds; and MPPI among 18- to 24-year-olds. Various
surveys we have seen show that the most-frequently cited reason for taking out PPI
relates to ‘peace of mind’. Appendix 2.6 provides further information on the characteristics of PPI consumers.
Organization of the PPI value chain
2.28 In this section we set out the basic structure of the PPI industry and provide an
overview of the different players in the sector.
2.29 There are four ways in which businesses sell PPI to consumers (and a business may
utilize one or more of these ways):
• The PPI is sold to the consumer by the business that is providing the credit
product, and the PPI underwriter is part of the same group of companies as the
lender. In this case the PPI seller is a distributor and is vertically integrated with
the underwriter of the insurance.
• The PPI is sold to the consumer by the business that is providing the credit
product, and the lender contracts with a dedicated underwriter to provide PPI for
the lender to distribute. In this case the PPI seller is a distributor but is vertically
separated from the insurance underwriter.
22
[] post-issues statement []; RBSG, post-issues statement summary of hearing, paragraph 9, said that it thought it
possible that a decline in penetration rates might have been slightly steeper due to a move to non-advised sales.
23
0.29 per cent of [] CCPPI customers who cancelled their policies by telephone between September 2007 and February
2008 cited ‘press’ as their reason for doing so. Alongside this reason, [] noted a number of other reasons cited by its
customers for the cancellation of PPI during the same period, including among others, the cost associated with PPI (87 per
cent) and alternative cover being relied on by customers (5 per cent).
24
HSBC post-Emerging Thinking hearing summary.
27
• The PPI is sold to the consumer by an intermediary, 25 who sources credit for a
consumer, and sources either the PPI associated with the credit product or PPI
from another underwriter (either from a panel of underwriters or from a preferred
underwriter).
• The PPI is sold to the consumer by a firm which either underwrites PPI itself or
sources it from an insurer. However, the firm offers the PPI to insure credit which
the consumer has taken out elsewhere. In this case we shall describe the firm as
selling PPI on a stand-alone basis.
2.30 We consider in paragraphs 2.31 to 2.58 the various types of businesses involved in
the supply of PPI.
Underwriters
2.31 The underwriters of PPI policies enter into contractual relationships with individual
consumers, and take on the risk that a consumer may be unable to meet financial
obligations arising from a credit agreement due to events stipulated in the policy
(typically ASU and often L, as explained in paragraph 2.3). To insure these risks,
underwriters charge the customer a premium. The underwriters are able to control
their levels of risk by stipulating benefits and exclusions in the insurance policies they
offer.
2.32 The largest underwriters of PPI in 2006, 2007 and the first half of 2008, by GWP, 26
are set out in Table 2.6. Not all of these underwriters underwrite all elements of PPI;
for example, Prudential underwrites only long-term insurance elements (life, critical
illness and long-term disability cover) of PPI policies. 27 Appendix 2.7 provides
summary information about these underwriters.
TABLE 2.6 The proportion of GWP underwritten in 2006, 2007 and the first half of 2008, by underwriter
per cent
Share of PPI underwriting
Underwriter
Lloyds TSB
Aviva
HBOS
RBSG
Barclays
AXA
Cardif Pinnacle
Genworth
Prudential
London General Insurance
HSBC
Others*
Total
Total GWP, £m
2006
2007
[0–5]
[10–30]
[10–30]
[5–10]
[10–30]
[0–5]
[5–10]
[0–5]
[10–30]
[0–5]
[0–5]
5.3
100.0
4,397.2
[10–30]
[10–30]
[10–30]
[5–10]
[10–30]
[5–10]
[0–5]
[0–5]
[0–5]
[0–5]
[0–5]
5.3
100.0
3,737.4
First half 2008 [10–30]
[10–30]
[10–30]
[10–30]
[10–30]
[5–10]
[0–5]
[0–5]
[0–5]
[0–5]
[0–5]
5.3
100.0
1,518.8
Source: CC analysis of data provided by underwriters.
*For 2007 and first half 2008, we assume that the proportion of total GWP accounted for by other underwriters is equal to the
2006 level.
25
For the purposes of this investigation we have chosen to distinguish between intermediaries, who source the credit being insured from someone else, and distributors, who provide the credit being insured themselves. As explained in paragraph 2.24, the total GWP is significantly different from that calculated by Mintel. 27
Appendix 2.6 provides further details. 26
28
2.33 PPI policies are typically not underwritten on an individual basis (with the policy costs
specifically tailored to the customer); rather they are underwritten such that the total
risk associated with all the policies that are issued is pooled. We heard of few
examples of companies moving towards pricing their PPI so that it reflects some
customer characteristics, such as age. 28
2.34 We consider the state of competition in the underwriting market in Section 7.
Distributors
2.35 For the purposes of this investigation distributors are lenders (for example, banks,
mortgage providers, credit card providers, motor finance companies) who offer PPI
combined with their underlying credit products, either at the point of sale or
subsequently (mostly in the case of CCPPI 29 and retail PPI 30 ). CCPPI is
predominantly sold either when the credit card is sold or when it is activated. We
consider both of these times to be a point of sale for the purposes of this
investigation. Lenders can also offer PPI on a stand-alone basis.
2.36 When a borrower agrees PPI with a bank or building society, the lender acts as a
‘distributor’ for an underwriter’s PPI policy. Appendix 2.8 provides a list of the
distributors of PPI who have provided evidence during the investigation. Not all
distributors offer all forms of credit, and hence they do not offer all forms of PPI.
2.37 The split in the share of supply of PPI between the largest PPI distributors in 2006,
2007 and the first half of 2008 is shown in Table 2.7. All brands sold by a distributor
are included (for example, RBSG includes the Royal Bank of Scotland and NatWest
(and its other) brands; Barclays includes Woolwich and its other brands etc). It can
be seen that in 2006 the 12 largest distributors of PPI were responsible for about
85 per cent of the policies sold (by GWP). Appendix 2.7 provides summary
information about these distributors. Major changes which occurred during this
investigation included: Northern Rock stopped selling personal loans and PLPPI, and
is reducing the size of its mortgage business significantly; MBNA ceased offering
personal loans (and associated PLPPI) in 2007; Nationwide temporarily suspended
sales of CCPPI and PLPPI in August 2007, with sales of PLPPI recommencing in
September 2008; and Barclays ceased offering its secured loans (and associated
SMPPI) in August 2008. In addition, some parties stopped selling single-premium
PPI (see paragraph 10.247).
28
In October 2006 HBOS introduced risk-based pricing on its PPI policies, which meant that the premiums were raised for
customers with a higher risk of making a claim, and lowered for those with a lower risk. Four criteria are used to determine the
risk profile of customers: age, gender, the size of the loan and the loan term. PPI price was linked to the risk that a customer
would make a claim on PPI (not to the risk of default on the loan); Lloyds TSB introduced age-related pricing for PLPPI in
March 2004; British Insurance prices its stand-alone PPI policies to reflect customer age. We were also told that some other
stand-alone PPI distributors used age-banded pricing.
29
We have been told that CCPPI is sometimes offered when a customer phones the credit provider to activate the credit card.
Mintel estimated that 81 per cent of CCPPI was purchased at the point of sale; of those who purchased CCPPI at a later date,
95 per cent bought it from the provider of the card—Mintel report Creditor Insurance, Finance Intelligence, January 2007.
30
SDGFS (formerly Everyday Financial Services (EFS)) told us that nearly [] per cent of its policies were sold to existing
rather than new customers.
29
TABLE 2.7 Share of supply of PPI in 2006, 2007 and first half 2008
per cent
Share of PPI distribution by GWP
Distributor
Lloyds TSB
HBOS
RBSG
Barclays
Cattles
MBNA
HSBC
Alliance & Leicester
Abbey
Capital One
Nationwide
Northern Rock
Other*
Total
2006
2007
First half
2008
[10–30]
[10–30]
[10–30]
[10–30]
[0–5]
[0–5]
[5–10]
[0–5]
[0–5]
[0–5]
[0–5]
[0–5]
16.2
100.0
[10–30]
[10–30]
[5–10]
[10–30]
[5–10]
[0–5]
[5–10]
[0–5]
[0–5]
[0–5]
[0–5]
[0–5]
16.2
100.0
[10–30]
[10–30]
[10–30]
[10–30]
[5–10]
[0–5]
[0–5]
[0–5]
[0–5]
[0–5]
[0–5]
[0–5]
16.2
100.0
Source: CC, based on data provided by parties.
*For 2007 and the first half of 2008, we assume that the proportion of total GWP accounted for by other distributors was equal
to the 2006 level.
2.38 Distributors take a commission from the insurance premium paid by the consumer,
allowing them to cover their costs and make a profit. The OFT suggested that
commission rates paid by insurers to downstream intermediaries look high compared
with other general insurance policies. We discuss commission rates further in
paragraph 2.62.
Vertically-integrated businesses
2.39 Vertically-integrated businesses are those organizations where the credit function
and some or all of the insurance underwriting are provided within the same group of
companies. There are four vertically-integrated groups in the PPI market: Lloyds
TSB, HBOS, Barclays and RBSG. A fifth group, HSBC, was partially vertically
integrated until it sold its Hamilton businesses to Aviva during the course of the
investigation.
2.40 Data we received from the parties showed that in the first half of 2008 over 50 per
cent of PPI was underwritten by businesses which are still vertically integrated (see
Table 2.6). Most of the vertically-integrated firms use third-party underwriters to a
greater or lesser extent; RBSG appears to be the only business which uses its inhouse insurer to underwrite 100 per cent of its PPI policies.
2.41 We consider the effect of vertical integration on the PPI market in Section 8.
Intermediaries
2.42 For the purposes of this report, an intermediary is a third party through whom
consumers identify a suitable type of PPI policy, whether with or without an
associated credit product. In this context intermediaries are used to purchase MPPI,
to a lesser extent, SMPPI, and some motor PPI. Intermediaries can distribute both
credit products and PPI (and other insurance policies) or PPI alone. However, they
neither underwrite the PPI policies nor finance the credit. Intermediaries make
available to consumers the credit products and PPI of one or more credit providers
and/or underwriters. These can be distributed under the brand name of the credit
30
provider or of the underwriter. The evidence we have received indicates that
intermediaries sell significant numbers of only one form of PPI, namely MPPI.
2.43 Our definition of intermediary would catch, among others, price comparison websites.
However, for the purposes of this report we are excluding independent price
comparison websites from our definition of intermediaries.
2.44 Price comparison websites gather and present price and non-price information to
enable consumers to search for and compare products in order to determine which is
suitable for their needs. Price comparison sites act as conduits providing a means by
which consumers may purchase products but do not sell their own products to
consumers. We see these sites, although not without shortcomings, as helping to
increase competition, by providing a transparent and regulated forum through which
price and non-price information is made available to consumers. For this reason we
have excluded price comparison sites as intermediaries even though they represent
a developing sales channel.
2.45 We have not excluded from our definition of intermediary price comparison sites
which are run by parties which also act as distributors, underwriters or stand-alone
providers and through which consumers can purchase that party’s product. The party
has conflicting incentives in terms of offering information on different products,
including its own, in an impartial manner. We do not see as the primary purpose of
these sites to be the provision of information to enable consumers to search and
compare and determine which product is suitable for their needs. We see the
provider-run comparison sites as a sales channel and these are primarily aimed at
selling the provider’s product by using a price comparison site style of marketing. We
do not see provider-run price comparison sites as contributing to competition and
have not exempted them from the remedies.
2.46 While some intermediaries may be tied to a particular lender and/or underwriter,
many more operate independently. This allows an intermediary the option of arranging a credit product from one lender and sourcing PPI from an underwriter not
associated with the lender.
2.47 Where an intermediary is directly regulated by the FSA it can source PPI from a
supplier and then sell it directly to a consumer.
2.48 The Association of Mortgage Intermediaries (AMI) told us that its members ranged
from sole practitioners through to large national organizations which employed
thousands of people. However a ‘typical’ mortgage intermediary firm would employ
two or three advisers and one or two support staff. If the firm were directly authorized
by the FSA, it might employ a compliance officer who could work on a part-time
basis.
2.49 The AMI also told us that small intermediaries commonly joined ‘networks’
(effectively umbrella organizations) which the FSA viewed as a single firm for
regulatory purposes. Examples of networks are Sesame, Legal & General Mortgages
and Open Works. The network is an entity which is regulated by the FSA and takes
regulatory responsibility for all of the firms in it. It also acts as a buying group to
achieve more competitive rates of commission and levels of service for those firms in
it. Other firms join loose affiliations which are known as mortgage clubs. The purpose
of these clubs is to negotiate on behalf of their members for improved levels of
service or product exclusives 31 from mortgage lenders. The mortgage clubs are not
31
AMI provided us with examples of exclusive products its members had access to.
31
exclusive and firms can belong to as many mortgage clubs as they wish. There are
also what the AMI termed ‘pseudo-networks’, which operate in the same way as
networks except that each firm is responsible for its own regulatory conduct. The
firms in a pseudo-network can take guidance and support from it, and it also offers
marketing assistance and information and acts as a group to increase its members’
buying power.
2.50 The networks would usually employ a team of specialists to survey the market for
both mortgages and MPPI (as well as other types of insurance policies) and prepare
a panel of providers for the network members to select from when advising
customers. The panel would be expected to be made up of providers which were
representative of the market as a whole. Neither the network nor any of its members
would have a contractual relationship with any of the providers on the panel. The AMI
told us that the majority of its members were independent mortgage advisers who did
not have contractual ties with insurance companies for types of PPI policies,
including MPPI.
2.51 A network can take on an Appointed Representative (AR) who then sells PPI on
behalf of the network. An AR is allowed to carry on certain regulated activities by the
network (his Principal) under a contract by which the Principal accepts responsibility
for the regulated activities carried on by its AR, which would include selling PPI.
Hence it would be expected that the AR only sources PPI through the network.
2.52 The intermediary network can also provide PPI to an intermediary that is directly
regulated by the FSA. In this case, it is possible that an intermediary could use
several different networks, given that it is independently authorized by the FSA.
2.53 As noted in paragraph 2.42, intermediaries appear to have made significant sales of
MPPI. Intermediaries often recommend their own choice of MPPI policy to sell with a
mortgage, and this appears to have resulted in credit providers seeing lower
penetration rates for their own MPPI policies. For example, Northern Rock told us
that 89 per cent (by value) of its mortgage applications in 2006 came from mortgage
intermediaries. Only [less than 10] per cent of these customers took Northern Rock’s
MPPI product with their Northern Rock mortgage (as opposed to taking out MPPI
from another provider or not taking out MPPI at all). At the other extreme, [over half
of the] customers who took out a Northern Rock mortgage from one of its branch also
took Northern Rock MPPI.
2.54 As at 18 April 2008 approximately 6,619 firms were authorized by the FSA to arrange
and advise on regulated mortgage contracts. The CML estimated that the UK’s
mortgage intermediary market provided 32 per cent of the 263,800 new MPPI
policies sold in the second half of 2006. 32 Only three intermediaries (Sesame
Limited, Legal & General Partnership Services Limited and Openwork Limited) held
more than 3 per cent of the intermediaries’ share of supply (by mortgage revenue)
each in 2006.
Stand-alone provision
2.55 For the purposes of this investigation, we considered that there are two ways in
which to provide stand-alone PPI. In both, PPI is offered to insure credit provided by
another business. In the first case the provider does not offer credit in any circumstances; it is either an insurer (such as Cardif Pinnacle) or a firm which contracts with
an insurer (such as British Insurance). In the second case the provider is a business
32
Source: www.cml.org.uk/cml/statistics, spreadsheet PPI2.
32
that does offer credit and PPI together, but also offers PPI on a stand-alone basis
with the aim of gaining PPI business from other credit providers. In this second case
the PPI policy could be one made available combined with its own credit products
(for example, MPPI policies from HBOS and Abbey) or to be sold on a stand-alone
basis (for example, Alliance & Leicester, Abbey).
2.56 Stand-alone provision of PPI appears predominantly to take place in relation to
PLPPI and MPPI; we have come across only two providers offering a stand-alone
CCPPI policy. Paragraph 3.40 and Table 3.3 provide details of the levels of standalone PPI sales achieved.
2.57 As providers of stand-alone PPI do not have access to the customers they seek at
the point of sale of the credit, they have to rely on other means of attracting
customers. Most appear to have an Internet presence, though some also make
telephone sales, and use price comparison websites (where these allow comparison
of stand-alone policies) or sponsored links on search engines to attract customers.
Others may seek to sell it to consumers with whom they already have an existing
relationship (for example, a bank may try to sell it to its current account customers,
knowing that some of them are likely to have credit with another provider).
2.58 We were told by several parties 33 that selling PPI away from the point of sale of
credit is very difficult, though it has also been put to us that it is becoming easier due
to increased use of the Internet by consumers. Several insurers and distributors have
entered and then exited the stand-alone PPI market, including Aviva, AXA and Marks
& Spencer. However, we are aware of at least one insurer which underwrites PPI for
sale combined with credit products and which also offers a stand-alone MPPI policy
(Cardif Pinnacle) and several distributors (two of whom launched stand-alone policies
in the second half of 2007). It was put to us that stand-alone PPI providers run a
higher risk of adverse selection than point of sale PPI providers, 34 resulting in standalone providers facing a higher claims ratio on their insurance policies than point-ofsale insurers do. 35 It was also put to us that the higher risk of adverse selection is not
an insurmountable barrier to the successful provision of stand-alone PPI.
How PPI is sourced and sold
Contractual relationships between underwriters and distributors
2.59 The evidence we have received from parties in the sector suggests that distributors
have a significant level of input into the design of PPI policies, 36 and set PPI prices
relative to their main competitors in the consumer credit markets.
2.60 The contracts for the provision of PPI services are generally put out to tender by
independent distributors and by distributors which operate a hybrid business model
(where part of the business is outsourced to underwriters). Vertically-integrated
businesses also put some contracts out for tender, and we were provided with
evidence that vertically integrated distributors conducted tenders for the provision of
PPI services to the distribution arm of their businesses. However, we were also told
33
See, for example, Alliance & Leicester, post-issues statement summary of hearing, paragraph 5; Genworth, post-issues
statement summary of hearing, paragraph 6; Citizens Advice, post-issues statement summary of hearing, paragraph 10.
34
Adverse selection, sometimes referred to as anti-selection, occurs as a result of information asymmetries between the insurer
and the customer. Customers who know they face a large risk of needing to make a claim on an insurance policy are more
likely to take out insurance than those who face small risks; however, the insurer does not know which customers fall into which
category and so cannot price its insurance premium accordingly.
35
Aviva told us that this was its experience in the stand-alone market; its claims ratio for its stand-alone products was significantly higher ([]) than for its products sold combined with credit.
36
For example, MBNA, post-Emerging Thinking hearing summary, paragraph 35.
33
that some vertically-integrated businesses tender more with a view to testing the
wider market, and ensuring that the in-house offer is competitive. 37
2.61 Contracts between underwriters and distributors tend to be negotiated on a contractby-contract basis (we were told that this reflected the individual nature of different
credit products that PPI policies were designed to insure). Contracts tend to be of
one to five years in length, and many have an initial period after which the contract
continues in force until one party gives notice.
2.62 Distributors are contractually entitled to a percentage of GWP as commission.
Typical commission rates are 50 to 80 per cent for PLPPI and CCPPI and 40 to
65 per cent for MPPI. The remaining GWP is passed to the underwriter to cover
expenses, including claims. In the event that claims levels are less than expected,
the resulting profit typically will be split between the underwriter and distributor
according to an agreed profit share percentage; typically 90 to 100 per cent in favour
of the distributor. A separate profit share arrangement will typically apply to any
investment income earned by the underwriter on premium income and may also
apply to tax benefits on life business.
2.63 The process by which insurers tender to underwrite PPI policies is set out in detail in
Appendix 2.9. Information on tenders and contract lengths can be found in Section 7.
The sale of PPI to customers
How PPI is sold
2.64 PPI is typically offered for sale by distributors combined with their credit products.
This cross-sell is most often done at the time the customer is making an application
for credit, or accepting an offer of credit—either of these points can be considered to
be the point of sale of credit (for credit cards it can also be sold when the card is
activated). However, the distributor has the opportunity to remind the customer of its
PPI offering at any time it has contact with the customer (for example, on every
statement it sends out, or when a customer contacts the distributor with a query).
2.65 This opportunity is only available to the distributor selling the credit, and more often
than not it results in the credit and PPI being sourced from the same distributor.
Distributors can include banks, building societies, car dealerships or retailers; a
distributor can provide credit using one or more brands (for example, Royal Bank of
Scotland and NatWest; Halifax and Bank of Scotland, etc; the 12 largest distributors
between them operate 44 brands). While most PPI policies are taken out at the point
of sale, PPI may be purchased as a ‘stand-alone’ policy, independent of the credit
product. Stand-alone PPI is discussed further in paragraph 2.55.
2.66
The OFT 38 found that PPI is a secondary product, and parties to the investigation
commented that the demand for PPI was driven (wholly or generally) by the purchase
of the underlying credit product. 39 We have not come across any instances where
purchase of PPI is actually a condition of credit being given. However, there is
37
Summary of a post-Emerging Thinking hearing with Company B, paragraph 15. Other companies said that they compared or
benchmarked their in-house underwriters against mainstream competitors or independent underwriters: see HBOS post-issues
statement hearing summary, paragraph 8, and Barclays post-issues statement hearing, paragraph 20.
38
OFT, Payment Protection Insurance; The OFT’s reasons for making a market investigation reference to the Competition
Commission; February 2007, p23.
39
Aviva, response to the issues statement, paragraph 2.1.1; Barclays, response to the issues statement, paragraph 3.1;
Genworth, summary of post-issues statement hearing, paragraph 5; Alliance & Leicester, summary of post-issues statement
hearing, paragraph 7; MBNA, summary of post-issues statement hearing, paragraph 6; ABI, summary of post-issues statement
hearing, paragraph 6.
34
evidence to suggest that some consumers were under the impression that their credit
application was more likely to be accepted if they bought PPI. For example, the CC
BMRB 2007 survey found that 25 to 47 per cent 40 of respondents thought that their
credit application was more likely to be accepted if they took out PPI (see paragraph
5.49). The FSA, in its September 2007 report, found that 92 per cent of the
businesses it visited in its mystery shopping exercise did make it clear that the
purchase of PPI was optional. 41 Appendix 2.5, on regulation, sets out that, if
purchasing PPI were a condition of the credit, the advertised APR would have to
reflect the cost of taking both credit and PPI. This may influence a distributor’s
decision on whether to make purchasing PPI a condition of sale.
The link between PPI and credit
2.67 The OFT suggested that competition was centred on the credit product, with the APR
used by a consumer as a means of comparison, and noted that the APR does not
indicate the price of a credit product including the PPI policy. We have heard different
views on the extent to which consumers take account of the price of PPI when
choosing their credit and PPI, or choose the most competitive credit offer and then
decide whether they can afford to take out PPI. 42 We consider the link between credit
and PPI in paragraphs 3.95 to 3.118.
Sales methods
2.68 The FSA authorizes PPI to be sold either on an ‘advised’ or a ‘non-advised’ basis, at
the discretion of the PPI provider. Any sales on an ‘advised’ basis, where an adviser
can make a personalized recommendation as to the suitability of a type of PPI policy
for the consumer’s needs, are subject to additional selling conditions compared with
those offering PPI on a ‘non-advised’ basis (for example, in an advised sale the seller
is required to provide a statement of demand and needs to the consumer). 43 Further
details on the difference between advised and non-advised sales can be found in
Appendix 2.5.
2.69 Table 2.8 shows how the largest distributors currently sell their PPI through the
branch/face to face and telephone channels.
40
This range covers MPPI, SMPPI, PLPPI and CCPPI holders.
FSA, The Sale of Payment Protection Insurance Thematic update, September 2007, p11; the FSA looked at PPI sold
combined with unsecured personal loans, revolving credit (such as credit cards and instalment finance), prime mortgages, other
insurance policies and secured loans.
42
For example, Nationwide said that customers actively considered the purchase of PPI when considering taking out a credit
product, and said that this was borne out by the consumer survey produced for the OFT (Nationwide, response to the issues
statement, p3). Northern Rock told us that customers tended to choose credit based on the APR, and then look at the prices of
loans with and without PPI (Northern Rock, summary of hearing, paragraph 15). Barclays said that customers generally made
their decision based on the affordability of the total monthly repayments of the loan and PPI (Barclays, summary of hearing,
paragraph 11). The CML believed that customers searched on the mortgage rather than MPPI, and suspected that the final
decision was based on the price of the mortgage or the mortgage plus PPI, rather than the price of the PPI. Lloyds TSB said
that customers tended to base their decisions on the price of credit, and then considered the price of the package with PPI
(Lloyds TSB, response to the issues statement, p2). Citizens Advice suspected that customers looked at the price of credit per
month, and if they could afford the PPI then they would buy that as well (Citizens Advice, summary of hearing, paragraph 8).
43
FSA, Advised and non-advised sales: insurance factsheet, October 2006.
41
35
TABLE 2.8 Sales processes used by distribution channel for the largest distributors of PPI
Distributor
Abbey
Branch/face to face Telephone
Alliance & Leicester
Barclays
Capital One
Non-advised (MPPI); non-advised
(PLPPI)
Non-advised
Advised
N/A
Non-advised (MPPI); non-advised (PLPPI)
Non-advised
Advised
Non-advised
Cattles
HBOS
HSBC* (HFC)
(LifeChoices)
Lloyds TSB
MBNA
Nationwide
Northern Rock
RBSG
Advised
Advised
Advised
Advised
Advised
N/A
Non-advised
Advised
Non-advised
Advised
Advised
Not applicable
Non-advised
Advised
Non-advised
Non-advised
Advised and non-advised
Non-advised
Source: CC, based on evidence provided by the parties.
*HSBC told us that it was currently not selling PPI except through its HFC brand. HSBC Bank currently offers to customers to
whom it previously offered MPPI, PLPPI and CCPPI, the chance to discuss their broader protection needs with one option
being to take LifeChoices (if it is the most suitable product), a short-term income protection product. Customers who were
previously offered MPPI are offered LifeChoices subject to suitability.
2.70 We were told by two companies that they had seen a drop in PPI sales when they
moved from selling on an advised basis to selling on a non-advised basis. 44
2.71 When advertising credit, lenders have to use APR as the method of showing the
price of the loan. Whilst the APR quoted on credit does not include the price of PPI,
the monthly repayment quoted sometimes has the price of PPI included in it
automatically, though we have been told that quotes with and without PPI are
provided. However, showing the monthly price of PPI does not give an indication of
the importance of the lifetime cost of the PPI policy. Which? provided us with data it
had gathered on the average total amount consumers would pay for a £7,500
unsecured personal loan, repayable over five years, with PPI and without PPI. It
looked at 54 types of PPI policies provided by 31 lenders or underwriters. The
average interest payable on the £7,500 loan over that time would be £1,566; the
price of PPI was less than the interest payable on the loan for only five of the 54 PPI
policies on the list. We calculated from Which?’s data that the average APR quoted
for loans on its list was 7.8 per cent. If the APR calculation had included the price of
PPI, the average APR would have been 18.8 per cent.
2.72 Using data we received from the largest distributors, we plotted the APR of credit
without PPI against the APR if PPI is included in the calculations, for a £5,000
personal loan repayable over three years (see Figure 2.1), and the ratio of the total
APR of PPI and interest against the APR of the interest only for the same PPI
policies in Figure 2.2.
44
[]. []. Also, RBSG, summary of hearing, paragraph 9, said that it thought it possible that a decline in penetration rates
might have been slightly steeper due to a move to non-advised sales.
36
FIGURE 2.1
Comparison of APR for personal loans with and without PPI—December 2006
data for a £5,000 loan repayable over three years
60
APR—including PPI (%)
50
40
30
20
10
0
0
5
10
15
20
25
30
35
40
APR—excluding PPI (%)
Source: CC, based on data provided by the parties.
FIGURE 2.2 Ratio of APR of PPI and interest to APR of interest for different types of PPI
policy
3.5
APR including PPI/APR loan only
3.0
2.5
2.0
1.5
1.0
0.5
0.0
1
4
7
10 13 16 19 22 25 28 31 34 37 40 43 46 49 52 55 58 61 64 67 70 73
PPI credit combination
Source: CC, based on data provided by the parties.
37
2.73 Figure 2.1 shows that for the majority of the personal loan products for which data
was available, taking out PPI leads to a significantly higher APR. In Figure 2.2 we
found that the APR including PPI was between 1.3 and 2.9 times as large as the
APR for the same loan without PPI—if the PPI was less expensive than the interest
on the loan we would expect the ratio to be less than two. It was less than two in
around one-third of the products we looked at. For the remainder of the products the
PPI element was more expensive than the cost of the underlying credit product.
2.74 The OFT suggested that the level of information available on PPI before the point of
sale of the credit product was poor; however, a number of parties emphasized to us
that information was available prior to the point of sale. Several parties 45 told us that
PPI was ‘sold not bought’ at the point of sale—that is, the initiative for the purchase
usually comes from the salesperson, rather than the consumer asking to purchase
PPI combined with credit. 46 Others, however, told us that this was not the case or
was not necessarily the case. 47 In its latest report on PPI, the FSA said that around
one-third of those firms it visited failed to ensure that consumers were given the basic
information necessary to make an informed decision about purchasing a PPI policy,
and that, in its latest mystery shopping exercise, less than half of shoppers were told
about the policy limitations and exclusions. 48
Distribution channels
2.75 PPI is sold predominantly by distributors and intermediaries (as noted in paragraph
2.42, intermediaries predominantly sell MPPI). Credit products and the associated
PPI are sold through several channels:
• face to face (eg through a branch network or at an intermediary’s office);
• telephone sales;
• the Internet; and
• the post.
2.76 Our analysis of data provided by the large distributors (see Table 2.9) showed that
for all sales channels there was a similar pattern to distribution across the four main
types of PPI, with the majority sold face to face, and smaller volumes sold over the
telephone and Internet (though the differences between sales channels were less
marked for CCPPI than for other types of PPI policy). 49 Northern Rock said that the
45
Barclays, summary of hearing, paragraph 24. Barclays did note that there is information available prior to the point of sale; CML, summary of hearing, paragraph 14; Which?, summary of hearing, paragraph 7; and FSA’s third thematic update, p12;
Aviva, response to the issues statement, p1. Aviva said that its comment that PPI is sold, not bought, was not made in the context of a discussion of sales methods. It was used to illustrate the nature of PPI as a secondary product, combined with the
credit product, as well as the fact that consumers do not naturally consider it. 46
We were told that at the point of sale of credit the customer is most receptive to thinking about, or likely to consider, whether they are likely to be able to meet their credit repayments. Post Issues Statement hearing summaries: Alliance & Leicester, Aviva, AXA, Cattles, Nationwide. 47
RBSG did not agree that PPI was sold and not bought (summary of hearing, paragraph 13), HSBC did not necessarily agree
(HSBC response to the issues statement, paragraph 4.10) and Northern Rock said that a large number of customers searched
for information on PPI prior to purchase, implying that it was bought rather than sold (summary of hearing, paragraph 10). 48
FSA, The Sale of Payment Protection Insurance: Thematic Update, September 2007. The FSA found that 40 per cent of firms visited that sold PPI on an advised basis failed to demonstrate that they had properly determined whether the particular PPI
policy was suitable. 49
Mintel (Creditor Insurance, Finance Intelligence, January 2007) estimated the split for PLPPI sales between sales channels
was 42 per cent face to face; 25 per cent over the telephone; 11 per cent over the Internet; 3 per cent by post; and 19 per cent through other means (including car dealerships, intermediaries and door-to-door lenders). We are content that our calculations,
which are based on data provided to us by distributors, are robust. 38
lower take-up of PPI on credit purchased on the Internet may reflect the lack of
advice given by a salesperson on whether to purchase PPI during these sales. 50
TABLE 2.9 Sales of new PPI policies by distribution channel, 2007, by number of policies
per cent
Personal
loans
Number of policies sold on a stand-alone basis
Number of policies sold via your company for use
with one of your credit products, of which
—In branch/face to face
—Telephone
—Internet
—Postal
—Other
Total number sold via a downstream intermediary
All new PPI policy sales
First-charge
mortgages
Credit
cards
Second-charge
mortgages
0.0
4.2
0.0
0.0
99.9
62.3
26.9
10.8
0.0
0.1
0.1
100.0
92.2
67.8
16.5
0.0
0.0
15.7
3.6
100.0
98.5
38.6
28.5
21.3
9.3
2.2
1.5
100.0
79.6
65.2
28.1
6.7
0.0
0.0
20.4
100.0
Source: CC calculations based on data submitted by those of the 12 largest distributors for whom a breakdown was available.
Notes:
1. Percentages as a proportion of all sales where a breakdown was available.
2. The category ‘other’ consists mainly of sales where the sales channel was unknown.
The insurance premium paid
2.77 The payment method for PPI policies is either ‘single premium’ or ‘regular
premium’. 51 A single premium policy requires the total price of the policy to be paid in
one lump sum at the beginning of the contract. This is usually done by adding the
cost of the PPI to the credit. This means that a customer accrues additional interest
costs as a result of the extra borrowing. Single premiums are usually used to pay for
PLPPI, motor finance PPI and some SMPPI (see Table 2.2).
2.78 With regular-premium policies a monthly premium is paid to maintain the policy. This
has the advantage of being able to cover a credit product, the price of which changes
from month to month (for example, CCPPI). Payments for PPI on credit cards and
mortgages are generally covered by regular-premium policies.
2.79 It was put to us that the main advantage of single-premium PPI is that the insurance
remains valid even if the customer misses repayments during the life of a loan.
Cattles told us that it considered regular-premium policies inappropriate for a majority
of its customers, arguing that those customers were more likely to have fluctuations
in their levels of disposable income and therefore required repayment ‘holidays’. PPI
cover paid for by a single premium is unaffected by any breaks in the repayment of
the loan. 52 When a monthly payment on a regular premium is missed, the contract
can in principle be declared cancelled by the provider which would mean that the
consumer would lose the benefit of the PPI cover, possibly at the precise time at
which the consumer most needs it. However, the Financial Ombudsman Service
(FOS) told us that it had not seen any evidence of this occurring and that, although
these sorts of conditions were included in some policies, it would have doubts about
whether this would be fair or reasonable if only one or two payments were missed
even if the policy terms permitted immediate cancellation. Further details of the
advantages of single-premium policies, to consumers and to businesses, can be
found in Appendix 2.10.
50
Northern Rock expressed this view in its response to the issues statement, paragraph 3.8. Table 2.2 sets out the proportion of policies sold in 2006 which were single premium and regular premium. 52
Cattles’ response to the issues statement (paragraph 3.9) and Cattles’ position paper, 2 March 2007 (paragraphs 3.7&3.8). 51
39
2.80 Other parties have said that single-premium policies disadvantage consumers. We
were told that single-premium policies can be more expensive than regular-premium
policies over the course of a loan as the total price of the PPI is added to the amount
borrowed by the consumer. This means that interest is charged on the additional
amount borrowed.
2.81 Consumers with single-premium policies are entitled to a rebate if they cancel the
policy at any time (for example, if they pay off the loan early, or decide to cancel their
PPI). We were told that the calculation of a refund takes into account the greater risk
to which insurers are exposed early in the life of a policy, because of the higher
insured balances outstanding, the initial expenses of the policy and the higher risk of
default due to adverse selection. We consider the impact of single-premium policy
rebates in paragraphs 5.63 to 5.75.
The price of PPI
2.82 For policies which cover a fixed loan amount, the price of PPI is typically quoted as a
monthly amount and the total price of the policy over the lifetime of the loan is also
given in many cases. For policies where the amount covered varies, such as CCPPI
and overdraft PPI, the price is quoted as pounds per £100 of the loan amount. 53
2.83 Further information on the price of PPI is given in Appendix 2.1. There are three
observations about the price of PPI worth mentioning here. First, there is great
variation in the comparative price of PPI between different PPI credit products. For
example, Defaqto 54 estimated that the average price per £100 of benefit for MPPI
was £4.76, for CCPPI was £11.70 and for PLPPI was £18.23.
2.84 Our second observation is that the pricing of single-premium PPI is significantly more
complex than the pricing of regular-premium PPI. There are at least three different
ways in which lenders calculate the premium rate of a single-premium policy. The
single premium charged on a loan will vary according a number of factors including
the APR on the loan, the term of the loan and the method of calculation used by a
particular provider. There is therefore no single-premium rate which customers can
use as a ‘common currency’ to compare the price of single-premium PPI offered by
different providers, without getting a specific quote on the price of credit with PPI
from several suppliers.
2.85 The complexity of the pricing of single-premium PPI is illustrated by two examples.
2.86
First, [] told us (in the context of a discussion about price caps) that it priced its
personal loans with PPI broadly on the following factors: []
2.87 Each of these factors, if multiplied together, gives rise to a large number of price
permutations and, therefore, possible price points. While [] does not use all of the
1,120 possible pricing points, this example shows the complexity of its PLPPI
pricing. 55
2.88 Second, by using the Loan Calculator on its website, we calculated the total monthly
cost of PLPPI as a £ per £100 of monthly benefit for NatWest. The results are shown
in Table 2.10 below.
53
It should be noted that credit is advertised based on its APR. Defaqto, Payment Protection Insurance, 2007—The Year of Reckoning, February 2007. Providers of regular-premium PPI may also offer PPI at different prices, but such variation is normally determined either by
differences in the level of cover offered (for example, whether a PPI policy is LASU, ASU, AS or U) or, less frequently, differences in the risk posed by the customer. 54
55
40
TABLE 2.10 Monthly cost of PPI as £ per £100 of monthly benefit
Amount borrowed (£)
2,000
5,000
Term (years)
1
2
3
4
5
9.67
12.92
16.07
20.83
25.90
9.34
12.19
14.80
18.83
23.06
Source: CC analysis of company website.
2.89 These examples show that, even with a policy offering identical cover, the price of
single-premium PPI from a provider can vary substantially even for an individual
customer, depending on the term of the loan and the APR.
2.90 Our third observation is that the price of PPI can vary significantly within a policy
type; the OFT commented that the difference in price might not always explain the
differences in PPI cover. As set out in paragraph 2.71, Which? provided us with data
it had gathered on the price of taking out a £7,500 unsecured personal loan,
repayable over five years. The data showed that the overall price of taking out PPI
from these providers varied from £1,104 to over £3,000 (a per month price varying
between £18 and £55), with an average price of £2,200. Our own analysis of price
variation is discussed in paragraphs 4.6 to 4.11.
2.91 Related to our last observation, we noted that whilst the price of PPI is typically
quoted as a monthly amount, when the customer receives this quote it is shown
alongside the combined monthly repayment of capital repayment and interest. The
customer is therefore not usually shown the monthly cost of PPI against the monthly
cost of interest on its own (or the APR of the PPI against the APR of the credit
interest). One exception to this is CCPPI where the cost for CCPPI is shown alongside the cost of interest on the customer’s monthly statement.
2.92 We were also told that the price of stand-alone PPI was significantly less than the
price of PPI bought at the point of sale. For example, British Insurance quotes standalone PLPPI at prices from £3.50 to £5.00 per £100 (for ASU cover), depending on
age (prices shown are for 18 and 64-year-old people), 56 and the Post Office told us
that showing the cost of a loan per £100 of cover would show that PLPPI obtained
from a distributor at the point of sale typically cost £25 or more per £100 of benefit,
which was five times more expensive than similar stand-alone cover from Post Office
Financial Services. 57 Evidence from Defaqto 58 suggested that the median figure for
PLPPI per £100 cover was £17.82, with a maximum of £28.25.
2.93 Various parties told us that the reason stand-alone PPI policies were often cheaper
than PPI sold at the point of sale of the credit product was primarily because they
were of lower quality; 59 though some said it might also be because distribution costs
were lower and to increase sales volumes when the stand-alone PPI policies were
newly launched. 60
2.94 HBOS said that it priced its stand-alone PPI (Life Stages) lower than its PPI policies
sold combined with credit because the stand-alone PPI policy had small sales
volumes and it was trying to drive demand and market share. Also, it said that
56
www.britishinsurance.com. Prices as of 4 January 2009. Post Office Financial Services, response to the issues statement, p5.
Defaqto, Payment protection insurance 2007—The Year of Reckoning, op cit.
59
Aviva, post-issues statement hearing summary. 60
Nationwide, post-issues statement hearing summary. 57
58
41
distribution costs for the stand-alone product were low because it was sold via direct
mail.
2.95 In 2007 HSBC began to withdraw its PPI policies from sale and to offer its customers
the chance to discuss their broader protection needs with one possible option being
to take its LifeChoices product. The prices of the old PLPPI and MPPI policies
appear higher than similar LifeChoices products. LifeChoices appears more
expensive than the old CCPPI cover. Table 2.11 shows a comparison between PPI
and LifeChoices offered by HSBC.
TABLE 2.11
Comparison between prices that were charged by HSBC for PPI policies and prices for comparable
LifeChoices policies
PPI (cost per month)
LifeChoices (cost per month)
Personal loan
(£5,000 over 3 years)
£24.11 for L, A, S, U and hospitalization
£16.15 for L, A, S and U for five years
(minimum period of cover)
Mortgage
(£120,000 over 25 years)
£41.58 for A, S and U
£39.20 for A, S and U
Credit card
(Debt of £300)
£2.37 for L, A, S and U
Benefit on death is the balance outstanding.
Benefit for A, S or U is 10% of outstanding
balance per month.
£12.72 for L, A, S and U
Benefit on death is £29,850.
Benefit for A or S is £187 per month.
Benefit for U is 10% of outstanding balance
per month.
Source: HSBC. PPI prices as at November 2007; LifeChoices prices as at May 2008.
2.96 The stand-alone IP policy sold by RBSG through its Churchill brand is priced at a
lower rate than its PPI, but offers more limited benefits (for example, it does not pay
out if the customer dies, and accident and sickness payouts are limited to a maximum of 12 times the monthly payout, whereas the RBS policy pays out for up to five
years). For a personal loan for £10,000 taken over five years, NatWest/RBS PLPPI
would cost £55.45 per month; whereas Churchill IP would cost between £6.87 and
£20.69 per month (depending on the age of the customer). 61
2.97 RBSG told us that providers of stand-alone PPI would not necessarily face the same
costs as distributors and intermediaries who offered PPI with credit; for example,
most providers of stand-alone PPI distributed via the Internet, meaning that they had
lower distribution costs.
2.98 Lloyds TSB said that generally stand-alone PPI policies were different from and
inferior to the PLPPI policies offered by Lloyds TSB and we should expect them to be
offered at a substantial discount. For example, it said that many income protection
policies did not cover critical illness or life risks; they only offered cover for 12 months
(not the term of a longer loan), and might be repriced at any time. In contrast, Lloyds
TSB’s PLPPI product received Defaqto’s highest score (five stars), but British
Insurance’s income protection product received only three stars, because it did not
offer critical illness or hospitalization benefits.
2.99 Nationwide said that a critical factor influencing the price of a stand-alone MPPI
policy was the fact that most providers of stand-alone PPI did not have to support the
cost of an extensive branch network and the associated costs of advisers in those
branches.
61
Prices as of second quarter of 2007.
42
2.100 Some stand-alone PPI policies do not appear to have been rated as highly as by
Defaqto as PPI policies sold in combination with credit. 62 Defaqto gives policies a
‘DNA Rating’ based on a number of scoring bands; it then gives a ‘Defaqto Star
Rating’ out of five. As mentioned in paragraph 2.98, British Insurance’s stand-alone
personal loan protection product received three Defaqto stars in 2008. The personal
loan protection product offered by Paymentcare which offers accident, sickness,
unemployment and hospitalization cover also received three stars from Defaqto. On
the other hand, Paymentcare received a five-star rating in 2007 for its MPPI policy
and a four-star rating for its CCPPI policy. Churchill’s short-term IP policy achieved a
five-star rating in 2008. We compare the quality of PPI policies sold at the point of
sale and on a stand-alone basis, using Defaqto scores as a measure of quality, in
Section 10.
Search
2.101 As noted in paragraph 2.64, most PPI is sold at the point of sale of the underlying
credit product.
2.102 The OFT was concerned that consumers did not appear to be shopping around for
PPI. However, we were told that the overall market for credit was competitive as
borrowers looked for the best deals (generally using the APR offered as a
comparison tool). 63 We were shown consumer research which found that for half of
all new loans the choice of provider was driven by a consumer’s existing relationship
with the provider, with interest rates being the other significant driver of choice. One
party suspected that this did not necessarily translate into high search rates for the
associated PPI. 64
2.103 We received conflicting views on whether consumers shop around for PPI. Some told
us that consumers only shopped around for the credit (and, as noted in paragraph
2.74, that PPI is sold, not bought). 65 Others told us that consumers increasingly
shopped around for the combination of credit and PPI, taking advantage of price
comparison websites, among other things, to find the best overall package. 66
2.104 The qualitative research carried out on our behalf by BMRB 67 found that participants
had generally shopped around for their primary credit products, but generally had not
considered payment protection options before applying for their chosen loan. There
were exceptions to this, however, particularly among those who had recently
undergone a change of circumstance and those who reported that they were
particularly anxious about personal finance. We look in more detail at the extent to
which consumers compare PPI policies in paragraphs 5.9 to 5.57.
2.105 We were told that to sell PPI at any point other than the point of sale would
significantly reduce the penetration of PPI and increase the likelihood of adverse
selection, 68 which was identified as a problem facing providers of stand-alone PPI
(see paragraph 2.58).
62
Defaqto said that the ratings were mathematically calculated using a series of scoring bands. The ratings concept is similar to
that used for hotels and restaurants. See Appendix 4.2 for a more detailed explanation of the Defaqto star ratings. 63
See, for example, Barclays’ response to the issues statement, paragraphs 5.3 and 5.4; Abbey response to the issues statement, paragraph 3.9; Abbey, summary post-issues statement of hearing, paragraph 14. 64
Alliance & Leicester, post-issues statement summary of hearing, paragraph 14. 65
For example, CML, post-issues statement summary of hearing, paragraph 20; Genworth, post-issues statement summary of hearing, paragraph 9. 66
For example, Barclays, post-issues statement summary of hearing, paragraph 26; HSBC, response to the issues statement, paragraph 4.15; Finance and Leasing Association, post-issues statement summary of hearing transcript, paragraph 19. 67
BMRB, Payment Protection Insurance Market Investigation Research Project, February 2007. 68
Aviva, response to issues statement, p2. 43
2.106 There seems to be a lack of advertising of PPI prior to point of sale which may
contribute to the fact that many customers do not shop around prior to that time.
Customers may be unaware of the types of protective cover available to them before
reaching the point of sale as the search focus is on the credit product. We have been
told that there are a number of initiatives aimed at informing customers (such as the
leaflet A Guide to Payment Protection Insurance, 69 which was produced earlier this
year). However, it is not clear how effective these initiatives are or will be, or how
they will affect customer behaviour prior to the point of sale.
69
This guide is an initiative between the ABI, BBA, CML, FLA and Protect. A copy can be found at:
www.abi.org.uk/BookShop/ResearchReports/Payment_Protection_Guide.pdf.
44
3. Market definition for the distribution of PPI
Summary
3.1 In this section we define the relevant market for the distribution of PPI. We do this
primarily by assessing the available evidence regarding the substitutability of PPI
policies offered by different PPI providers and by assessing the substitutability of PPI
and alternative types of insurance. We also consider the relationship between PPI
and credit and the implications of this relationship for market definition.
3.2 The evidence in this section should be read in conjunction with our findings in
Section 4, where we seek to verify our findings on market definition by looking at
indicators of the level of competition between providers of PPI. Our determination of
the relevant market takes account of all the evidence available to us and is consistent
with the evidence in the round.
3.3 We find that PPI distributors and intermediaries are not competitively constrained by
other PPI providers or by alternative types of insurance. We find this because:
• The internal documents and oral evidence that we received from the distributors
indicated that the responsiveness of demand for distributors’ PPI policies to
changes in the price of those policies was low.
• Our assessment of the evidence on consumers’ search and switching patterns
indicated that relatively few consumers shop around for PPI policies or
combinations of PPI and credit.
• The results of the CC GfK NOP 2008 survey of purchasers of PPI policies
indicated that limited numbers of purchasers of PPI policies compared two or
more PPI policies before their purchase. Our analysis of the results of this survey
indicated that changes in the price of a PPI policy would result in a relatively small
change in the sales of that policy.
• Our analysis of the distributors’ sales data showed that demand for a distributor’s
PPI policies was not as responsive to changes in its PPI price as we would expect
in a competitive market. 1
• There was little evidence of competition on non-price factors.
• The high margins earned on PPI policies indicated that the responsiveness of
demand for PPI to changes in PPI prices was low.
3.4 We consider that the above factors are compelling evidence to conclude that the PPI
demand faced by distributors and intermediaries is not as responsive to changes in
their PPI prices as we would expect in a competitive market. 2 This in turn indicates
that there is only limited substitutability between PPI policies offered by different
distributors and intermediaries and also limited substitutability between PPI and
alternative insurance.
1
Hereafter we refer to the sensitivity of the demand for PPI faced by an individual distributor to changes in the price(s) charged
by that distributor as ‘the responsiveness of PPI demand to PPI prices’. Similarly other references to ‘the sensitivity of demand’
in this section refer to the sensitivity of an individual provider’s demand to a change in its own PPI prices.
2
If distributors and intermediaries were in direct competition with each other we would expect that small changes in prices
would result in companies winning or losing customers to their rivals; as customers substitute away from high-priced policies to
cheaper policies offered by rival providers. We would therefore expect the responsiveness of demand for a provider’s policy, to
the price of that policy, to be relatively high.
45
3.5 However, as well as assessing the cumulative effect of all competitive constraints on
distributors of PPI policies, we also assessed each potential source of constraint on
distributors of PPI policies individually. 3 In particular, we assessed the evidence on
each potential substitute for a PPI policy sold by an individual distributor. The
alternatives we considered were other PPI policies offered on a stand-alone basis by
distributors, PPI policies offered by stand-alone providers, PPI policies offered by
mortgage intermediaries and alternatives to PPI such as income protection
insurance, critical illness insurance and life insurance. In each case we concluded
that these individual possibilities for substitution were not likely to impose a
significant competitive constraint on distributors and intermediaries when selling PPI
policies to their credit customers.
3.6 We also considered the potential constraint imposed on PPI distributors by the
complementary relationship between PPI and the credit products that they insure. If
high PPI prices induce consumers to switch to another provider of credit and PPI,
then such a PPI distributor will lose the profit made on the PPI and on the sale of the
underlying credit product. This can lead to an additional incentive for distributors to
offer lower PPI prices. However, we found this incentive to be low because of the
following factors:
• For the reasons set out in paragraph 3.3, we found that the responsiveness of
consumers in general to changes in PPI prices was low. The responsiveness of
credit sales to changes in PPI prices was lower still. 4
• Evidence from the distributors’ internal documents did not show that distributors
were deliberately or consciously constraining their PPI prices because of the
prospect that high PPI prices might damage their credit sales.
• The CC GfK NOP 2008 survey indicated that the majority of search was for
combinations of PPI and credit; however, the incidence of search was low. CC
analysis indicated that this level of search was either insufficient, or only
marginally sufficient, to constrain current high prices.
• CC analysis of distributors’ sales data did not show any compelling evidence that
increases in PPI prices in the past had resulted in a significant reduction in credit
sales.
• The profit margins that PPI distributors earn on the underlying credit product are
far lower than those earned on PPI. If firms lose a sale of a combination of PPI
and credit because of high PPI prices, then the loss of the PPI element of that
combination is relatively more important and the loss of the credit element
relatively less important in terms of overall profits. 5 Credit sales therefore have a
proportionately lower impact on profits, and so firms’ incentives, than the loss of a
PPI sale.
3
Whilst we have assessed each potential source of substitution individually, it is the cumulative effect of all substitution
possibilities which is determinative for our market definition exercise since it is the effect of a price rise on the level of sales
achieved by a distributor which will determine whether a price rise (SSNIP) above competitive levels is profitable.
4
The responsiveness of PPI demand to changes in PPI prices (the own-price elasticity of demand) is driven by substitution
between PPI policies on a stand-alone basis, and also by substitution on the basis of a combination of PPI and credit. The
responsiveness of credit demand to a change in PPI prices (the cross-price elasticity of credit demand with respect to PPI
prices) is driven by substitution of combinations of PPI and credit only and will therefore be lower than the own-price elasticity. It
therefore follows that if we find a low own-price elasticity, the competitive constraint from credit is also likely to be low.
5
Although this also depends on the relative proportions of the price of the combination made up by PPI and credit respectively.
For example, in the case of combined PPI and unsecured personal loans, the PPI element of the combination accounts for
more than half of the overall price of a combined PPI and loan and also the profit margins on credit are low. The combination of
these two factors means that firms’ incentives are relatively unaltered by the prospect of losing credit sales. Where high prices
result in lower sales of combined PPI and loans, it is therefore the loss of PPI sales which drives the impact on profits, and the
loss of credit sales is relatively unimportant. This topic is discussed in greater depth in Appendix 3.7.
46
3.7 Our analysis of the cumulative constraints imposed by all of the potential sources of
consumer substitution listed in paragraph 3.5 showed that distributors’ and
intermediaries’ current prices were either unconstrained or only marginally
constrained. Combined with the evidence in Section 4 on the general level of PPI
prices, this led us to the conclusion that the relevant product market was no wider
than the sale of PPI by a distributor or intermediary to its own credit customers.
However, we found that whilst PPI sold by distributors and intermediaries to their
credit customers is not competitively constrained by stand-alone PPI, asymmetric
constraints mean that stand-alone PPI is constrained by PPI policies sold by
distributors and intermediaries.
3.8 We concluded that the geographic market was the area within the UK in which a
distributor or intermediary sells PPI.
Introduction
3.9 We considered the appropriate product and geographic market definitions, applying
the methodology set out in our guidelines. 6 We do not regard market definition as an
end in itself, but rather as a framework within which to analyse the effects of market
features.
3.10 Our guidelines state that the key to market definition is substitutability—the extent to
which consumers can readily switch between products, or suppliers can readily
switch their facilities between the supply of alternative products. 7
3.11 In assessing the likelihood of this substitutability, we applied the SSNIP (small but
significant non-transitory increase in price) test (also known as the hypothetical
monopolist test). The principle behind this test is that a market is defined as a
product, or collection of products, the supply of which can, hypothetically, be
monopolized profitably. The CC normally uses a 5 per cent increase above competitive levels for the SSNIP test because, in many instances, an increase in the price of
a product of around 5 per cent might reasonably be judged to have a significant
effect on consumers’ expenditure. This is the figure we used in our analysis. 8
3.12 In a market investigation, applying the SSNIP test in practice can be made difficult
because of the cellophane fallacy. 9 If a market is less than fully competitive, prices
may already be in excess of competitive levels. At higher price levels consumers
tend to display a higher propensity to substitute to other products than they would at
lower price levels. In other words, the elasticity of demand tends to be higher at
higher price levels. Evidence of substitutability at current price levels may
erroneously lead to the conclusion that a SSNIP from the competitive level would be
unprofitable, when the reverse is true. This could lead to a market definition that is
too broad. 10 If the SSNIP test shows that a price rise would be unprofitable at current
prices, because two goods are substitutable at current prices, this may be attributed
to genuine substitutability, or alternatively to the cellophane fallacy. Such a finding
would, therefore, be inconclusive. However, if the test shows that a price rise at
current prices would be profitable, and that there would be insufficient substitution to
alternative products, this indicates that these products are poor substitutes under all
6
CC3. CC3. 8
That is to say, we looked at the effect of a 5 per cent rise in price of PPI, not a 5 per cent rise in price of a combination of credit and PPI together, nor a 5 per cent rise in credit APR. 9
CC3, paragraph 2.9. 10
CC3, paragraph 3.29. 7
47
circumstances and this can be taken as good evidence that the products in question
constitute distinct economic markets.
3.13 Market definition can seem a complex, technical exercise for those that are not
familiar with competition analysis. Often, the markets that are identified for the
purposes of competition analysis are different from those that are commonly referred
to by businesses or consumers. In many cases, businesses or consumers refer to
entire industries or product sectors when using the term ‘market’. 11 However, the
more structured approach that we employ in this section, as well as in other CC
investigations, ensures that we have defined the market in a way that provides a
sound framework for analysing the extent of competition.
3.14 In addition, the definition of the relevant market is not necessarily unique. It is
possible that substitutability between products affects those products in different
ways and the competitive constraint on each product is different. This can lead to an
asymmetric market definition which is different depending on the starting point for the
definition exercise. 12
3.15 We looked first at the product market, using the SSNIP test starting from the
narrowest possible market definition—different levels of PPI cover sold by a
distributor to insure a particular credit product it offered consumers—and working
progressively wider. We looked separately at the role played by mortgage
intermediaries, and the effect they have on the market for MPPI and hence for the
market definition for that product. We then looked at the link between credit and PPI
in order to understand the extent of any constraint imposed on PPI prices by the
relationship between distributors’ PPI prices and sales of credit and whether, as a
result, credit products and PPI policies should be considered as being within the
same relevant economic market. We also looked at whether the characteristics of
consumers of non-standard credit products were such that a different product market
definition is appropriate for providers of PPI for those consumers.
3.16 We then went on to consider the geographic market.
3.17 Finally, an alternative view on how to assess the competition between PPI providers
was put to us. This was that we should look at the extent of competition between PPI
providers and consider whether it needs improving—in effect, not using our market
definition as the tool for analysing the extent of competition between providers. We
considered whether this was an appropriate way forward, and whether our
conclusions would be affected by whether we used our market definition or used this
alternative methodology (see paragraphs 3.142 to 3.150).
11
For the purposes of defining a market for competition analysis, we are most interested in those options that customers have
for switching their purchases now or in the near future. We will also take into account options that might become available to
consumers in the longer term, but this will be in the context of our competition analysis rather than in defining the market. This
approach can result in differences between the way in which competitive constraints are identified for the purposes of market
definition compared with the way in which a business might perceive the competitive constraints it is facing over the longer
term. In some cases this difference may have little practical effect. However, in other cases, differences might arise as a result
of different assessments of the extent of these longer-term competitive threats.
12
For example, if the level of substitution between product A and B is sufficient to constrain product A, but insufficient to
constrain product B (for example, because of thinner profit margins, or because B is sold in greater volumes), then product A
will sit in a wider market for both A and B. However, as B is unconstrained by substitution to A, the market definition for product
B would be narrower, consisting of product B only.
48
Product market
The responsiveness of PPI demand to PPI prices
3.18 As a first step in defining a market, we often assess the responsiveness of demand
to changes in price (the own-price elasticity of demand). 13
3.19 The responsiveness of demand to changes in price is dependent on the level of
substitutability and so is an important piece of evidence when defining the relevant
market. Where demand is relatively responsive to prices (the elasticity of demand is
high) small changes in price result in large changes in the level of sales achieved. A
high elasticity of demand therefore indicates a high degree of substitutability and
potentially wider markets. Similarly, where demand is relatively unresponsive to
changes in prices, it is likely that substitutability is low and a narrower market may be
appropriate. The responsiveness of demand to changes in price is driven by the sum
of all the constraints on a firm. 14
3.20 In the case of PPI, the overall responsiveness of demand for PPI policies sold by a
distributor or intermediary to changes in its PPI prices will be driven by substitution
away from the firm to other substitute products. This could include substitution to
stand-alone PPI, to alternative types of insurance, consumers choosing not to take
out insurance at all and PPI sales lost when consumers substitute to other
combinations of PPI and credit.
3.21 Evidence on the own-price elasticity of demand can therefore provide information on
the cumulative effect of all constraints on a firm but it cannot tell us about the nature
or relative strength of individual constraints, such as substitution to a particular rival
firm or alternative product. In paragraphs 3.34 to 3.118 we considered separately the
evidence on the potential sources of competitive constraint we identified.
3.22 We received evidence from a number of the parties regarding the responsiveness of
demand to changes in the PPI prices that they charge. The majority of distributors
told us that there was limited responsiveness of PPI demand to changes in PPI
prices, or that they were unable to determine the responsiveness of demand to the
PPI prices that they charge. Changes in PPI prices that they had introduced in the
past had resulted in little identifiable change in their sales of PPI policies.
3.23
[] told us that in June 2005 it increased the price of its PLPPI policy, and that by
late 2005 its penetration rate had declined from around 60 per cent to less than
50 per cent. Although there was a reduction in the credit rate in July 2005, which may
have meant that [] was attracting more price-sensitive customers who may have
been less likely to take PPI, [] said that it believed that it lost PPI customers as a
direct result of the increase in PPI prices. We did not find this evidence persuasive.
First we noted that there was a downward trend in PPI penetration rates over time
which could also be a reason for the reduction in PPI sales observed by []. In
addition we note that even were we to find that, in this instance, PPI penetration was
13
The own-price elasticity of demand is the proportionate change in sales of a PPI policy to a change in the price of that PPI
policy.
14
This will include substitution to any and all alternatives to a PPI policy sold by a distributor or intermediary, and also lost sales
through customers choosing not to buy insurance at all. The own-price elasticity of demand is therefore a measure of the
cumulative effect of all competitive constraints on a firm. Therefore, where the own-price elasticity of demand is low; all of the
individual constraints which make up the total effect on PPI demand must also be weak since, in sum they result in a relatively
small loss in sales as a result of a distributor raising its PPI price. One distributor ([]) said that the CC only calculated
elasticities of demand for PPI, without seeking to calculate the extent to which customers would divert to other providers of PPI
or to other types of insurance products in response to a relative price increase. For the reasons set out above we do not
consider this necessary, and we note that nonetheless we have considered each of the potential competitive constraints on
distributors individually in paragraphs 3.34–3.136.
49
affected by PPI prices, this may not necessarily indicate that demand is sufficiently
responsive to warrant a wider market definition. We assess the effects of PPI price
changes, accounting for general trends in the data and the degree of substitution, in
Appendices 3.2 and 3.3. [] found that when it increased its CCPPI charge from 69p
per £100 of outstanding balance to 89p and then 99p, for sub-prime customers, there
was no decline in response. It noted, however, that the increases took place between
1999 and 2001 and might not represent current market conditions; moreover, it did
not believe that this was representative of the wider market.
3.24 HBOS introduced risk-based pricing for PLPPI in late 2006. This resulted in some
price changes to certain segments of PPI consumers. HBOS said that the results of
risk-based pricing showed a significant impact on both PPI penetration and take-up
of credit from customers within the HBOS sales channel. Because the analysis
suggested that a price rise would be profitable, HBOS concluded that there must be
substantial additional commercial pressures affecting pricing that this analysis was
not capturing. 15 However, we noted that HBOS did not assess the effect of PPI prices
on the aspects of PPI demand which its analysis did not cover (for example, by
assessing the impact of PPI price changes on the volume of loan illustrations). We
noted that on the aspects of PPI demand its analysis did cover the impact of changes
in PPI prices was small. In the absence of any direct evidence of PPI prices having a
substantial impact on demand we found HBOS’s arguments regarding the sensitivity
of PPI demand demonstrated by its analysis unconvincing.
3.25 Another distributor ([]) told us that it had run a pricing pilot on what it thought were
its most price-sensitive customers which resulted in a 62 per cent change in
penetration in response to a 34 per cent change in price for its PLPPI. Alliance &
Leicester told us that demand for PPI was relatively price inelastic and changes in
price in recent years had little effect on penetration rates. RBSG told us that it was
very difficult to assess the effect of PPI prices on demand for PPI, but in its view
consumers in general had a low sensitivity to PPI prices compared with their
sensitivity to credit product pricing changes.
3.26 This view was confirmed by the distributors’ internal documents. We found numerous
examples within the distributors’ pricing and strategy documentation of the low
responsiveness of PPI demand to PPI prices. Appendix 3.1 provides a fuller
description of our findings regarding the distributors’ internal documentation. A
summary is provided in Table 3.1.
15
HBOS estimated that the elasticity of demand implied by these results was less than -0.56. However, HBOS told us that its
analysis only measured part of the elasticity of demand, because it did not measure the potential impact of PPI prices on the
total demand for protected credit and because it did not capture the impact of changed prices on pre-point-of-sale searchers.
HBOS told us that the low elasticities that it had found with its analysis must mean that there were substantial additional
commercial pressures affecting pricing which were not picked up by its analysis; as otherwise, with elasticity estimates on this
scale, one would have expected HBOS to have profitably raised prices by a significant proportion.
50
TABLE 3.1 Price sensitivity quotes
PPI provider
PPI product
referred to
Price sensitivity quotes
[Distributor A]
CCPPI
‘Rough’ testing in 2000 suggests that there is little price sensitivity for PPI, but that
prices over [] do hit conversion
[]
PLPPI
Price increase ... penetration continued to increase & no impact from price increase ...,
price increase in April shown no impact on sales. [] experience price has very little
impact on propensity to take up of UPL
[]
CCPPI
... for price rises we have seen very little change in demand and we have no experience
of price reductions and whether this would result in any significant increase in demand
[]
MPPI
MPPI [] is simply priced at [£5 to £6] per £100 of monthly cover, on which we earn
[]% commission. This is towards the higher-priced end of the market (comparing to
c.£3.00 for cheaper standalone broker products) Our current penetration levels suggest
pricing is not a significant barrier to the sale. However, we will continue to benchmark
our pricing against that of similar providers and ensure we do not become exposed to
external attention from regulators, the press consumer groups, who focus on the
payment protection market on a regular basis.
[]
PLPPI
[] price rise introduced with no adverse reaction.
[Distributor A]
CCPPI
HR test results: new customers seem to have no price sensitivity
[]
PLPPI
It is not anticipated that a price increase would be detrimental to overall penetration
rates. Industry experience indicates that PPI take up is largely price inelastic.
[]
CCPPI
Sales would not decrease with the higher prices—sales are not price-sensitive and the
product is sold to need*
[]
PLPPI
[PL]PPI sales are relatively price inelastic; the focus for consumers is primarily when
comparing the cost of the main loan product against competitor offerings. However
there is an external focus on the relative value for money.
[]
PLPPI
Loans PPI: reduce discount from []% to []%—very little impact on PPI sales†
[]
PLPPI
Price increase in unemployment cover []. Previous experience of [] increase is little
impact on sales
[Distributor A]
CCPPI
Main reason for customers cancelling is price
[]
PLPPI
The PLP price increase from [] per £100 to [] per £100 was implemented as
planned on []. We received no adverse customer reaction whilst adding a further []
to monthly earned premium.
[]
PLPPI
In [], commission levels were reduced from []% to []% in order to generate
additional [] value and increase strike rate. No material increase in strike rate was
achieved. Whilst the monthly cost of [] to the customer will increase by c.15% ...
customers' total monthly loan repayment loan with [] will increase by c.1–2% ... no
detrimental impact on loan volumes or strike rate is envisaged.
[]
CCPPI
Cards: PPI is rarely bought on the basis of price and an increase is not expected to
significantly impact take-up or cancellations. [] cards increase price of PPI product in
[] and did not experience any negative effect on either take-up of new policies or
cancellations
[]
PLPPI
Despite reducing PPI discount we have seen an increase in take up ... largely down to
attracting less price-sensitive customers with higher APR; Propose to remove discount
to further test price sensitivity
Source: CC, quotes taken from documents provided by the parties.
*[]
†[]
3.27 In addition to an analysis of the distributors’ internal documents, we also undertook
an extensive analysis of the distributors’ sales data for PLPPI, MPPI and CCPPI
using econometric techniques. The purpose of this analysis was to determine the
level of responsiveness of demand for a particular PPI policy to changes in the price
of that policy. The results of this analysis are provided in Appendices 3.2 and 3.3.
Our analysis showed that for PLPPI there was little evidence of price sensitivity and
we concluded that, at the low values for the elasticity that we found, PLPPI
substitution would be insufficient to constrain prices at their current levels. For MPPI
51
we found some evidence of price sensitivity but these results were highly sensitive to
the specification and range of PPI policies used in the estimation. For credit cards we
analysed data on individual customers provided by one large distributor ([]). We
found some evidence that there had been a statistically significant response to an
increase in CCPPI prices in 2007; however, in the example where we found a
statistically significant response the magnitude of this response was low, indicating a
low responsiveness of CCPPI demand to CCPPI prices.
3.28 We commissioned GfK NOP to undertake a survey of purchasers of PLPPI, MPPI
and SMPPI. The results showed relatively small numbers of purchasers shopped
around for PPI prior to purchase of their current policy (see Table 3.2). This is
consistent with modest elasticities of demand. Our estimates of the own-price
elasticities of demand implied by our survey results were between –1.2 and –2.2. Our
analysis of the GfK NOP survey results, including our consideration of the degree of
substitution necessary to define a wider market, is presented in more detail in
Appendix 3.9.
TABLE 3.2 GfK NOP 2008 survey results and implied elasticities of PPI demand
MPPI
Comparers
Non-comparers
Others
Comparers:
- Compared combination of loans and stand-alone PPI policies - Compared combination of PPI and loans only
- Compared stand-alone PPI policies only
- Only made simultaneous credit applications - Other method To summarize
Comparers
- Stand-alone only - Bundle (including only made simultaneous credit applications and other method)
Implied elasticity of PPI demand* per cent
SMPPI
PLPPI
21.3
65.4
13.3
11.3
80.8
7.9
12.3
82.3
5.4
7.2
7.0
4.9
0.6
1.5
1.6
6.9
0.6
1.8
0.6
2.5
6.5
1.6
0.2
1.5
21.3
4.9
11.3
0.6
12.3
1.6
16.4
10.7
10.7
–1.28
–2.16
–1.71
Source: CC analysis.
Note: The figures in the table were corrected for consumers who wrongly classified themselves. See Appendix 3.4: Evidence
on search, paragraphs 18–21.
3.29 Based on the evidence from these four sources we concluded that the responsiveness of PPI demand to changes in PPI prices is low. This indicates that there is only
limited substitutability between PPI policies sold at the point of sale and other alternatives and that price rises would result in only limited substitution of any form. We
considered that the level of demand responsiveness to PPI prices implied by the
evidence from these four sources was inconsistent with effective competition
between distributors for PPI.
The substitutability of different levels of PPI cover sold at a point of sale
3.30 Having considered the available evidence regarding the cumulative constraints on
PPI providers, we go on to consider each potential source of competitive constraint
individually.
52
3.31 We begin by examining the narrowest conceivable product market, that of an
individual policy. 16
3.32 As noted in paragraph 2.3, some PPI distributors offer different levels of PPI cover at
the point of sale of their credit products. Consumers may be offered a choice of cover
levels which may include a combination of U cover, A and S cover and L cover.
3.33 It was unclear from the available evidence whether consumers would substitute
between these various options in response to changes in relative prices. However,
there appeared to be no particular barriers to the distributor present at the point of
sale offering any configuration of cover levels to its own credit customers. We
therefore considered that supply-side substitution within a distributor would warrant a
wider market to encompass all the cover levels offered at the point of sale by that
distributor.
The substitutability of PPI policies offered by different distributors
3.34 We considered next whether the potential for substitution between PPI policies
offered by different distributors constrained distributors from raising prices. There are
two ways in which consumers can substitute between distributors’ PPI policies. First,
they can do so directly by substituting to a stand-alone PPI policy, or short-term IP.
Second, consumers can substitute between combinations of PPI and credit.
3.35 In this section we consider substitutability of the first type—between PPI sold in
combination with credit and stand-alone PPI (or short-term IP). We consider this for
all types of PPI policies (for example, PLPPI policies, MPPI policies), and then focus
on whether the presence of mortgage intermediaries makes a difference to the
analysis for MPPI products. We discuss substitutability of the second type, between
combinations of PPI and credit (from paragraph 3.90).
3.36 First we considered whether PPI policies sold at different points of sale could
constrain each other. Most PPI policies sold at the point of sale are not available on a
stand-alone basis. We found evidence of only four MPPI policies sold in combination
with mortgages that were also sold to customers with mortgages held elsewhere (of
the largest distributors, Abbey, Alliance & Leicester, Barclays and HBOS currently
offer MPPI for sale in this way), accounting for less than 4 per cent by number of
MPPI policies sold and less than 2 per cent by GWP of MPPI sold). We did not find
that these provided a constraint.
3.37 We found that many distributors monitored the prices of PPI sold by other distributors
to their own credit customers. Whilst this evidence could be consistent with a degree
of substitutability between PPI policies sold at different points of sale, we did not
consider this evidence conclusive. First, the costs of monitoring are low, it may
therefore be in the interests of distributors to monitor rival PPI offerings even if the
competitive threat that they present is relatively weak. Furthermore, we considered
that monitoring of other distributors’ PPI prices was not necessarily inconsistent with
a lack of substitution between PPI policies sold at different points of sale. Distributors
16
Barclays said that given the economic features of PPI and evidence gathered by the CC this candidate market was not a
viable starting point. Barclays said that the primary evidence in the provisional findings report suggested that the correct
starting point should be the combined credit and PPI product seen as a bundle. RBSG said that it only made sense to look at
PPI and credit together. RBSG said that this was because PPI was marketed together with the underlying credit product and
the majority of customers preferred to buy credit and PPI together. RBSG said that the products were priced and profitability
was monitored as part of the overall credit and PPI product family. Our approach to market definition, where there is a
complementary relationship between products, is set out in Appendix 3.7. In this appendix we set out the conditions under
which it is appropriate to conclude there is a wider ‘system’ market for the combined credit and PPI product. We assess
whether those conditions are met in paragraphs 3.90–3.129.
53
may monitor prices for other reasons, for example, to avoid any bad publicity that
might arise through having PPI prices significantly above those of other distributors. 17
We therefore considered that this evidence was insufficient to conclude that there
would be sufficient substitutability in the event of a price rise to render a price rise
above competitive levels unprofitable.
Substitutability of stand-alone PPI and short-term income protection
3.38 As discussed in paragraph 3.36, most PPI policies sold at the point of sale are not
available on a stand-alone basis. However, there are other PPI policies and shortterm IP policies sold on a stand-alone basis. These are mostly sold by companies
which do not offer credit—either underwriters (such as Cardif Pinnacle), companies
specializing in selling insurance (such as Paymentcare and British Insurance), or
companies selling a variety of products (the Post Office). There have been some new
entrants into the sale of stand-alone PPI or short-term IP since the inquiry began:
Barclays, whose Plan B product, sold with Barclaycard, can also be extended to
provide CCPPI on other credit cards; Churchill (part of RBSG), which has started
selling short-term IP; and HSBC, whose LifeChoices product offers short-term IP.
Plan B is sold both to insure Barclays credit cards at the point of sale and on a standalone basis. HSBC mortgage customers are offered LifeChoices at the point of sale
subject to suitability, and personal loan and credit card customers are offered a
referral to discuss their protection needs (not just those related to the credit product)
with a LifeChoices purchase one of the possible outcomes; LifeChoices is also
offered on a stand-alone basis. Our focus in this section is on their substitutability for
other PPI policies when they are sold on a stand-alone basis. Where short-term IP is
offered at the point of sale of credit, we consider that it is effectively PPI sold at the
point of sale. 18
3.39 Stand-alone PPI and short-term IP are predominantly aimed at, or suited for, protecting stable credit repayments such as mortgage and loan repayments. We found only
two providers offering stand-alone substitutes for CCPPI sold at the point of sale:
Paymentcare and, as of 2007, Barclays (which is available to customers who hold a
Barclaycard and want to insure other credit cards as well).
Evidence on the number of policies sold
3.40 We looked at the number of stand-alone PPI and short-term IP policies sold.
Table 3.3 sets out the sales of different policies. On the basis of sales data from each
provider, we found that [20,000–30,000] stand-alone MPPI policies were sold in
2007. Four providers reported annual sales of between 0 and 2,500, one reported
sales of 2,500 to 5,000, and one provider reported sales of over 10,000 stand-alone
MPPI policies. One provider was unable to provide details of its stand-alone MPPI
sales. Preliminary figures suggest that stand-alone MPPI sales will increase in 2008,
although this growth is driven almost entirely by the sales of a single provider. For
stand-alone PLPPI, we identified two providers, one of whom reported sales of
between 0 and 2,500 in 2007, the other sales between 2,500 and 5,000 in 2007.
Total stand-alone sales for PLPPI reached approximately [2,500–7,500] in 2007. For
CCPPI, we identified two different products; [0–2,500] stand-alone CCPPI policies
17
This was reflected in the internal documents of one distributor ([]). We found the following quote in an internal document:
‘MPPI [] is simply priced at [£5 to £6] per £100 of monthly cover, on which we earn []% commission. This is towards the
higher-priced end of the market (comparing to c.£3.00 for cheaper standalone broker products) Our current penetration levels
suggest pricing is not a significant barrier to the sale. However, we will continue to benchmark our pricing against that of similar
providers and ensure we do not become exposed to external attention from regulators, the press consumer groups, who focus
on the payment protection market on a regular basis’.
18
See Appendix 2.3.
54
were sold in 2007, with both distributors reporting annual sales of between 0 and
2,500. Finally, we found that a few large distributors offer short-term IP policies. We
found that 10,000 to 15,000 policies were sold in 2007. Four short-term IP providers
reported sales of between 0 and 2,500 and three providers reported 2,500 to 5,000
policies sold. Some of these sales were made as a result of referrals made during the
sale of mortgages, personal loans and credit cards (see Appendix 2.3). Again
preliminary figures for sales during the first half of 2008 suggest that short-term IP
sales will likely increase during 2008.
55
TABLE 3.3 Provision of stand-alone PPI and short-term IP policies
Sales
Credit product
2006
2007
First half
2008
Barclaycard Plan B for
multiple cards
Credit Card Payment
Protection Insurance
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
Company
Name of policy
Credit cards
Barclays
Credit cards
Paymentcare
Personal loans
Personal loans
British
Insurance
Paymentcare
Loan Payment Protection
Insurance
Loan Payment Protection
Insurance
First-charge mortgages
Abbey
First-charge mortgages
Alliance &
Leicester
First-charge mortgages
Barclays
First-charge mortgages
First-charge mortgages
First-charge mortgages
British
Insurance
Cardif Pinnacle
*
HBOS
First-charge mortgages
HSBC†
First-charge mortgages
Paymentcare
Paymentcare Mortgage
Payment Protection
Bill Protector (until May
2007)
Mortgage Payment
Protection Insurance (from
February 2008)
Openplan protect and
mortgage care
Mortgage Payment
Protection Insurance
Helpucover
Total Mortgage Protection
Plan
Mortgage Repayment
Protector
Mortgage Payment
Protection Insurance
Second-charge mortgages
British
Insurance
Short-term IP
Short-term IP
Short-term IP
Barclays
HBOS
HSBC†
Short-term IP
Short-term IP
MBNA
NHI Services
Ltd
Post Office
Financial
Services
RBSG
Short-term IP
Short-term IP
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
Lifestyle Protector []
Bill Protector
LifeChoices (launched
June 2007)
Extra Cash (pilot)
Payprotect
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
Lifestyle
[]
[]
[]
Churchill Income Protection
[]
[]
[]
Source: CC based on information from the parties.
*
Following correspondence, we discovered that most of the MPPI sales reported by HBOS to be stand-alone were in fact sales
to remortgage customers. HBOS said that the number of actual stand-alone sales was in reality extremely small, estimating
that around [] per cent of the sales which it reported as being stand-alone were true stand-alone sales. We have calculated
the figures in Table 3.3 on this basis.
†HSBC told us that it was not currently selling PPI except through its HFC brand. HSBC currently offers to customers to whom
it previously offered PLPPI and CCPPI the chance to discuss their broader protection needs with one possible outcome being
to take LifeChoices (if it is the most suitable product), a short-term income protection product. Customers who were previously
offered MPPI are offered LifeChoices subject to suitability. The sales figure quoted here is derived by multiplying the total
number of LifeChoices sales with an Accident, Sickness or Unemployment component by [] per cent—the proportion of
LifeChoices sales which were identified in Appendix 2.3 as ‘stand-alone’ in the first half of 2008.
Notes:
1. N/A = not applicable.
2. N/P = not provided.
3.41 For PLPPI and CCPPI, stand-alone sales are very small compared to the total number of PPI policies sold by distributors; the 12 largest distributors between them sold nearly 1.5 million PLPPI policies and almost 1.3 million CCPPI policies in 2007 (implying that the stand-alone market accounts for less than 0.5 per cent of total PLPPI sales, and less than 0.1 per cent of total CCPPI sales). After performing some rough adjustments to account for sales by smaller distributors and intermediaries (using data on the total value of premiums earned by these parties in 2006), we 56
estimated MPPI sales in 2007 to be somewhere in the region of 300,000. This
suggests that while stand-alone sales may make up a more significant proportion of
the MPPI market than is the case for other PPI markets (a little under 9 per cent), the
extent of MPPI policies sold on a stand-alone basis is still very small.
3.42 The evidence indicated to us that, in practice, sales of stand-alone PPI and shortterm IP were small relative to the size of the market and do not currently act as a
competitive constraint on distributors selling any PPI product at the point of sale.
They are therefore not within the same market as PPI sold at the point of sale.
3.43 It was put to us that sales of stand-alone products were growing and would
increasingly become a constraint (though we were provided with no internal documentation showing that distributors thought that they would grow substantially). We
were told that the recent introduction of stand-alone products by Barclays, HSBC and
RBSG was likely to impose significant competitive constraints on distributors and
intermediaries when selling PPI in combination with credit, because, in contrast to
other attempts at stand-alone entry, these providers already had access to large
numbers of credit customers to whom they might attempt to cross-sell PPI. 19 We
were told that these companies would be unlikely to launch stand-alone policies
unless they were expected to be a commercial success, and the fact that three large
high street banks had all launched stand-alone policies was, in their view, telling. 20
We consider the extent to which we expect stand-alone sales to constrain sales of
PPI policies at the point of sale in the future in our section on the analysis of
competition (see paragraphs 5.120 to 5.134).
Evidence of consumer switching
3.44 We looked next at evidence to see whether consumers switch to stand-alone PPI.
3.45 We looked first at evidence on consumer awareness of the possibility of buying PPI
from providers other than their credit provider. The CC’s BMRB 2007 qualitative
research found that there was low awareness among the group participants that it is
possible to shop around for a separate or different PPI policy to the one they had
been offered with their product. Although group participants were aware that it was
possible to shop around for other insurance policies, they had not thought it possible
to do so for PPI. Exceptions to this pattern were PPI consumers who had used a
financial adviser for the purchase of their PPI (which was mainly MPPI).
3.46 The CC’s GfK NOP 2008 quantitative survey found that many PPI consumers who
did not compare products from different providers believed that they could only buy
the PPI policy from their credit provider—42 per cent of MPPI non-comparers (based
on 792 respondents), 72 per cent of SMPPI non-comparers (based on 1,836) and
73 per cent of PLPPI non-comparers (based on 2,359). 21
3.47 The CC BMRB 2007 quantitative survey gave similar, though slightly lower, figures
for each category. When asking all consumers (both those who compared products
and those who did not) whether they believed that they could only buy a PPI policy
from their credit provider it found that 31 per cent of MPPI consumers (based on 477
respondents), 61 per cent of SMPPI consumers (based on 617), 60 per cent of
PLPPI consumers (based on 588) and 65 per cent of CCPPI consumers (based on
601 respondents) thought that PPI could only be bought from the credit provider.
19
LECG paper in support of Barclays’ response to Emerging Thinking, paragraphs 2.11&2.12. LECG paper in support of Barclays’ response to Emerging Thinking, paragraph 2.16.
PPI search behaviour: a research report for the Competition Commission, GfK NOP, April 2008, p22. The bases are shown in
survey tabulations published on the CC website. 20
21
57
3.48 In addition, we found that there was little documentary and statistical evidence
relating to switching PPI on a stand-alone basis but the evidence we found was
relatively consistent, suggesting that switching of PPI by consumers is limited and if
any switching occurred it was most likely to occur for MPPI.
Evidence of search
3.49 Although the evidence on the number of policies sold was, in our view, conclusive on
the extent to which stand-alone PPI and short-term IP are competitive constraints on
PPI sold at the point of sale, we also looked at evidence of search for PPI policies on
a stand-alone basis and for short-term IP. Further details of evidence on search can
be found in Appendix 3.4. We focused on the four products for which we had survey
evidence on search: MPPI, PLPPI, SMPPI and CCPPI.
3.50 Non-quantitative evidence from the parties was mixed. As set out in Appendix 3.4,
some parties told us that there was little or no search activity for PPI policies, and
others told us that consumers did search in varying amounts.
3.51 Turning to individual types of PPI, we looked first at search for MPPI policies. The
evidence we received from parties suggested that search for PPI was most prevalent
on MPPI, and it was suggested that this could be because of the role of mortgage
intermediaries (see paragraph 3.64).
3.52 The CC GfK NOP 2008 survey found that 21 per cent of MPPI consumers who had
not used a mortgage intermediary to source their MPPI compared two or more MPPI
policies or loans with MPPI in making their choice of provider. Two-thirds of these
consumers compared combinations of credit and PPI and just over half compared
stand-alone products (some of these consumers compared both).
3.53 Table 3.4 shows the search evidence for MPPI, PLPPI and SMPPI.
TABLE 3.4 Search undertaken by a sample of consumers that bought MPPI, SMPPI and PLPPI
per cent
Comparers
Of whom:
Comparers who compared combinations only
Comparers who compared stand-alone only
Comparers who compared stand-alone and
combinations
MPPI
PLPPI
SMPPI
21
12
11
33
23
53
13
61
5
34
20
14
Source: CC analysis based on the CC GfK NOP 2008 survey.
Notes:
1. Comparers based on MPPI 1,257, SMPPI 2,387 and PLPPI 3,156 responses. Methods of comparison based on MPPI 76, SMPPI 64 and PLPPI 60 responses. 2. The figures in the table were corrected for consumers who wrongly classified themselves. See Appendix 3.4: Evidence on
search, paragraphs 18–21.
3.54
We considered this survey to offer a robust estimate of search. However, we did note
other surveys which suggested that the proportion of consumers who shopped
around, or obtained multiple quotes, was higher (26 to 44 per cent). These surveys
did not split out those who searched on the combination of credit and PPI and those
who searched between MPPI policies.
3.55
The data on search activity on stand-alone MPPI policies did suggest some level of
search, but when compared with the low level of stand-alone MPPI sales we did not
58
find that there were sufficient consumers comparing an MPPI policy sold with credit
with stand-alone policies to constrain MPPI distributors’ prices. The information we
received from intermediaries indicated that they carried out limited search when
creating panels of providers of PPI and its equivalents such as short-term IP.
3.56 We looked next at search for PLPPI policies. The CC GfK NOP 2008 survey found
that 12 per cent of PLPPI consumers compared two or more policies in making their
choice of provider. However, of these comparers 13 per cent focused only on
comparing PPI elements, and half focused on comparing combinations of PPI and
loans only. One in five compared both combinations of credit and PPI, and credit and
stand-alone policies.
3.57 Other surveys on PLPPI search provided a very wide range for the proportion of
PLPPI consumers who engaged in search (but did not split out those who searched
on the combination and those who searched between PLPPI products). Estimates
ranged from 4 to 50 per cent (though this latter figure was from a survey conducted
on customers of a company which sought to attract customers by being top of ‘bestbuy’ tables for credit and for PPI, and so would expect to attract a high proportion of
actively-seeking consumers).
3.58 We considered that our survey provided a robust estimate of search behaviour, and
thought that this indicated that there was very little search on stand-alone PLPPI
policies.
3.59 We looked next at sales of SMPPI policies. The CC GfK NOP 2008 survey found that
11 per cent of SMPPI consumers compared two or more policies in making their
choice of provider. However, of these comparers, 5 per cent compared stand-alone
policies only; 61 per cent compared combinations of PPI and credit only; and 14 per
cent made comparisons of both combinations PPI and credit and stand-alone PPI
policies. We considered this to indicate a very small amount of search by comparing
stand-alone policies.
3.60 Finally, we looked at evidence on CCPPI search activity. Surveys conducted for the
OFT and Lloyds TSB estimated the number of customers who shopped around to be
5 to [10–20] per cent. However, we noted that CCPPI consumers can take out a new
card, with CCPPI, without having to cancel an existing card (by transferring the
outstanding balance to the new card). In addition, CCPPI consumers only pay a
premium if there is an outstanding balance and are able to switch without duplicating
costs. To the extent that this type of potential ‘shopping around’ between credit cards
already held by consumers constitutes search, it is unlikely to be picked up by
surveys. Therefore these statistics were less likely to be significant indicators of
search activity.
3.61 We found that the extent of search through comparing a PPI policy with one or more
stand-alone policies was low, with no more than 10 per cent of consumers searching
in this way for MPPI. We concluded that this evidence was in line with our evidence
on the number of stand-alone PPI and short-term IP policies sold, in indicating that
such policies do not constrain the prices of PPI policies sold at the point of sale.
Conclusions on whether stand-alone PPI and short-term IP policies constrain PPI
sold at the point of sale
3.62 The evidence on the number of stand-alone PPI and short-term IP policies was, in
our view, compelling evidence that they do not constrain PPI policies sold at the point
of sale. The evidence on the extent that consumers switch PPI policies and search
59
for the best policy was in line with this. We therefore conclude that stand-alone PPI
and short-term IP policies do not constrain policies sold at the point of sale. 22
3.63 Whilst stand-alone PPI and short-term IP providers do not constrain the prices of PPI
policies sold at the point of sale, we concluded that they do compete to win
customers from across the range of PPI providers, both those who offer PPI in
combination with credit and those who offer it on a stand-alone basis. As such, the
relevant market for stand-alone providers is the supply of PPI to all consumers, and
they are constrained by providers who offer PPI with credit. Our market definition
exercise for the stand-alone provision of PPI is therefore asymmetric—stand-alone
providers are constrained by providers who offer PPI with credit, but stand-alone
providers do not impose a competitive constraint on PPI sold with credit.
Sales of MPPI by mortgage intermediaries
3.64 As noted in paragraphs 2.42 to 2.51, some mortgages are sold via mortgage
intermediaries, who offer MPPI with the mortgage. Not all intermediaries sell MPPI,
or offer it to all their clients. We focus here on those occasions when MPPI is offered
to a client by an intermediary, and the resulting impact on competition.
3.65 If consumers considered purchasing PPI before approaching an intermediary, and
hence had an idea of the products available, as they do with the mortgage product,
then it could help to ensure that they receive the best deal possible. However, the
evidence we received led us to conclude that it is more likely that consumers tend to
use intermediaries for their skills in search for a good mortgage deal. 23 The AMI told
us that consumers that used intermediaries had a better understanding of mortgages
than other financial products but consumers’ knowledge of payment protection was
‘considerably below the understanding of how the mortgage works’. We therefore
considered the role of intermediaries in the sale of MPPI and whether their presence
increases the level of search for MPPI to the extent of widening the market beyond a
distributor-specific one.
• Sales of MPPI by intermediaries
3.66 We looked at how many sales of MPPI were made by mortgage intermediaries.
Tables 3.5 and 3.6 show the proportion of MPPI sales made through intermediaries
compared with other channels. In 2006, 26 per cent of MPPI policies in force and
31 per cent of MPPI policies sold were sold by intermediaries. The proportion of
MPPI policies sold by intermediaries, calculated as the number of new policies sold
in a year and the number of policies in force, has increased considerably since 2002
(after a drop in sales from 2000 to 2002). Over the same period the proportion of
MPPI sales made by lenders decreased. This appears to be driven by intermediaries’
increasing share of mortgage sales. 24 However, the actual number of policies sold by
intermediaries shows a slightly different trend. The most sales made by
22
One distributor ([]) said that our approach to market definition suffered from a binary fallacy and that our approach implied
that any firm with market power should be defined as operating in a separate market, which it considered was inconsistent with
our guidelines. We did not agree that our approach implied that any firm with market power would necessarily operate in a
separate market and we considered our approach to be consistent with our guidelines.
23
Northern Rock told us that its experience was that intermediaries acted on behalf of the customer to search and compare the
market for the mortgage deal that best suited the customer’s needs. Abbey told us that consumers who used intermediaries
would have the ‘shopping around’ undertaken for them. [A large distributor] said that in 2007, 73 per cent of its mortgages were
sold through mortgage intermediaries who, in effect, searched and compared products on behalf of its customers through their
sourcing systems. [Another large distributor] told us that in 2007 it received 47 per cent of its first-charge mortgage volume from
intermediaries. This suggested to [] that, in relation to first-charge mortgages, intermediaries took the place of customer
search.
24
[]
60
intermediaries were in 2003 (163,000); in the three years since then for which this
data is available the number of sales achieved has been reasonably stable at
between 155,000 and 160,000.
TABLE 3.5 New MPPI policies
2000
2001
2002
2003
2004
2005
2006
MPPI provision by channel (% of total)
Gross new
mortgage
advances
MPPI
policies
MPPI policies
% of all new
advances
Intermediaries
Lenders
Insurers
1,681,000
2,087,000
2,511,000
2,766,000
2,606,000
2,142,000
2,253,000
574,100
697,100
875,000
926,100
707,900
541,200
506,900
34
33
35
33
27
25
22
25
19
14
18
22
30
31
74
71
68
63
65
59
59
1
9
18
19
13
12
11
Source: CML Research and Association of British Insurers, www.cml.org.uk/cml/statistics.
Note: The data provided in the first two columns of data is also published by the ABI. However, there are some small
discrepancies between the two published sources.
TABLE 3.6 MPPI policies in force
MPPI policies in force
2001
2002
2003
2004
2005
2006
Mortgages
outstanding
Total
% of all
mortgages
% provided via
intermediaries
% provided
via lenders
% provided
direct
11,247,000
11,364,000
11,452,000
11,512,000
11,595,000
11,719,000
2,456,900
2,570,300
2,717,500
2,492,800
2,456,600
2,297,900
22
23
24
22
21
20
15
14
19
22
23
26
80
80
75
77
76
71
5
6
6
1
1
2
Source: CML Research, Association of British Insurers, www.cml.org.uk/cml/statistics.
Notes:
1. Number of mortgage advances represents the Survey of Mortgage Lenders’ estimate of the number of loans for house
purchase and remortgages.
2. The data provided in this table is also published by the ABI. However, there are some small discrepancies between the two
published sources with regard to the first two columns of data.
3.67 We also looked at the number of mortgages sold by intermediaries, and the extent to
which intermediaries sold a distributor’s MPPI policy in combination with that
distributor’s mortgage. The analysis of this, based on data provided by five of the
largest mortgage distributors, is in Appendix 3.5. We found that in 2006, 43.8 per
cent of their mortgage sales were made through intermediaries. However, only
6.4 per cent of their PPI sales were made through an intermediary. According to the
CML, 31 per cent of MPPI policies in 2006 were sold through intermediaries. This
data shows that intermediaries tended not to source MPPI from the mortgage
supplier. Rather, intermediaries appear to source MPPI from providers that do not
sell credit products. Our analysis showed that nearly all MPPI sales made by
intermediaries were in combination with the sale of a mortgage; very few were standalone MPPI sales (see Appendix 3.5).
3.68 This evidence indicated to us that when intermediaries sell a MPPI policy with a
mortgage, they do not usually sell that distributor’s MPPI policy. Evidence we
61
received from HBOS 25 and Northern Rock (see paragraph 2.50) made the same
point.
• The MPPI policies sold by intermediaries
3.69 Having found that intermediaries tend not to sell the distributors’ MPPI policies, we
looked at what they do sell. There are two ways in which an intermediary or
intermediary network searches for a PPI policy. Either they periodically search the
market, identify a policy or range of policies and contract the distributor(s) to be their
PPI provider(s), or they search across the market, or a subset of the market, for each
individual consumer.
3.70 We looked at the range of MPPI policies sold by the 20 largest mortgage intermediary networks. Our analysis of this is shown in Appendix 3.5. We found that an
intermediary or intermediary network will typically sell one PPI policy or a small
selection of PPI policies. There are exceptions, as in some circumstances intermediaries could source PPI from a wider number of PPI providers. We also found
that the PPI offering for each intermediary does not change very often; rather
intermediaries tend to sell the same PPI providers policies for a considerable period.
For example, Connells has sourced its PPI from Allianz for more than five years and
Countrywide told us that it had only used two PPI providers since 1992.
3.71 Our analysis was not entirely in line with a survey conducted by the AMI, which
suggested that most intermediaries offer more than one MPPI policy (see Appendix 3.5). However, we concluded that the AMI survey results were likely to be more
representative of very small intermediary networks or individual intermediaries, as we
were told that the survey was a poll of intermediaries and the AMI told us that the
respondents were typically small firms of three or fewer employees. This would mean
that large networks and intermediaries are under-represented, and we concluded that
the AMI figures therefore overstate the percentage of intermediary sales based on a
wide panel of providers.
3.72 We considered why intermediaries tended not to offer distributors’ PPI, preferring
other MPPI policies. Our analysis of commission payments to intermediaries (see
Appendix 3.5) suggested that those distributors who offered smaller commissions to
intermediaries tended not to be included on the panel of lenders from which the
intermediaries sourced products. This meant that the PPI policies offered by those
distributors would not usually be available to consumers who bought mortgages from
intermediaries.
• Conclusions on the sales of MPPI by intermediaries
3.73 We found that sales of mortgages and MPPI by intermediaries is significant, at up to
31 per cent of all MPPI sold in a year. Intermediaries tend not to sell the MPPI policy
offered by the mortgage distributor. We found that intermediaries tend to offer a static
25
HBOS told us that ‘[t]he dominant role of brokers in this market, however, results in the mortgage and the insurance being
typically sourced from different manufacturers, albeit sold as part of the same single sales process or transaction’. Data from
Northern Rock showed that although 89 per cent of its mortgage sales were made through intermediaries, only 20 per cent of
sales of its MPPI were made through the same channel.
62
and narrow range of MPPI policies, suggesting a limited amount of search on the part
of intermediaries. 26
3.74 We concluded, therefore, that the presence of intermediaries did not introduce a level
of search which would lead us to a wider market definition for MPPI. Further, we
concluded that intermediaries do not engage in sufficient search for their customers
to be able to rely on this search to receive a competitive offer. We concluded that the
relevant product markets for the sale of MPPI by intermediaries were the supply of
MPPI by mortgage intermediaries to their own customers. 27
The substitutability of other insurance products with PPI
3.75 Having found no evidence of a constraint on a distributor or a mortgage intermediary
selling a PPI product from other PPI or short-term IP sellers, we considered whether
alternative types of insurance offered a competitive constraint.
3.76 The alternative policies that consumers may potentially substitute for PPI include:
(a) IP insurance; 28
(b) personal accident insurance (PA);
(c) CI;
(d) term life insurance (TA); and
(e) decreasing life insurance (DA).
3.77 Views among the major distributors regarding the competitive constraint imposed by
other alternative policies were mixed (see Appendix 3.6). Some distributors did not
think that there were any types of insurance policies that were an effective substitute
for PPI policies. Others considered that, whilst alternative policies did exist, these
were only partial substitutes and offered an insufficient constraint to warrant inclusion
in the appropriate market definition. Some distributors, however, considered that
alternative policies, and in particular IP, did offer an alternative to PPI and had the
potential to provide a powerful constraint.
3.78 There appeared to be a broad consensus among the distributors that IP policies are
the closest alternative to PPI, and therefore the most likely to provide a competitive
constraint on the pricing of PPI policies.
26
Nationwide said that it was wrong to conclude that an intermediary with two MPPI providers on its panel placed no competitive
constraint on the MPPI market because intermediaries would choose between the providers on their panel, even if there were
only two providers on it, to identify the product that best met their customers’ needs. Nationwide said that this introduced a
degree of search for MPPI that was not present in other PPI markets. We did not agree that substitution as described by
Nationwide would confer a competitive constraint on the MPPI market. Substitution of this type would not constrain the pricing
of other MPPI providers because intermediaries would not substitute between providers on the panel and those outside the
panel. Furthermore this type of substitution would not seem to constrain the pricing of PPI policies within the panel, since the
intermediary will have a measure of control over prices through the negotiated commission with each panel member.
27
Nationwide said that the CC's conclusion about the level of search that intermediaries carried out on behalf of their customers
was over-generalizing from the information provided by a small number of intermediaries. It said that the CC's conclusion
ignored the competitive constraint provided by the significant number of intermediaries who carried out more extensive search.
Lloyds TSB said that, even if our proposed approach to market definition was otherwise correct (which it disputed) we should
draw a distinction between the relevant product markets for the sale of PLPPI and CCPPI, and the market for the sale of MPPI.
In its view it would be appropriate to conclude that there was a single market for the sale of MPPI and that this market was
effectively competitive. We do not agree that the available evidence shows that the sale of PPI through mortgage
intermediaries results in effective competition for the sale of MPPI, or that a wider market definition for the sale of MPPI is
appropriate.
28
This is distinct from short-term IP, which is considered in paragraphs 2.12–2.14.
63
3.79 We considered how similar IP was to PPI. Both generally provide cover against A
and S. IP and PPI also both provide a monthly payment rather than a lump sum
(which is more common for PA, CI, TA and DA policies).
3.80 In general, IP policies have longer deferment periods and will pay benefits to the
customer for a longer period (for example, until they return to work, or in some cases
until retirement). 29 They are also typically underwritten individually on the basis of
customer age, health, employment and other demographic factors. 30 PPI policies are
generally pool underwritten (though, as noted in paragraph 2.33, a small number of
distributors have moved to some degree of price variation depending on age and
gender of the customer).
3.81 PPI policies usually relate to a specific underlying credit product, so, for example, the
level of cover is tailored to the payments on a loan or the outstanding credit card
balance. Alternative policies such as IP do not have this feature, although for certain
types of PPI policies where repayments are stable or predictable (eg MPPI, SMPPI,
PLPPI and motor PPI) this can be replicated by taking out IP cover for a similar
amount.
3.82 Finally, in terms of the risks covered by PPI, there appear to be few alternative
policies that cover the same risks as a typical PPI policy. In particular, IP, PA, CI, TA
and DA policies rarely cover unemployment. 31
3.83 We looked at whether PPI distributors took account of alternative types of insurance
when pricing their PPI policies. We sought evidence on the factors taken into account
when setting prices. Typically these factors included costs, underwriting
performance, the level of cover offered and competitor pricing. None of the examples
of distributors’ price comparisons and monitoring that have been provided to the CC
included monitoring of alternative types of insurance (see Appendix 3.1). For
example, [] provided an example of a management report which is used by [] to
assess the competitiveness of its PPI and credit products, which showed that it
monitored PPI products (but not any alternative types of insurance).
3.84 We also considered whether providers of alternative types of insurance policies could
impose a competitive constraint on PPI policies through supply-side substitution.
3.85 The CC guidelines state that supply-side substitution occurs when a price rise
prompts other firms to start supplying, at short notice, an effective substitute for the
policy in question. 32 Supply-side substitution occurs in the short run with little or no
investment required.
3.86 As any insurance provider seeking to switch to the supply of PPI would not have
access to the credit point of sale in the way that a distributor which already sells the
underlying credit product would, general insurers are only likely to be able to
substitute into the supply of stand-alone PPI.
29
[] said that IP policies allowed the customer to select the deferred period before benefit payments started (up to 112
weeks). This allows PPI and IP policies to be combined. According to [], PPI products had a set deferment period of 14 to 60
days. [] said that IP would pay a monthly benefit up to retirement if required whereas PPI provided shorter-term protection, ie
12 to 24 months or (in cases of accident or sickness) for the term of the personal loan.
30
[] told us that IP policies offered cover to a specific age whereas PPI policies offered cover related to the term of lending.
[] said that IP was underwritten and premiums varied by risk whereas PPI was not risk-based and premiums only varied
depending on the credit amount.
31
It is possible that customers could try to replicate a PPI policy by purchasing a suite of products; however, this would be likely
to result in very high transaction costs, as customers would have to fill out multiple applications. As IP, CI and TA tend to be
individually underwritten, the applications processes for these products may take longer.
32
CC3, paragraph 2.19.
64
3.87 The evidence suggested to us that whilst providers of IP, PA, CI, TA and DA
insurance could potentially substitute into the supply of policies which cover similar
risks and have similar terms and conditions to PPI, they are likely to face significantly
increased distribution costs as a result of their lack of access to the point of sale for
the underlying credit product. We discuss the point-of-sale advantage in paragraphs
5.88 to 5.119. It would therefore appear unlikely that the major PPI distributors are
constrained by the potential for supply-side substitution. 33
3.88 We found very little evidence of demand-side substitution between PPI and
alternative types of insurance for all types of PPI. We also found no evidence of such
policies being considered by PPI distributors when scanning the market. The
evidence we received showed that PPI distributors did not consider such policies
when scanning the market (see paragraph 3.83), indicating that they did not, in
practice, consider these alternative types of insurance as potential competitive
constraints.
3.89 We therefore concluded that these alternative types of insurance were not
competitive constraints and, therefore, not in the relevant economic markets. 34
Substitution between combinations of PPI and credit
3.90 As discussed in paragraph 3.34, there are two main ways in which consumers can
substitute between PPI policies. First, consumers can substitute directly between PPI
policies on a stand-alone basis and, second, consumers can substitute between
combinations of PPI and credit. In this section we consider the likely competitive
constraint as a result of substitution of the second type.
3.91 Distributors of PPI typically supply both PPI and the underlying credit product. Where
consumers react to changes in PPI prices by substituting to other combinations of
PPI and credit, the distributor loses both the sale of a PPI policy and also the sale of
a credit product. Where this substitution is sufficient to render a PPI price rise
unprofitable, it is appropriate to define a wider market to include combinations of PPI
and credit.
3.92 In order to determine whether substitution between combinations of PPI and credit is
likely to impose a competitive constraint on PPI distributors when selling PPI, there
are therefore two important questions. First, we need to assess the likely response of
consumers to a change in PPI prices; in other words, if distributors change their PPI
prices, to what extent does this lead to consequential losses of sales of combinations
of PPI and credit? 35 Second, we need to consider the extent to which any loss in
sales of combinations of PPI and credit, as a consequence of high PPI prices, would
alter the incentives of distributors to charge a high price for PPI, which in turn
depends on the relative profitability of credit and PPI. For a fuller discussion of how
the complementary relationship between PPI and credit can competitively constrain
PPI pricing, see Appendix 3.7.
33
A lack of access to the credit point of sale would also impose additional distribution costs on existing providers of PPI seeking
to supply the credit customers of another distributor. Whilst each distributor has access to its own credit customers, it has only
indirect access to the credit customers of others.
34
We note that existing distributors, with access to credit customers at the point of sale, could potentially substitute between the
supply of various types of insurance to those customers, including all the alternatives above. This raises the prospect that the
product market, at the point of sale, may be wider than PPI. However, the incumbent distributor would still have a monopoly
position over the supply of any insurance to those credit customers. This aspect of market definition does not therefore affect
our assessment of competition between suppliers; we therefore did not consider this aspect of market definition further.
35
In other words some estimate of the cross-price elasticity of credit demand with respect to PPI prices.
65
3.93 In order to assess the likely impact of PPI prices on consumer demand for
combinations of PPI and credit, we looked at the following sources of evidence:
• oral and documentary evidence from distributors;
• evidence from the distributors’ sales data;
• evidence of consumer switching behaviour; and
• evidence of consumer search behaviour.
3.94 We then assessed the likely impact that substitution between combinations of credit
and PPI would have on the profitability of distributors.
Oral and documentary evidence
3.95 The evidence we received from the distributors was mixed. A number of distributors
told us that they believed that PPI was part of a wider system market which included
the underlying credit product (see Table 4 of Appendix 3.1 for examples). Some other
distributors noted that PPI and the underlying credit product were linked, but stopped
short of saying that consumers’ decisions in the (primary) credit market would be
affected by prices in the (secondary) market for PPI. For example, Abbey said that
PPI and the underlying credit product were interrelated in the sense that the number
of sales of PPI was directly related to the number of sales of loans. 36 Abbey also said
that it would be dangerous to look at the PPI market in isolation, as it was a
secondary product, which was strongly influenced by the credit product. 37
3.96 However, a number of distributors told us that there was no link between their PPI
prices and demand for their credit, which we consider would indicate that PPI and
credit are in separate markets. For example, Alliance & Leicester said that, if it
changed the price of its PPI, it would not affect the demand for the underlying credit
product and that there was no link between the credit decision and the decision to
take out PPI. Similarly, Lloyds TSB said that if it increased the price of PPI, this
would be unlikely to affect sales of the corresponding credit products in the short
term (though in the longer term it said that dissatisfaction among customers with the
value for money offered would be likely to have a knock-on effect on the reputation of
the group as a whole, with consequential effects on sales of other products).
3.97 Lloyds TSB said that credit and PPI should be viewed as one market as the products
were marketed and purchased together. However, it also noted that competition had
driven down prices of the underlying credit product, and lenders had compensated to
a degree by increasing the price of PPI. Our view is that divergent effects on prices,
as described by Lloyds TSB, are consistent with a low sensitivity of credit demand to
changes in PPI prices.
3.98 We assessed internal documentation supplied to us by the parties in relation to three
key areas: documents that showed how the distributors set their PPI prices;
documents that showed how the distributors view PPI internally, and how they set
their PPI strategy; and evidence of how the distributors assess the performance of
their PPI businesses internally. Our analysis is explained in detail in Appendix 3.1.
3.99 The key findings of our analysis were that:
36
Abbey, summary of hearing, paragraph 2.
Abbey, summary of hearing, paragraph 1.
37
66
• There is some evidence that firms analyse the price of credit and PPI together and
the majority of the distributors say that they do not run PPI as a separate
business.
• There is, however, no primary documentary evidence that firms consider that an
increase in PPI prices will have a significant impact on credit sales.
• When planning PPI price rises, distributors generally do not consider that there
will be a significant effect on their sales of credit.
• However, there is evidence that firms acknowledge that credit and PPI demand
are closely related. 38
3.100 The evidence from distributors’ internal documentation indicated that, generally,
distributors view the demand relationship between PPI and credit to be asymmetric;
they are of the view that changes in their PPI pricing will not affect demand for their
credit products significantly, although we saw some evidence that they do perceive
that changes in credit pricing will affect demand for PPI.
Evidence from the distributors’ sales data
3.101 We assessed the distributors’ sales data for evidence that price changes for PPI in
the past had resulted in an effect on sales of combinations of PPI and credit. We
looked at sales data for personal loans, mortgages and a sample of data from one
large distributor for credit cards.
3.102 In the case of personal loans, we found little evidence that credit sales (and therefore
combinations of credit and PPI) were sensitive to PPI prices. For credit cards, we
were unable to find clear evidence that demand for PPI, or demand for a combination
of PPI and credit, was sensitive to PPI prices. For mortgages, we found some
evidence that demand may be sensitive to PPI prices. However, this result was
sensitive to the precise econometric specification used, and so we could not place
much weight on this finding. A full discussion of our econometric analysis of the
distributors’ sales data is provided in Appendices 3.2 and 3.3.
Switching of the combination of credit and PPI
3.103 In terms of switching the combination of credit and PPI from one provider to another,
we were told by many providers that if switching of PPI occurred, it took place at the
same time that consumers switched or cancelled the underlying credit product.
However, the evidence we have received about why people switch the combination
of credit and PPI has been extremely limited. Lloyds TSB commissioned a survey
that found that if a mortgage was moved to another company then the PPI would
naturally follow, as it was usually linked to the mortgage. It also found that those
consumers who held a mortgage with PPI had slightly less propensity to switch a
fixed-rate mortgage upon maturity than consumers without PPI. [A large distributor of
PPI] records data on switching for personal loans, which showed in the year to June
2007 that 67 per cent of consumers settling early had PPI and 21 per cent of
consumers who settled their loan early repaid using funds from a loan provided by a
competitor.
38
Given that the price of credit is likely to affect sales of credit, it follows that there would be a knock-on effect on sales of PPI.
67
3.104 We looked at data on the PPI termination rates (the level at which consumers cancel
PPI), which has been broken down by the reason for cancellation. Evidence on low
levels of early settlement would tell us that there is not much switching of the creditPPI combination. However, evidence of early settlement does not necessarily provide
insight into the extent that consumers switch the credit-PPI combination as there is
no evidence on what has happened to these customers; they may have switched or
they may have just settled the term early and not taken out credit or the credit-PPI
combination elsewhere. We calculated from data received from the main providers of
PPI that 49 per cent of SMPPI and 33 per cent of PLPPI consumers cancelled PPI
due to the early settlement of the loan (see Table 4.2 for further details). 39
3.105 The evidence we received on switching of the combination of credit and PPI was too
sparse for us to be able to form a clear view on the extent to which consumers switch
the credit-PPI combination.
3.106 We therefore looked at the extent of switching of credit; if we found little switching of
credit we would, by extension, have evidence of little switching of the combination of
PPI and credit. However, the reverse would be less informative as it would not be
clear whether switching of credit was affected by PPI prices. Overall, evidence we
received from the parties suggested that they thought that people with credit and PPI
might be less likely to switch than people with credit alone, but we saw nothing to
suggest that there were significant differences in switching behaviour.
3.107 The evidence we received on mortgages suggested that consumers do switch
mortgage providers, and usually cancel the MPPI associated with the original
mortgage when switching (as noted in paragraph 3.36, there are few portable MPPI
policies). The evidence on personal loans suggested that there was a degree of
switching of the loan, with several surveys provided to us by parties suggesting that
around 9 to 10 per cent of new loans were consumers switching from another loan
provider (and that, upon switching, consumers would cancel their current PLPPI).
3.108 Overall we had very little direct evidence of the level of switching of combinations of
PPI and credit. We saw some evidence that there was switching of credit, particularly
for mortgages. However, this is of limited use because we are unable to determine
the extent to which this switching is affected by PPI prices. We note in paragraphs
3.18 to 3.29 that, in general, PPI demand is unresponsive to changes in PPI prices;
we therefore expect that the level of switching of the combination of credit and PPI on
the basis of PPI prices is low, and the available direct evidence on the level of
switching does not lead us to question that expectation.
Consumer search
3.109 Having considered the extent to which consumers switch credit and PPI at the same
time, we moved on to look more generally at the extent to which consumers compare
combinations of credit and PPI, which they might do before they take out credit or
when deciding to switch providers of credit and PPI.
3.110 In considering consumers’ search behaviour, we assessed three sources of
information. We considered evidence from consumer surveys, evidence from the
providers and also evidence of search behaviour from Internet traffic through price
comparison websites. Our main aim was to understand the extent to which
consumers searched for a combination of credit and PPI together, searched for PPI
39
The same dataset showed that only 6 per cent of MPPI holders settled their mortgage before term, which seemed low to us,
in light of evidence we had received on the extent to which mortgage holders switch mortgage providers.
68
on its own, or searched for credit and then bought PPI from the credit provider
without searching for PPI.
3.111 We found that where consumers undertake comparisons of PPI policies, they are
more likely to undertake a comparison of combinations of PPI and credit. Our survey
provided evidence on where those consumers who told us that they did compare PPI
products conducted their search activity. 40 Those consumers who had compared
combinations of credit or stand-alone PPI products focused their search activity on
Internet comparison websites (the most-cited information source for PLPPI and
SMPPI and the second most-cited source for MPPI), individual providers (either
direct quotes from them, leaflets from them or using their websites) and Internet
search engines (such as Google or Yahoo!). Independent financial advisers (IFAs)
were a less-cited source of information. 41
• Survey evidence
3.112 We looked for evidence that consumers searched on the combination of credit and
PPI. The CC GfK NOP 2008 survey was the only survey we saw that split out search
between those who compared combinations of credit and PPI and those who
compared a PPI product with one or more stand-alone PPI products (see paragraphs
3.52 to 3.59 for the results of that survey for the latter type of search).
3.113 The survey showed that most people who compared PPI products did so by
comparing combinations of PPI and credit (or comparing both those and PPI with
stand-alone products). 15 per cent of MPPI respondents, 9 per cent of PLPPI
respondents and 10 per cent of SMPPI respondents searched on the combination in
some form (either comparing combined PPI and credit only, comparing both
combined PPI and credit and stand-alone policies or making simultaneous credit
applications).
3.114 Those MPPI consumers who did compare products on average compared 3.5
combinations of credit and PPI, taking a median time of 2 to 3 hours to carry out their
searches. PLPPI consumers on average compared rather more combinations of
credit and PPI—4.9—taking the same median time of 2 to 3 hours to carry out their
searches. SMPPI consumers made the most comparisons, on average 6.1
companies, taking a median time of 4 to 5 hours to complete their searches.
• Evidence from the parties
3.115 The evidence from parties was mixed on the extent to which consumers searched on
the combination of credit and PPI or searched on the credit APR and simply took the
PPI sold with the chosen credit. Appendix 3.4 sets out the evidence we received on
this.
3.116 This evidence on search for the combination of credit and PPI suggested to us that if
consumers do search, they are more likely to compare combinations of credit and
PPI than search separately for PPI. This is perhaps not surprising given the current
market structure, with nearly all PPI sold at the point of sale of credit. However, the
evidence did not indicate to us that there were significant levels of such search for
40
PPI Search Behaviour: A research report for the Competition Commission, GfK NOP, April 2008, p18.
MPPI holders who had sourced their policy from an IFA were excluded from this survey, as the purpose of the survey was to
understand the search behaviour of those who had conducted their own search activity. As such, references to using IFAs for
information imply that, whilst the IFA may have provided information, they did not sell the MPPI policy to the customer, and may
not have sold PLPPI and SMPPI to people who consulted them on those products.
41
69
PLPPI and SMPPI. For MPPI, there was a higher degree of search on the
combination. The level of search on MPPI was not such that we felt able to conclude
that, in and of itself, the level of search was either significant or insignificant, without
looking in the round at all the indicators of competition between MPPI providers. We
therefore assessed the sensitivity of PPI demand implied by the survey results (see
Table 3.2). In the case of PLPPI, SMPPI and MPPI we found that the sensitivity of
PPI demand to PPI prices that was implied by these levels of search was low.
• Evidence from the use of price comparison websites
3.117 As noted in paragraph 3.109, consumers who compared different PPI policies cited
Internet comparison websites as one of their main sources of information. We looked
at the search patterns of consumers who used Internet comparison websites, in order
to assess the extent to which these websites were used to search for either standalone PPI products or PPI products in combination with credit. Details of our analysis
of the use of Internet comparison websites can be found in Appendix 3.8.
3.118 Our analysis of data obtained from Internet comparison websites led us to conclude
that some consumers do use Internet comparison websites to search for
combinations of PLPPI with credit; however, the number of people doing this is small
in absolute terms and only a small fraction of all those search for credit. We also
noted that, even with the launch of the new FSA comparison tables, it remains
difficult for consumers to search for bundles of PPI and credit, or to compare the
prices of the PPI component of several different combinations of credit and PPI. We
did not find that the use of Internet comparison websites was significant in terms of
creating competition between PPI providers, and saw nothing to suggest to us that
this would change in the short to medium term.
The relative profitability of credit and PPI
3.119 As noted in paragraph 3.22, in addition to the likely scale of the response of demand
to a change in PPI prices, it is also important to assess how any likely change in
demand would affect distributors’ profits and therefore their incentives to raise prices.
3.120 We assess the profitability of the four main types of PPI (PLPPI, MPPI, SMPPI and
CCPPI) in paragraphs 4.59 to 4.77 and in Appendix 4.4, and the profitability of retail
PPI in paragraphs 6.99 to 6.138. We found that profits are high for the distribution of
all five types of PPI.
3.121 We discuss the profitability of the underlying credit products for each of these types
of PPI in paragraphs 4.80 to 4.94 and Appendix 4.5. We found that the profitability of
personal loans was very low, and in 2006 was loss making, without PPI income. We
found that both mortgages and credit cards were profitable without PPI income, but
that these products were not as profitable as PPI.
3.122 We therefore found a consistent pattern across the three main types of PPI that the
credit element of the combination of credit and PPI is less profitable than the PPI
element. This has the following implications for our market definition.
3.123 First, high margins on PPI imply that there is a relatively low sensitivity of demand for
PPI to PPI prices. This would imply that there would be only limited substitution,
whether to other insurance products, or to other combinations of PPI and credit, in
response to an increase in the price of PPI.
70
3.124 Second, because profit margins on credit are relatively low, any lost credit sales
make a relatively small difference to the profitability of raising PPI prices. We discuss
the importance of relative profit margins in principle in Appendix 3.7 and in relation to
specific analyses we carried out in Appendix 3.2, Appendix 3.3 and Appendix 3.9.
Conclusion on the relationship between credit and PPI
3.125 We set out in Appendix 3.7 that the complementary relationship between PPI and
credit has the potential to constrain distributors’ PPI pricing and so should be taken
into account when assessing the relevant market for a secondary product like PPI.
However, we find that, in the case of PPI, the constraint on PPI prices from the
complementary relationship with credit is modest, and is insufficient to warrant a
wider ‘system’ market for credit and PPI. This is because of two factors:
• First, the effect of PPI prices on demand for a combination of PPI and credit is
low. Our assessment of the evidence from the parties’ internal documents, from
the parties’ sales data and from survey evidence indicates that demand for PPI or
for combinations of PPI and credit is generally unresponsive to changes in PPI
prices.
• Second, the main types of credit on which PPI is sold exhibit only modest
profitability, particularly when compared with the profitability of the PPI sold on
those products. As a result, the incentive for firms to restrain PPI prices, because
of the complementary demand relationship with credit, is lower than it would be
with higher profit margins on credit.
3.126 For PLPPI, we are of the view that the likely effect on demand for combined PPI and
credit of an increase in PLPPI prices would be low. This is because the internal
documents that we saw showed that distributors did not perceive that PPI demand
was responsive to PPI prices, and because our analysis of sales data and survey
evidence showed that, whilst there was evidence of a modest effect, PPI prices were
unlikely to have a significant impact on demand for PPI or credit. In addition, our
analysis of the profitability of PLPPI distribution showed that profit margins on PPI
are high, which is consistent with a low responsiveness of demand to changes in PPI
prices. Finally, the revenue earned on personal loans makes up less than half of the
total revenue that a distributor will earn on a combined PPI and loan, with PPI income
the larger part. This, combined with very low profit margins on personal loans, meant
that the constraint on PPI prices from the complementary link with credit was likely to
be low.
3.127 We also are of the view that the effect on demand for credit card balances of a
change in CCPPI prices would be modest. In our view, the way credit card credit and
CCPPI are priced is very different—one in terms of APR, the other in terms of cost
per £100 of outstanding balance—and makes comparisons of prices of combinations
of credit and PPI difficult and consequently it would seem unlikely that consumers
consider the combined cost of credit and PPI when choosing which credit card to
use. That said, there would seem to be few barriers, in principle, for those consumers
with multiple combinations of PPI and credit cards, to switching combined PPI and
credit balances between their cards in response to a change in relative CCPPI
prices. However, the documentary evidence we received suggested that distributors
perceived the responsiveness of demand to changes in PPI prices was low. This was
supported by evidence from our analysis of the customer data from a large distributor
of CCPPI which showed that the responsiveness of demand to a change in the price
of CCPPI was low. In addition to this, our analysis of the profitability of CCPPI
distribution indicated that this was highly profitable and that there were high profit
margins on CCPPI. This is consistent with a low responsiveness of demand for
71
combined PPI and credit balances to changes in PPI prices. We therefore concluded
that there was unlikely to be a strong relationship between CCPPI prices and
demand for credit card balances and that consequently the constraint on CCPPI
prices from that demand relationship was likely to be low.
3.128 Our analysis of second-charge mortgages and SMPPI showed that there was
evidence of some search taking place; however, the levels of search were low and
our analysis in Appendix 3.9 showed that this was only marginally sufficient to
constrain prices at current levels. The documentary evidence we saw showed that
distributors do not behave as if they believe there will be a large effect on demand for
secured credit of a change in PPI prices—distributors raise prices without taking
account of the possible effect on credit sales. In addition to this, our analysis of the
profitability of SMPPI distribution indicated that this was highly profitable and that
there were high profit margins on SMPPI. This is consistent with a low responsiveness of demand for combined PPI and credit balances to changes in PPI prices. We
therefore considered that the constraint on SMPPI prices from the complementary
demand relationship with credit would be low.
3.129 Our analysis of mortgages and MPPI showed that there was some evidence of a
demand relationship between MPPI prices and demand for mortgages. The CC GfK
NOP 2008 survey showed that there was some search for combinations of MPPI and
mortgages. The incidence of this search behaviour was higher than for other types of
PPI for which survey evidence was available. However, the level of substitution
implied by these results still implied a relatively low responsiveness of demand for
combinations of PPI and mortgages to changes in PPI prices. Our analysis in
Appendix 3.9 showed that this would either be insufficient or be only marginally
sufficient to constrain prices at current levels. In addition we would not necessarily
expect all of the consumers identified as comparers in the survey to change their
purchase in response to a small increase in the price of PPI. 42 Our analysis of the
profitability of MPPI distribution indicated that this was highly profitable and that there
were high profit margins on MPPI. This is consistent with a low responsiveness of
demand for combined PPI and credit balances to changes in PPI prices. We
therefore concluded that, whilst the constraint on MPPI prices from the complementary demand relationship with mortgages may be higher than for other types of PPI,
this was still insufficient to define a wider systems market for combined PPI and
mortgages.
The position of non-standard credit consumers
3.130 One provider (Cattles), in explaining why it thought that competition took place on the
combination of credit and PPI, said that our market definition was particularly
inappropriate for the non-standard sector. 43 It gave three main reasons: 44
• PPI was more important to non-standard consumers, something it said was
reflected in the fact that the penetration rate of PPI for its customers was
significantly higher than the average PLPPI penetration rate. We were told that the
implication of the evidence on this point was that the price of PPI would be
constrained by its influence on the overall demand for the credit and PPI together.
42
Appendix 3.9 paragraphs 34–37. There is no standard definition of a non-standard customer. Datamonitor defines a non-standard customer as someone who is systematically refused credit from mainstream lenders (banks, building societies and large finance houses), whatever the size and nature of their application. (Datamonitor: UK non-standard and sub-prime consumer credit 2006). Sub-prime customers are a subset of non-standard customers. 44
Submission from Cattles plc—Market Definition (response to Emerging Thinking). 43
72
• Many of its customers had prior experience of credit and PPI, which it thought was
not consistent with a view of PPI distributors having a captive market of
consumers who did not understand the benefits on offer.
• Non-standard consumers are more likely to need to manage their finances
carefully, and are more likely to be sensitive to the level of monthly payments,
including the cost of PPI. As a result consumers are more sensitive to monthly
cost than the credit APR or the price of the insurance.
3.131 Cattles said that these reasons meant that it was more likely that non-standard
consumers would consider the price of the combination of PPI and credit and so, for
non-standard credit providers, increases in PPI prices were more likely (in principle)
to result in a loss of PPI and credit sales.
3.132 We looked first at the evidence on PPI being more important to non-standard
consumers. Cattles provided evidence that the penetration rate among non-standard
consumers was significantly higher than average penetrations shown in CC survey
data, though it had declined by ten percentage points [] between 2002 and 2006. 45
Cattles said that demand for credit would be influenced by the price of PPI; some
consumers might be unwilling to take out credit without the protection provided by
PPI.
3.133 However, whilst non-standard consumers may display a greater propensity to take
out PPI, there is no obvious reason why those consumers who do take out PPI
should be any more or less responsive to changes in PPI prices than standard credit
consumers who take out PPI. The evidence we received suggested that nonstandard PPI consumers are not responsive to changes in PPI prices (see paragraph
3.23), and we have seen nothing to suggest to us that a significantly higher number
of non-standard consumers would react to an increase in PPI prices than would
standard consumers. As noted in paragraph 3.135, non-standard consumers showed
a lower propensity to search than standard consumers, which would seem to indicate
that the reverse is true and a lower proportion of non-standard consumers would
substitute in response to an increase in PPI prices. This was reflected in the
experience of another large distributor, []. This distributor found that when it raised
its CCPPI charge by 29 per cent, then a further 11 per cent, for sub-prime customers,
there was no decline in demand (see paragraph 3.23). The evidence we saw did not
convince us that, for non-standard consumers, the price of PPI would be constrained
by its influence on the overall demand for the credit and PPI together.
3.134 We considered next the relevance of consumers’ experience. We were told by
Cattles that it experienced noticeably higher penetration rates for PPI among
customers who had previously had a Cattles loan, and had taken out PPI before (with
Cattles or someone else) than among customers who had not. However, we did not
find this evidence compelling. Unless consumers make a claim on their policy, it is
difficult to see why a repeat customer should exhibit a higher propensity to shop
around on PPI than a customer with no experience of PPI. Consumers who claimed
successfully on PPI policies may have a better understanding of the benefits of a
policy as a result of the claims experience and so may be more likely to shop around
on PPI; however, as pay-out rates on PLPPI polices are less than 15 per cent, which
implies that most PLPPI policies are not claimed on, we did not find that the
incidence of repeat customers is sufficient evidence to support the contention that
non-standard consumers would show a greater propensity to shop around on PPI.
45
Submission from Cattles plc—Market definition, paragraph 42 (response to Emerging Thinking).
73
3.135 Finally, we looked at the importance of monthly payments to non-standard
consumers. Evidence from the CC GfK 2008 survey showed that, for both PLPPI and
SMPPI, a significantly lower proportion of the non-standard consumers stated that
they compared combinations of credit and PPI than did standard consumers. 46 This
suggested to us that, whilst it may be true that non-standard consumers need to
manage their finances particularly carefully, this did not result in their comparing the
costs of PPI policies more carefully than prime consumers—indeed, the opposite
proved to be the case. Cattles provided us with a presentation from research in 2006
among people who had been offered a Cattles loan but did not subsequently take it
out, which indicated that a large proportion of these people had considered the
overall cost of credit and PPI as a package. We did not find the analysis of the survey
data convincing evidence that non-standard consumers were more likely than others
to consider the overall cost of a credit and PPI package when choosing a credit
provider.
3.136 The evidence we saw suggested to us that, whilst a higher proportion of nonstandard consumers take out PPI, non-standard consumers do not exhibit
significantly different behaviours from standard consumers. We conclude therefore
that our market definition, of PPI sold by a particular distributor, is appropriate for
non-standard consumers.
Geographic market
3.137 All the distributors of PPI that we received evidence from considered the relevant
geographic markets to be the UK. We have received no evidence that would indicate
that there are any regional differences in prices or policy offerings which may suggest
narrower geographic markets, nor has any distributor indicated to us that they are
planning to implement such differences in the future.
3.138 We do not consider that defining regional geographic markets would alter our
analysis to a significant extent. We therefore, consider that it is appropriate as a
framework for the analysis of competitive constraints, to define the relevant
geographic market as the supply of PPI by distributors and intermediaries to their
customers within the UK. For providers that do not sell across the whole of the UK,
the relevant market is the area within the UK within which they do sell PPI.
Conclusion on market definition
3.139 We conclude that, for all types of PPI policies, the relevant product market is the sale
of PPI to an individual distributor’s, or intermediary’s, credit customers by that
distributor or intermediary. The reason for this is that we found that the competitive
constraints being imposed on distributors and intermediaries by other providers of
PPI, by providers of short-term IP, by providers of other types of insurance products
and by consumers choosing not to purchase PPI were not sufficiently strong to
warrant a wider product market. Further, we found that the constraint on PPI prices
as a result of the complementary demand relationship with credit was not sufficiently
strong to warrant a wider systems market including credit and PPI.
3.140 For stand-alone PPI and stand alone short-term IP providers, we concluded that they
competed to win customers from across the range of PPI providers, both those who
46
For PLPPI: only 6.9 per cent of 303 non-standard customers compared on the combination, compared with 15.2 per cent of
2,838 standard customers, a statistically significant difference at the 95 per cent confidence level. For SMPPI: only 11.4 per
cent of 1,236 non-standard customers compared on the combination, compared with 17.7 per cent of 1,135 standard
customers, also a statistically significant difference at the 95 per cent confidence level.
74
offer PPI in combination with credit and those who offer it on a stand-alone basis. As
such, the relevant markets are asymmetric. We concluded that stand-alone providers
are constrained by competition with distributors, but the scale of substitution from
distributors to stand-alone providers is insufficient to competitively constrain the
distributors. Stand-alone providers therefore operate in a wider economic market
including all providers of PPI.
3.141 We conclude that the relevant geographic market for each product market is the
geographic area within the UK in which the distributor is active.
The impact of our market definition on our analysis of competition
3.142 It was put to us that, even if our analysis of market definition found that there were
narrow retail markets, it would be more appropriate to identify wider markets for the
purposes of conducting our analysis of competition between PPI providers. We were
told 47 that:
Brand-specific market definition does not provide a useful platform for
assessing how a market functions or the ‘extent of competition between
suppliers’, which the CC wishes to examine after reaching a view on the
relevant market(s) for the sale of PPI. For example, the number of
competitors in a market defined as ‘Lloyds TSB-branded PLPPI’ will be
limited to one and Lloyds TSB is unavoidably a monopolist in this market;
by definition there will be no ‘intra-market rivalry’ or ‘competition between
suppliers’ in the relevant product markets provisionally defined by the
CC.
3.143 It was suggested that the more appropriate way forward was to look at whether there
were impediments to competition between PPI providers by looking at a productspecific market (for example, all PLPPI).
3.144 As set out in paragraph 3.2 and in our guidelines, 48 market definition and the
assessment of competition between providers are linked and will overlap
significantly. As such, the definition of the relevant market is a reflection of our
findings in Section 4, regarding the level of competition between PPI providers, as
well as a framework within which to assess that evidence. Given that the bulk of the
evidence in this report points to limited competition between distributors, we consider
it sensible to reflect that evidence in our market definition. We do not regard a wider
market definition, which implies powerful competitive constraints between the
providers in that market (sufficient to render a small price rise above competitive
levels unprofitable), as defensible in the light of all of the evidence we have seen.
3.145 For example, to conclude that the relevant market was wider to encompass types of
PPI (PLPPI, MPPI, CCPPI etc) we would need to see evidence that the policies sold
by distributors at the point of sale are substitutable. Given the very low sales of
stand-alone PPI, the limited availability of stand-alone policies and the lack of
evidence of substitution to stand-alone policies (see paragraphs 3.38 to 3.43), we do
not regard this market definition as sustainable. Similarly, in order to define a wider
systems market for PPI and credit (combined PPI and loans, combined PPI and
mortgages and combinations of PPI and credit card balances), we would need to see
evidence that there was substitution between combinations of PPI and credit, and
47
Lloyds TSB response to Emerging Thinking, paragraph 2.10(i).
CC3, paragraph 1.22.
48
75
furthermore that PPI prices influence substitution of this type. For the reasons set out
in paragraphs 3.90 to 3.108, we do not find that this is the case.
3.146 We do not agree that the market definition we have found would not provide a useful
tool for looking at competition in the markets. 49 In reaching our conclusions on
market definition, we considered the competitive constraint imposed by stand-alone
provision of PPI and alternative products to PPI. Further, we looked at whether the
competitive constraint on PPI distributors from the complementary relationship with
credit was sufficiently strong to warrant a wider system market for credit and PPI.
3.147 Whilst we have found a narrow market definition, we do not consider it necessary to
examine all the markets individually, and reach views on the state of competition
within each one. We considered, where appropriate, whether markets were in some
way different from each other, because of, for example, the consumers they served
(we looked at whether non-standard consumers were different from prime consumers
in paragraphs 3.130 to 3.136) or the reasons consumers used the seller (see our
analysis of intermediaries in paragraphs 3.64 to 3.74). However, we saw no
differences which led us to come to the view that different economic markets for a
given PPI policy operated differently as a result of the behaviour of the consumer.
3.148 Similarly, we saw nothing to suggest that providers’ strategies were different in ways
which might lead to different competitive conditions in different economic markets.
For each type of PPI, we saw similar types of PPI policies being sold under similar
conditions in similar ways. We were content, therefore, that by identifying the
indicators of competition for each type of PPI policy we would be looking at the
indicators of competition for each of the economic markets for each distributor or
intermediary; similarly, by looking at the factors affecting competition for each type of
product, we would be looking at the factors affecting competition in each of the
economic markets for that product.
3.149 As noted in paragraph 3.143, one alternative approach to market definition that could
be used would be to define a market ‘no wider’ than the sale of a particular type of
PPI, and then assess the level of competition between the distributors active in that
frame of reference. This approach has been used in other CC inquiries. However,
this approach would, on the same evidence, lead us to the same conclusions
regarding the level of competition between providers as the approach we have taken.
Irrespective of whether the analysis is looking to see whether competition is effective
in a product market no wider than the supply of PPI, or whether there are external
constraints meaning that competitors in a distributor-specific PPI market cannot raise
prices above competitive levels, we are looking at the same companies selling the
same products, looking at the same indicators of competition and using the same
economic tools. We do not, therefore, find that the choice of market definition used to
frame the analysis of unilateral market power—among those definitions we consider
are defensible on the facts 50 —would have led to a different result.
3.150 Abbey told us that our market definition could potentially have significant
consequences for future investigations, including cartel cases. It said that our market
definition implied that a price-fixing agreement between PPI providers might be
49
HBOS said that the CC’s market definition had important implications regardless of the assessment of competition that
followed and that it would set a dangerous precedent to conclude a narrow market definition in the market for PPI when the
evidence was in fact uncertain. For the reasons set out in paragraph 3.149 we do not agree that defining a narrow market for
PPI, instead of a ‘no-wider’ market including all PPI policies, would alter our conclusions. Moreover we do not agree that our
market definition in this case would set a precedent for other cases, since the facts of those cases will be different. Finally, we
do not agree that the evidence is uncertain, nor do we agree that the ‘certainty’ of the evidence should constitute reason to
conclude a wider or a narrower market.
50
See paragraph 3.145.
76
unproblematic as the PPI providers did not compete with each other. In our view, our
market definition suggests that, even if PPI providers were inclined to such
behaviour, they would be unlikely to form a price-fixing cartel as it would not be in
their interests to do so. We received no evidence of the existence of such a cartel for
PPI.
77
4. Indicators of the extent of competition between PPI providers in the
supply of PPI
Summary
4.1 We found in Section 3 that there was very little evidence that PPI distributors
imposed a competitive constraint upon each other. In this section we assess a series
of indicators of the level of competition between PPI distributors.
4.2 We found that:
• There is little variation in PPI prices over time, and there is little evidence of PPI
distributors seeking to win sales from each other by competing on PPI prices.
• There is no clear evidence of PPI distributors or intermediaries seeking to win
sales from each other by competing on non-price factors such as quality
innovation or choice. We consider that it is unlikely that there is vigorous
competition on non-price factors because these factors are particularly difficult for
consumers to observe.
• There is very limited advertising of PPI. Advertising spending is on the credit
product and where PPI is included in advertising literature, the focus of advertising
is predominantly on the underlying credit product.
• PPI prices and the prices of combinations of PPI and credit are dispersed even
when differences in quality are taken into account. We find that the evidence
shows that there is no clear relationship between PPI price and quality in PLPPI
and CCPPI and only a relatively weak relationship between price and the quality
of MPPI.
• Despite this dispersion in prices and quality, and the apparent benefits to
searching and switching that this would seem to confer, the incidence of
substitution between PPI policies, or bundles of PPI and credit is low.
• The claims ratios on PPI policies are low, reflecting a high mark-up of price over
claims cost. This suggests that prices are high, which we considered by looking at
the profitability of the value chain in further detail.
• Costs of distributing PPI are very low relative to the price of PPI resulting in high
profits. We consider the profitability of PPI underwriting in Section 7.
4.3 These findings indicated to us that prices, and other competitive variables, are less
favourable to consumers than would be the case in a well-functioning market with
vigorous and effective rivalry between suppliers.
4.4 We consider that these findings are consistent with our finding in Section 3 that there
is limited evidence of substitutability between PPI policies and that PPI distributors
and intermediaries are not competitively constrained by the potential for consumers
to substitute between their products. These findings are also consistent with our
conclusion that there are a series of individual markets where credit suppliers have
market power over the supply of PPI to their own credit customers.
78
Introduction
4.5 As noted in paragraphs 3.139 to 3.141, we found narrowly-defined product markets.
However, it was put to us that there is competition between distributors of PPI, so we
look here at indicators of the extent, if any, of competition between PPI providers.
This can provide us with a check on whether our market definition is accurate, or
whether competition takes place more widely (which would be at a PPI product-wide
level). More generally, looking at the indicators of competition allows us to consider
the extent of competition between PPI providers.
Variation of PPI prices over time
4.6 We looked at how PPI prices had changed over time, to see if this showed evidence
of competition between providers.
4.7 We looked at price changes 1 of all PPI policies for which we obtained usable data
from the 12 largest distributors. The results of this analysis are shown in Table 4.1.
TABLE 4.1 Analysis of PPI price changes, 2002 to 2006
PLPPI
1-year premium
5-year premium
Products with usable
data
Products with no price
change
Products with any price
change
Of which:
Products with some
negative price change
Number of price
changes
Total
Up
Down
Growth in prices
Number of products
Total growth (median)
Total growth (mean)
MPPI
CCPPI
SMPPI
79
84
26
27
18
28 (35%)
27 (32%)
15 (58%)
18 (67%)
18 (100%)
51 (65%)
57 (68%)
11 (42%)
9 (33%)
0
12 (15%)
12 (15%)
2 (8%)
1 (4%)
0
95 (1.22 per
product)
82 (86%)
13 (14%)
121 (1.44 per
product)
108 (89%)
13 (11%)
17 (0.65 per
product)
15 (88%)
2 (12%)
27 (1 per
product)
26 (96%)
1 (4%)
0
35
11.8%
14.5%
37
11.5%
16.3%
20
0%
9.6%
23
3.9%
4.7%
0
0
Source: CC, based on data provided by the 12 largest distributors of PPI.
4.8 This analysis showed that over half of MPPI and CCPPI policies saw no price change over the period in question, and only one CCPPI policy and two MPPI policies saw a price decrease during the period. There were no price changes in any of the 18 SMPPI policies we looked at. For PLPPI there was more price variation, with between 10 and 15 per cent of all price changes being downwards. However, the evidence suggested to us that for all types of PPI, the variation in prices over five years was very limited. 2
1
We note that prices are not affected by inflation, because the premium is calculated as a proportion of the credit product.
One distributor (Abbey) told us that subsequent to this period it had reduced the price of its CCPPI from 79p per £100 of
outstanding balance to 59p per £100 of outstanding balance with the objective of increasing take-up and to respond to the
growing competitive threat. We note that a price change is not sufficient evidence, in itself, to demonstrate that there is effective
competition, nor is it sufficient to demonstrate that there is significant customer substitution between PPI policies. This
additional data did not change our analysis.
2
79
4.9 In addition to this general trend for PPI prices to remain stable or rise, we noted that
price changes were infrequent.
4.10 We considered reasons why prices had not varied much. The parties told us that
while the price of PPI might not have varied a great deal during the period 2002 to
2006, the level and quality of the cover did, such that consumers were getting better
value for money (see Appendix 4.1 for details of the views of the parties). We noted,
though, that terms and conditions were not advertised to any significant degree;
parties did not appear to try and win business by promoting the terms and conditions
of their policies. Moreover, we noted that there had been increasing numbers of
complaints to the Financial Ombudsman Service, 3 increasing amounts of adverse
media coverage, and increasing regulatory scrutiny and intervention in PPI by the
FSA. We considered that the need to treat customers fairly, and try to address
negative views on their record on this point, were more likely to be the primary
drivers for many of the policy changes we saw, rather than trying to provide better
value for money to gain a competitive edge. The indicators of non-price competition
are discussed more fully in paragraphs 4.12 to 4.23.
4.11 In a competitive market we would normally expect to see price variations over time
as firms seek to win business from each other. The level of price variation over time
that we saw was consistent with there being few significant competitive pressures on
PPI providers.
Indicators of non-price competition
4.12 In order to make an assessment as to the extent to which PPI distributors engaged in
non-price competition, we considered evidence supplied to us to identify: any new
PPI policies which had been introduced, whether there had been any innovations
within existing policies, the rationale for product change or innovation, and whether,
and if so how, distributors advertised and marketed their PPI policies.
New PPI policies
4.13 There was some evidence that firms had offered new policies in recent years, for
example: Lloyds TSB piloted PPI for a homeowner loan; Northern Rock introduced
SMPPI in June 2004; and HBOS launched a Total Mortgage Protection policy, which
incorporated MPPI, in 2001, overdraft PPI in June 2003, and stand-alone policies
known as Life Stages and Life Style in 2006, which offered accident, sickness and
unemployment cover.
4.14 Also, as noted in paragraph 3.38, three new policies have been introduced during the
course of this investigation: HSBC launched a short-term IP policy, LifeChoices, in
May 2007; 4 RBSG launched a short-term IP policy through its Churchill brand in May
2007; 5 and in September 2007 Barclays launched its ‘Plan B’ CCPPI policy, to cover
Barclaycard and/or other credit cards.
3
See Appendix 2.5. See Appendix 2.3 for further details. 5
See Appendix 2.3 for further details. 4
80
Innovations within existing products
4.15 Innovations within existing PPI policies included changes in the level of cover and
other benefits, the introduction of tiered PPI policies, changes in pricing structure and
the introduction of additional services.
4.16 We considered innovation in the supply of PPI to have been relatively rare. However,
some examples of policy changes presented to us were: 6
• Lloyds TSB increased the benefits of its PLPPI policies to include payment
holidays and amended the terms and conditions of the life and critical illness
cover.
• In 2000, Barclays introduced a CCPPI policy which offered a half-price half-benefit
step-down option.
• In 2003, Capital One introduced variants within its CCPPI offering—‘intro’, which
was a standard policy (but with a three-month payment-free period when taken
out); ‘joint’, which was designed to cover a second card holder; and ‘classic’,
which was a reduced cover policy.
• Northern Rock tiered its PLPPI policy in June 2005 to offer Bronze (life) and Silver
(life, sickness and accident) with its existing Gold offer; and in 2006 added
another Silver component (unemployment and life) to the policy.
• Some companies have moved to age-based pricing (see footnote to paragraph
2.33).
4.17 We also saw evidence that some distributors had offered additional services to their
customers, such as back-to-work job services initiatives. 7
Reasons for introducing new and amended PPI policies
4.18 We reviewed the internal documents provided to us by parties for evidence of
whether distributors were monitoring, and reacting to changes to, their competitors’
PPI policies. We found the following examples:
• [] researched the viability of offering customers three levels of cover from
Bronze to Gold for its PLPPI offer, with the aim of increasing sales. The
documents noted that the policies would be similar to those offered by Northern
Rock, HBOS and Sainsbury’s Bank.
• Barclays reduced the exclusion period on pre-existing conditions from 24 months
(five years for chronic, back and psychological conditions) to 12 months in order to
align itself with other large banks. (However, we noted that this was done after
Citizens Advice had publicly highlighted this discrepancy in Barclays’ policy.)
• HBOS had undertaken a benchmarking analysis with regard to its mortgage sales,
and the policies cross-sold with mortgages.
4.19 The internal documents we reviewed identified only a few examples of competition
between PPI providers taking place on product innovation. Many decisions to amend
6
Other examples of changes in PPI terms and conditions were provided to us. See, for example, HSBC hearing, 17 July 2007, paragraph 31; Lloyds TSB hearing, 26 July 2007, paragraph 33. 7
81
product designs were based on the results of consumer surveys and regulatory
requirements, rather than competitive pressure from other distributors. 8
Advertising/marketing
4.20 We looked at evidence of how PPI distributors advertised and marketed their
policies. We found few examples of distributors actively promoting their PPI or of
advertising campaigns specifically featuring PPI policies. PPI-specific promotions
appeared to have been limited to more passive forms of advertising such as inbranch leaflets, information on websites or inserts in customer mailings. For example:
• Alliance & Leicester had advertised MPPI in brochures in branches, though the
brochures did not include policy pricing. 9
• Lloyds TSB did not generally include MPPI in its advertising, though basic PPI
information was provided in promotional materials for other credit products. Also,
in June 2007 a Lloyds TSB customer leaflet referred to key differences between
its policies and those of its main competitors, and included a competitor
comparison table.
• Northern Rock had not included PPI prices or features in its advertising or
marketing campaigns in recent years. However, it had run promotional discounts
on its PPI policies across its distribution channels from time to time.
• Nationwide had sent out a customer mailing in which it had compared personal
loan repayments, with and without PPI, with Barclays, NatWest, RBS and HSBC.
It also conducted an MPPI direct mailing campaign, highlighted the costs of loans
with and without PPI in advertising, and provided information at the point of sale.
• Abbey had promoted its PPI policies via posters and direct mail. Abbey had also
included in the marketing material for its MPPI policy a table comparing it with the
MPPI policies of HBOS and RBSG.
• RBSG had promoted its stand-alone Churchill PPI policy.
• RBSG also told us that in January 2007 Tesco Personal Finance (TPF) had
implemented an advertising campaign (on television, in-store, via email and mail
inserts) which directly compared its PPI offering with those of HBOS and Lloyds
TSB, and claimed that TPF’s loan with PPI was cheaper.
4.21 Distributors appeared to rely on advertising of credit in order to sell both credit and
PPI. We saw little evidence that credit advertising advertised PPI at the same time.
For example, despite large numbers of consumers buying SMPPI and PLPPI at the
same time as taking out a loan, firms advertised the APR of the credit (as they have
to do), but rarely showed, in addition, the price of the PPI or the price of the bundle in
advertising.
4.22 Further, when PPI pricing is provided to consumers by distributors and
intermediaries, it is done in a way which does not facilitate a clear understanding of
8
Barclays said that the Defaqto rating system constituted a further competitive constraint. This was because to maintain a five
star rating a policy had to be in the top 10 per cent of policies, so providers would need to constantly innovate in order for their
policies to maintain a five-star rating (Barclays’ response to provisional findings, p11). We note that Defaqto changed its
approach to rating on the 1 February 2008. After that date, to achieve a five star rating, policies had to achieve an absolute
score rather than simply being in the top 10 per cent of scores.
9
Alliance & Leicester has since begun advertising its stand-alone MPPI product.
82
the relative cost of the product. When PPI cost is shown to a consumer, it is shown
as a monthly cost, and shown alongside the monthly repayment of the loan. Hence
consumers see the monthly cost alongside the cost of both capital repayment and
interest on the loan, rather than seeing it compared with the monthly interest
payment. If they saw the latter comparison consumers might gain a clearer
understanding of the relative cost of PPI.
4.23 Due to the limited introduction of new types of PPI policies, the lack of advertising of
PPI and the lack of evidence that distributors react to their competitors with regard to
PPI, we formed the view that distributors do not compete to any significant degree on
the non-price aspects of PPI.
Price dispersion
4.24 We found in paragraph 4.8 that prices of individual PPI policies did not vary much
over time. We looked at the relationship between price and policy characteristics
across PPI policies, to see whether any price dispersion could be accounted for by
different terms and conditions of policies (in effect, whether policy price reflected
quality of product).
4.25 In a competitive market where consumers searched effectively, it might be expected
that variations in price of differentiated products could be explained by variations in
non-price features of the products. This is because in a well functioning market we
would expect that differences in quality-adjusted prices would be competed away
over time as consumers substitute away from policies that have a high price relative
to their quality, to lower price and/or higher quality policies. Details of our analysis
can be found in Appendix 4.2.
4.26 As discussed in paragraphs 5.12 and 5.13, we found price variation between PLPPI
policies, for both PLPPI on a stand-alone basis and the credit-PLPPI bundle. There is
price dispersion for MPPI, though it appeared to be less pronounced than that of
PLPPI. 10 We observed minimal price variation between CCPPI policies but observe
significant variation in benefit calculations between policies, suggesting that
consumers may still benefit from search.
4.27 In addition, we considered whether there was a correlation between PPI price and
product quality that could explain some of the dispersion in PPI prices. We used the
Defaqto star rating scores, and star ratings, created using the system used by
Defaqto up to the end of January 2008, as a measure of quality. Defaqto gathers
information on many different characteristics of PPI policies and uses a benchmarking system to score policies for quality across a wide range of features and
benefits. It then assigns a star rating, from one to five stars. A number of parties told
us that they thought that Defaqto ratings were a relatively poor measure of PPI policy
quality. Whilst we recognized that Defaqto’s rating system was not perfect (and that
there is no perfectly objective method by which we could measure policy quality), we
considered it appropriate to use it in our analysis because we note that it is common
practice for some distributors to use their Defaqto star ratings as a signal of product
10
Stand-alone providers were not contained within the dataset.
83
quality in their marketing materials. 11 We also noted that Defaqto’s star rating system
is the only available measure of product quality. 12
4.28 For PLPPI, the main way we measured PLPPI price was cost per £100 of monthly
repayment. We found that the number of Defaqto stars awarded to each PLPPI
policy did not appear to be correlated with the price of the product. Similarly we found
no correlation between the Defaqto score and price.
4.29 For CCPPI we looked at two ways of measuring price: the monthly premium per £100
of outstanding balance; and, because different CCPPI policies pay a different
percentage of the balance outstanding on the credit card in the event of a claim, the
cost per £100 of benefit. By considering the relationship between the monthly
premium per £100 of outstanding balance and the level of cover per £100 of
outstanding balance, we found that premiums are similar even though the level of
cover and payout rate varies considerably.
4.30 For MPPI the price measure used was the cost per £100 of cover. For MPPI we
found a positive relationship between the Defaqto star rating for an individual MPPI
policy and its price, and similarly between Defaqto score and price. However, there
was a large amount of variation in prices of different products around this relationship. For all of the three types of PPI for which data was available we found a weak
relationship between policy quality, as measured by Defaqto, and the prices charged
for those policies. This dispersion in prices could have been caused by any of several
factors, including:
(a) the selection of characteristics used to generate the Defaqto star rating might not
be a good measure of the value consumers put on characteristics. This could be
because of one or more of the following factors:
(i) the equal weighting that Defaqto uses for each criterion does not reflect the
relative value consumers place on each characteristic;
(ii) other characteristics that are not used in the Defaqto star rating mask the
correlation between price and the Defaqto star rating;
(iii) consumers value other non-PPI specific variables such as the provider’s
brand or customer service; and
(iv) the risk profile of the underlying loan’s customers, and the customers who
take out the PPI policy, could affect the correlation; or
(b) there is a weak relationship between the price and the quality of the PPI policies.
4.31 The evidence suggested to us that PPI prices showed a level of dispersion across
products, that there was no clear correlation between PPI price and product quality
for PLPPI and CCPPI, and that there was a positive relationship between MPPI price
and product quality, but with a large amount of variation. The price dispersion we
found was in line with evidence we saw from Defaqto, which found a large degree of
11
For example, on 11 May 2008 providers such as Barclays, Lloyds TSB and NatWest advertised on their websites that their
PLPPI products had a five-star Defaqto rating. RBSG told us that the Defaqto seal of approval was used in its scripts and
marketing literature, showing how it competed on quality as well as price (RBSG response to Emerging Thinking, paragraph 3).
12
This reflects in part the difficulty in establishing an objective measure of product quality. For example, Defaqto records 24
separate characteristics which it considers affect the scope and quality of cover on a PLPPI product. Similarly, there were 24
separate characteristics each for MPPI CCPPI.
84
price variation for PPI policies. 13 This is consistent with findings from Defaqto, the
OFT 14 and Which?, all of which have found that variations in the price of PPI were
not always explained by differences in cover.
Search
4.32 The extent to which consumers search for PPI can be an indicator of competition,
especially in the context of markets where there are information asymmetries, there
are significant switching costs and there is little evidence of consumers switching. 15
We considered the extent to which consumers compared combinations of PPI and
credit with stand-alone policies We concluded in paragraph 3.61 that the extent of
search through comparing a PPI policy with one or more stand-alone policies was
low, only seeing more than 10 per cent of consumers searching in this way for MPPI.
We also considered the extent to which consumers compare bundles of credit and
PPI. We concluded in paragraph 3.116 that the evidence did not indicate that there
were significant levels of such search for PLPPI and SMPPI. For MPPI, there was a
higher degree of search on the bundle. The level of search on MPPI was not such
that, in and of itself, we felt able to conclude that it was either significant or
insignificant, without looking in the round at all the indicators of competition between
MPPI providers. We assessed the sensitivity of PPI demand implied by the survey
results (see Table 3.2). In the case of PLPPI, SMPPI and MPPI we found that the
sensitivity of PPI demand to PPI prices that was implied by these levels of search
was low. We consider all the evidence on MPPI together at paragraph 4.95, where
we noted that most search takes place on combinations of MPPI with mortgages and
that Internet comparison websites are not configured to allow consumers to search
for such combinations.
4.33 We consider that this lack of search is an indicator of a lack of competition.
Switching
4.34 Evidence of consumers switching between PPI policies may suggest a constraint on
PPI prices; this applies both to the case when consumers switch PPI on a standalone basis, and to the case when credit and PPI are switched together as a bundle.
However, in the latter case, in order to conclude that switching of the PPI and credit
bundle acts as a constraint on PPI prices, we would need evidence that a sufficient
number of consumers were considering the bundled price (ie they are considering
the PPI component as well as the credit price) when deciding whether to switch. If
consumers primarily base their choice of credit-PPI combination on the credit price
(the APR of the loan), then switching is less likely to constrain PPI prices.
4.35 The level of switching does not necessarily provide unambiguous information
regarding the level of competition prevalent in a market. It is possible that there may
be very little switching in a well-functioning market. For example, where firms’ prices
adjust very rapidly in response to price changes implemented by their rivals, there
will rarely be an incentive for consumers to switch, since prices are constantly
adjusted so as to remain competitive. Evidence regarding switching therefore needs
13
Payment Protection Insurance 2008: The Party’s Over Defaqto, February 2008, Chart 12. Defaqto found that the cost per £100 of monthly benefit varied between £2.15 and £6.50 for MPPI, £5.00 and £28.33 for CCPPI, and £10.62 and £24.40 for PLPPI. 14
The OFT mystery shopping exercise found that it was difficult to compare products due to differences in cover but still noted dispersion. OFT, Summary report of mystery shopping exercise, October 2006, Annex E. 15
CC3, paragraphs 3.47&3.51. 85
to be considered in the context of the observed dispersion in prices (paragraphs 4.24
to 4.31).
4.36 Evidence on switching of CCPPI has not been volunteered and is not apparent to us.
This is because, unlike PLPPI, SMPPI and MPPI policies, it is not necessary to
cancel either the credit or PPI in order to take out a new CCPPI policy. A consumer
pays no premium when there is no balance on a credit card, so there is less incentive
to cancel CCPPI when transferring a credit balance to a new card. [] provided us
with an analysis of its customer level sales data. [] told us that this analysis
showed that demand for insured credit cards and insured credit card balances was
responsive to PPI prices, which could be indirect evidence of switching. However, we
had some concerns regarding the analysis undertaken by []. Our own examination
of the same data did not show that demand for insured balances was affected by
credit prices. A full discussion of our analysis of [] credit card data is provided in
Appendices 3.2 and 3.3.
4.37 We looked at the extent of switching of the bundle of credit and PPI in paragraphs
3.103 to 3.108. Here we focus on the extent of switching of PPI on its own. Details of
our analysis of switching can be found in Appendix 4.3.
Switching PPI without switching credit product
4.38 We looked at the extent to which consumers switch just the PPI policy, but keep the
loan with their existing provider.
4.39 The rate at which PPI policies are terminated whilst the credit is maintained would
give an upper bound to the level of switching of PPI without switching credit (it would
provide an overestimate of switchers because the figures would include people who
cancelled PPI without taking out a replacement policy). We noted that for CCPPI it
was possible either to terminate the PPI policy or to stop using the credit card—so
termination data would show an upper bound for CCPPI switching subject to this
behaviour. Table 4.2 shows our analysis of PPI terminations, based on data obtained
from the largest providers of PPI.
TABLE 4.2 Breakdown of terminations for PPI (2002 to 2006) as a percentage of policies at start of the year
per cent
PLPPI
MPPI
SMPPI
CCPPI*
Breakdown of termination
Cooling-off period
Settled loan to term
Settled loan early
Retained loan but stopped using PPI
Defaulted on loan
Other
1
11
33
1
3
3
1
1
6
9
0
8
1
1
49
1
0
2
1
9
6
8
Potential switchers—scenarios
Retained loan but stopped PPI only
+ cooling off
+ cooling off + other
1
2
5
9
10
18
1
2
3
6
7
15
Source: CC calculations based on data from the parties. The dataset has been restricted to those distributors that have
provided data categorized by cancellation reason.
*This table looks at switching of the PPI product whilst not switching the underlying credit product. We noted that, in contrast to
the other products, for CCPPI, when switching the bundle it is not necessary to terminate either the CCPPI or the credit card.
4.40 We found that 1 per cent of PLPPI and SMPPI policies in force at the start of a year
were cancelled whilst the loan was maintained; the figures for MPPI and CCPPI were
9 per cent and 6 per cent respectively. Adding in those customers who terminated
86
their PPI policies in the cooling-off period and the ‘others’ category (which may or
may not include some relevant switchers) increased the levels of possible switching
to 5 per cent (PLPPI), 18 per cent (MPPI), 4 per cent (SMPPI) and 15 per cent
(CCPPI). We concluded that these numbers represented an upper bound on the
number of people switching PPI policy without switching the underlying credit.
4.41 The low levels of switching during the cooling-off period noted in Table 4.2 indicated
to us that few consumers are switching within this initial period when they do not
incur any switching costs.
4.42 Consumers switching PPI without switching credit have to switch to a stand-alone
product—either stand-alone PPI or short-term IP. We looked at sales of these
products to obtain a better estimate of switching of PPI without switching credit.
Table 3.3 shows sales of stand-alone PPI and short-term IP.
4.43 Of the large distributors, only four offer MPPI on a stand-alone basis. Abbey, Alliance
and Leicester and HBOS were the only distributors to report stand-alone MPPI sales
and these represent [less than 7.5 per cent] of all the MPPI policies sold by the 12
largest distributors. Barclays offers CCPPI on a stand-alone basis, although reported
sales are tiny compared with the size of the market (less than 0.1 per cent of total
CCPPI sales). None of the 12 largest distributors offers any other PPI policy on a
stand-alone basis (though, as noted in paragraph 3.38, HSBC now offers to credit
card and personal loan customers the opportunity to discuss their protection needs
(not just those relating to the credit product acquired) with a financial planning
manager, with LifeChoices one of the possible products offered to those customers
who would previously have been offered PLPPI and CCPPI, and as we found in
Appendix 2.3, LifeChoices is a form of PPI. In addition, RBSG’s Churchill brand
offers short-term IP on a stand-alone basis).
4.44 As noted in Table 3.3 and paragraphs 3.40 to 3.43, there are few sales of standalone PPI and short-term IP. We estimate total sales of between 40,000 and 50,000
for 2007 (some of which were sold as a result of referrals made at the point of sale of
mortgages, personal loans and credit cards, and a significant proportion of which
were sold by one existing stand-alone provider that has seen very significant growth
in its PPI sales during 2007 and the first half of 2008), compared with total PPI sales
of over 3 million for the year.
4.45 We received limited evidence from parties on the extent to which consumers switch
PPI without switching credit at the same time. Most said that this was limited, though
two respondents suggested that it was more common for MPPI than other types of
PPI. The OFT survey found that 14 per cent of respondents had cancelled their PPI
but only 2 per cent had switched to another provider 16 and Abbey provided a report
illustrating that less than 1 per cent of customers cancelled MPPI and took out
insurance elsewhere.
4.46 We noted that some new stand-alone products have been launched since our
investigation started, but that sales to date have been low (see paragraphs 5.120 to
5.134) and we saw no reason to find that these new products would cause significant
numbers of consumers to switch their PPI offer in the short to medium term (see
paragraph 5.134).
16
OFT. Twenty-four respondents switched; 150 cancelled; 1,057 surveyed. This is similar for all categories but the sample sizes
are very small. It found that 67 per cent of respondents knew that they could cancel their PPI policy; 9 per cent thought they
could not and 27 per cent did not know. A further 14 per cent of customers had successfully cancelled policies and the main
reasons cited for doing so were that it was too expensive (43 per cent) or circumstances had changed (39 per cent).
87
4.47 We concluded that there was little switching of PPI policies without switching the
credit at the same time. On switching of the bundle, as noted in paragraph 3.108, we
found little direct evidence, and it is difficult for us to come to a definitive view on the
incidence of switching on the bundle. The evidence did show that there was little
switching of bundles of personal loans and PLPPI, but this was insufficient for us to
draw firm conclusions regarding other types of PPI.
4.48 In paragraphs 3.18 to 3.29 we found that the responsiveness of distributors’ PPI
demand to PPI prices was low. This would be inconsistent with significant switching
of combinations of PPI and credit in response to PPI prices. Because we have little
direct evidence of switching of the bundle, we cannot discount the possibility that
there may be some significant churn in combinations of PPI and credit. However, we
consider that it is unlikely that the level of churn is affected by PPI prices, given the
other evidence we have seen. Where that churn is driven by other factors, such as
changes in the APR, or in consumer decisions to extend or consolidate debts, for
example, this will not impose a constraint on PPI prices.
4.49 The evidence of switching does not therefore cause us to alter our conclusions
regarding the general level of competition between PPI providers.
Claims ratios
4.50 The OFT’s document setting out its reasons for making a reference to the CC 17 noted
that average claim ratios (defined as claims paid as a percentage of GWP) were low,
at around 20 per cent. Our analysis, based on data we collected from the six largest
underwriters, showed that the average claims ratios for 2006 varied by type of PPI:
for MPPI it was 28 per cent but for PLPPI and CCPPI the average ratios were as low
as 15 and 11 per cent, with an average of 14 per cent. 18 This would leave over 80
per cent of GWP (over 70 per cent in the case of MPPI) to cover the costs, including
commission payments made by underwriters to distributors, and profits of
underwriters and distributors.
4.51 In assessing whether a policy provides good value for money to a consumer, the
claims ratio is a useful indicator. It shows how the proportion of the premium paid by
the consumer is used in payment of claims. 19
4.52 For most kinds of insurance the claims ratio is over 50 per cent. Using data from the
ABI 20 on net written premium and net written claims, we calculated the claims ratios 21
for 2006 for different insurance types. 22 Our results are shown in Table 4.3.
17
Payment Protection Insurance: The OFT’s reasons for making a market investigation reference to the Competition
Commission, February 2007, OFT899.
18
The claims ratio averaged across all products and over the five-year period 2002 to 2006 was 16 per cent. For PLPPI the
average over that period was 15 per cent, for MPPI 34 per cent and for CCPPI 13 per cent. The average claims ratio declined
over the five-year period.
19
One party ([]) told us that claims ratios were not necessarily a correct proxy for profitability, saying that it did not take into
account differences in costs other than claims and/or revenues other than premiums linked to the insurance policies. It
suggested other ratios which might be more appropriate because they included costs associated with the provision of PPI
([]). Despite the limitations of the claims ratio indicator, we nevertheless believed that it provided a simple and clear indicator
of value for money for consumers.
20
Statistical overview of UK insurance in 2006. Data bulletin, September 2007.
21
The ABI presented data on net written premium whereas our analysis of claims ratios was based on GWP. We did not
consider that these differences had a material affect on our analysis.
22
We noted that the claims ratios, as reported by the ABI in its Data Bulletin Overview of the UK Insurance Market in 2007 (July
2008) were very similar to the claims ratios shown for 2006, with the exception of property which increased to 77 per cent,
apparently as a result of the severe floods in the summer of 2007.
88
TABLE 4.3 Claims ratios for different insurance policies
Insurance type
Motor
Liability
Accident and health
Property
Pecuniary loss*
Claims ratio % (net claims incurred/ net written premium) 78
66
65
54
34
Source: CC, based on data from the ABI Data Bulletin Statistical Overview of UK Insurance in 2006 (September 2007).
*Business that covers a range of policies, which, subject to policy limits etc, reimburse the policyholder for direct financial loss,
which they suffer. The loss must be of money or income etc, and not of goods which are covered under property business.
Pecuniary loss includes the following lines of business: Assistance; Creditor; Extended Warranty; Legal Expenses; Mortgage
Indemnity; Pet Insurance; Other personal financial loss; Fidelity and contract guarantee; Credit; Suretyship; Commercial
contingency.
4.53 Where claims ratios are below 50 per cent, this means that more money paid by
consumers is going to pay expenses and profit of the providers than is paid out in
claims to policyholders.
4.54 As noted in paragraph 4.50, we calculated that the average claims ratio in 2006 for
PPI was approximately 14 per cent, varying between 11 and 28 per cent depending
on the PPI product. To the extent that consumers are aware that the claims ratio is
this low, consumers who would rationally buy this insurance (ie ones who think that
the insurance is good value for money at current prices) are likely to be those who
place a high value upon it and have one or more of the following characteristics:
• high aversion to risk;
• high consequential losses in the event of being unable to make payments; and
• asymmetric information: private knowledge (or belief) that they are an aboveaverage risk (adverse selection—see paragraph 2.55).
4.55 It was not, however, clear to us that information on claims ratios is readily available to
consumers.
4.56 In order to assess whether current prices of PPI are high relative to a competitive
price, we need to make an assessment of costs. Notwithstanding the low claims
ratio, the price could be competitive if this kind of insurance were much more costly
to distribute and administer than other kinds of insurance. For example, for this to be
the case, it would appear to require distribution costs to be three to four times higher
than those of motor insurance.
4.57 Using data from 2006 from large underwriters and distributors, we analysed a typical
£100 of GWP paid by a customer 23 in order to examine how much of it went towards
costs and how much went towards profit. Figure 4.1 shows our estimate of the split of
£100 of premium. Further detail on the profitability of underwriting and the profitability
of distribution can be found in Appendices 7.1 and 4.4 respectively.
4.58 We estimated that £14 of every £100 was typically paid out in claims. Commission
and profit share payments to the distributor amounted to £68. This left the
underwriter with £18 to cover expenses other than claims and make a profit. We did
23
We analysed data from all the large underwriters for 2006 for all PPI products.
89
not find that underwriters had earned unreasonable levels of profit (see paragraphs
7.38 to 7.43).
4.59 In the next section, on profitability, we look in more detail at the profitability of PPI
distribution, including the costs incurred by a distributor in selling a PPI policy. We
found that, based on 2006 data, out of £68 in commission and profit share payments
made to the distributor, costs of just under £11 were incurred in selling a PPI policy.
This left £57 as an additional contribution to common costs and profit, before tax and
capital costs, out of £68 to the distributor.
FIGURE 4.1
Typical split of £100 gross written premium (2006)
Claims paid by underwriter
Underwriter expenses and profit
Distributor fully-allocated selling costs
Amount retained by distributor
Source: CC based on 2006 data provided by the parties.
Profitability
4.60
The CC guidelines 24 state that it is normally helpful to consider the effectiveness of
competition by examining the outcome of the competitive processes in the particular
market. One of these outcomes is profitability. The section in the CC guidelines on
profitability states that a situation where, persistently, profits are substantially in
excess of the cost of capital for firms that represent a substantial part of the market
could be an indication of limitations in the competitive process. 25
4.61
We were interested in examining the profitability of both sections of the value chain—
underwriting and distribution—in order to conclude whether, and if so, at what point in
the value chain, profits were high. This section deals with the profitability of
distribution. We look at the profitability of underwriting from paragraph 7.38.
How we examined profitability of PPI distribution
4.62 Our conclusion on market definition is that PPI is in a separate market from the
underlying credit product (see paragraph 3.139). Our analysis of profitability has
therefore focused on identifying the revenues and costs incurred in the sale of PPI,
as distinct from those incurred in the sale of the credit product.
24
CC3, paragraph 3.78.
CC3, paragraph 3.82.
25
90
4.63
Most distributors said that we should consider the profitability of PPI and the credit
product together, and not try to analyse PPI separately. We disagreed with this view.
Distributors were generally of the view that the correct market definition was that of
PPI and credit combined (the so called ‘system market’). Distributors’ preference for
assessing profitability on a system basis was consistent with their view on market
definition. Whilst we acknowledge that current business practice is to integrate the
sale of PPI into the credit sale, it is clear that PPI is capable of being sold separately.
We do not find that it is so inextricably linked with the credit sale as to render any
attempt to analyse its costs and revenues meaningless.
4.64 Nevertheless, determining the extent of costs to be allocated to PPI is not a straightforward exercise, as many costs are shared with various other activities of the
distributor. In these circumstances, we are of the view that the relevant costs for the
purposes of our assessment of profitability are the fully-allocated costs. 26
Conceptually, the fully-allocated cost approach involves distributing the entire cost
base across relevant business lines; thus the fully-allocated costs include not only
direct costs but also an apportionment of common costs. Fixed and common costs
are fully recovered across the various business lines. Our fully-allocated cost
estimate is based on information from one provider’s activity-based costing (ABC)
system.
4.65 We concentrated on the three main types of PPI policies (personal loan, first-charge
mortgage and credit card). These products comprised a substantial proportion of total
sales of PPI in terms of GWP. 27 Our analysis focused on evidence from the 12 main
party distributors (see Table 2.7). These parties sell the majority of PLPPI (77 per
cent) and CCPPI (84 per cent). They also sell a substantial proportion of MPPI
policies (54 per cent). The remaining 46 per cent of MPPI policies is sold through
smaller distributors (such as smaller banks and building societies) and mortgage
intermediaries.
4.66 Our analysis focused on the economic profitability of PPI (the post-tax profits less a
charge to reflect the cost of capital required to support this business line). We
favoured this measure of profitability since it is widely used in the banking industry as
a key measure of business performance and shareholder value creation. We also
calculated return on equity capital, being post-tax profits divided by equity capital.
This gives an indication of the magnitude of returns in percentage terms and is widely
used and reported in the banking industry. We decided that the use of a combination
of economic profits and return on equity capital afforded an economically meaningful
picture of both the absolute and relative (in relation to the cost of capital) level of
profits earned from PPI distribution.
4.67 One distributor (Barclays) said that the net present value and internal rate of return
(IRR) were commonly accepted and well-established methods for measuring the
profitability of an economic activity and were suitable for assessing profitability in
competition cases. We accept that under a number of conditions, the IRR is
theoretically superior to other measures of profitability such as return on capital.
However, in this particular case our choice of methodology was dictated to a large
extent by the availability of data. As outlined above, most providers said that they had
little or no information about the costs (whether of a revenue or capital nature)
incurred in the supply of PPI. We did not believe that in the circumstances of this
case an IRR-type assessment of profitability would have been possible, but if it had
26
This approach is endorsed by Oxera’s report for the OFT on Assessing profitability in Competition Analysis, July 2003, which
states: ‘In a profitability assessment in competition policy, the FDC (fully distributed cost) is often the relevant starting point’.
27
In 2006, these three products made up 83 per cent of the GWP (48, 12 and 23 per cent respectively) underwritten by the
largest six underwriters.
91
been, we do not believe that it would have produced a substantially different picture
from the one portrayed by our assessment of return on capital. This is because many
of the limitations of accounting measures 28 such as return on capital apply to a lesser
extent to an established, mature and non-capital-intensive business such as PPI
distribution. In any case, we do not regard our estimates of economic capital or return
on equity capital as definitive computations of profitability which could be said to be
correct with 100 per cent certainty. Rather, we consider them to be indications of the
levels of profits in the industry, which may be expected to provide reasonable results
in this particular case, given the nature of the industry and the careful consideration
we have given to ensuring that they fit the purpose in hand.
4.68 One distributor ([]) disagreed with our approach to the analysis of profitability,
stating that although economic profit and return on equity were widely used by banks
at a corporate or business unit level as key measures of business performance and
shareholder value, contribution was a far more widespread and meaningful measure
of individual product performance. We disagreed with using contribution as a
measure of profitability as it did not take account of all costs 29 or the cost of capital
involved in the distribution of PPI. It also could not be compared against an
appropriate benchmark of the cost of capital.
4.69 Some distributors (HBSC, HBOS) said that we should assess the profitability of
‘protected credit’ sold with PPI and ‘unprotected credit’ sold without PPI. We disagreed with the approach of including credit in our assessment of profitability. As
stated above, we did not consider that PPI was so inextricably linked with the sale of
the underlying credit as to render any attempt at cost allocation meaningless. We
also considered that information about the profitability of PPI and credit combined
would not provide a clear indication of the extent of competition in the distribution of
PPI. We do, however, discuss the profitability of the PPI and credit combined from
paragraph 4.83 and in more detail in Appendix 4.5.
4.70 Whilst we understood that for routine management reporting purposes PPI was not
assessed as a separate profit centre, for the purposes of our inquiry we concluded
that considering any available information on the separately identifiable revenues and
costs incurred in the distribution of PPI was a useful exercise. In particular, we
concluded that an assessment of the revenues generated, and the costs incurred, as
a result of the sale of PPI could give a useful indicator of the state of, and nature of,
competition. However, for reasons relating to cost allocation explained in paragraph
4.64, we did not ask the distributors to create a hypothetical profit and loss account
on this basis for the purpose of this investigation. With this in mind, on reviewing the
board minutes and supporting strategic planning material submitted to us as
evidence, we did find some references to the profitability of PPI as distinct from the
credit product. We also found that one significant distributor had developed a model
of market economic profitability for the retail banking business (called the ‘Market
Economics Model’) which included PPI as a separate profit pool.
4.71 We thought that this Market Economics Model provided a useful framework for the
assessment of profitability of the distribution of PPI. However, we decided that it was
appropriate to adapt it so that we could use our own inputs and assumptions in the
model. To this end, we collected data from the 12 main distributors of PPI and
revised the model where necessary to take account of data from those parties,
including considering all revenues and costs which were related to the distribution of
PPI. We considered the following inputs for the revised model:
28
For a full discussion of the possible limitations, see, for example, Oxera’s report for the OFT on Assessing profitability in
Competition Analysis, op cit.
29
We note Oxera’s report for the OFT, ibid, and in particular its discussion of other measures of profitability.
92
• the share of premium income earned by distributors;
• levels of interest income earned on PPI premiums;
• the operating costs associated with the distribution of PPI;
• the effect of PPI on impairment losses; and
• the appropriate capital base and cost of capital for the distribution of PPI.
4.72 We recognize that profits of individual firms may fluctuate over time and may at
times, even in competitive markets, exceed the cost of capital by a substantial
amount. Nevertheless, the process of entry and rivalry should see very high levels of
profits eroded to normal levels over time. We did not attempt to analyse the profits of
individual firms in this investigation since we were not able to obtain firm-specific data
on costs or capital requirements. The work we carried out is an assessment of the
combined profitability of firms representing a substantial part of the share of supply of
PPI. 30
Results of profitability analysis
4.73 From the revised model of economic profitability we calculated that the 12 largest
distributors made economic profits on PPI of £1.4 billion in 2006, representing an
RoE of 490 per cent. Of the total economic profits of £1.4 billion, PLPPI contributed
£645 million, CCPPI contributed £336 million, MPPI contributed £112 million, and
SMPPI contributed £126 million. Other forms of PPI made up the remaining
£147 million. The RoEs for PLPPI, CCPPI, MPPI and SMPPI were similar to the
overall level of 490 per cent. Further detail can be found in Appendix 4.4.
4.74 We also carried out sensitivity analysis on the results, first by flexing the costs of
selling PPI and second by flexing the equity capital allocation.
4.75 As explained in paragraph 4.64, we estimated the costs of selling PPI on a fullyallocated basis as distinct from the costs of selling the underlying credit product, on
the basis that these costs would be incurred irrespective of whether or not PPI were
sold. Parties argued that this approach was not appropriate as most costs were
common to both sales. Thus, we carried out sensitivity analysis, first regarding costs
and second regarding capital allocation. We estimated the maximum costs shared
between selling PPI and credit and added those to the cost of selling a PPI policy
and examined the effect of including the resulting figure in the revised model.
Second, we doubled the equity capital allocation in the revised model. The result of
this exercise was still greatly in excess of any reasonable estimate of the cost of
equity.
4.76 We found that the distribution of PPI is a low-risk activity for the distributor. Our
review of contracts between underwriters and distributors, and responses from the
parties, indicates that the insurance risk is borne by the underwriter, at least in the
short term. In the medium to long term, the distributor suffers the risk of future
commission and profit share income streams being lower than otherwise rather than
actual losses, and thus lower than anticipated profits. We noted that an underwriter’s
downside is unlimited (subject to contractual changes) whereas its upside is limited
30
We note that approximately 46 per cent of MPPI is sold through smaller distributors (such as banks and building societies)
and mortgage intermediaries. We did not include costs paid to mortgage intermediaries for the sale of MPPI in our model. To
the extent that these exist, the profitability could be overstated.
93
to a small portion of profit share and investment income. In contrast, a distributor’s
downside is limited to reduced or no profit share, but it will not share in losses
suffered by the underwriter, while its upside is unlimited: the larger portion of profit
share and investment income accrues to the distributor (see Appendix 7.1 for more
details).
4.77 We also calculated that the 12 largest distributors made economic profits on PPI of
£1.4 billion in 2007, representing an RoE of 495 per cent. Further detail can be found
in Appendix 4.4.
Documentary evidence on profitability
4.78 In our review of available evidence we looked at documents such as parties’ board
minutes and strategy documents, and analysts’ research notes. Whilst we did not
seek to rely on these comments as direct evidence on the profitability of PPI, we
noted that they supported the conclusions of our analysis that PPI is highly profitable.
Conclusions on profitability
4.79 We looked at the profitability of PPI on this basis, using a model of market economic
profitability supplied by one significant provider ([]), a revised model incorporating
certain amendments to that model, and evidence from board papers and strategy
documents. The results of our own analysis and the material that we have seen from
parties on the profitability of PPI consistently indicate that:
• PPI distribution is highly profitable. Distributors earn a high proportion of the total
income from PPI premiums and in comparison the costs incurred in selling PPI
are low.
• The distribution of PPI is a low-risk activity for the distributor (the risk lies with the
underwriter) and consequently the additional capital required to support the PPI
distribution business is relatively low.
4.80 We considered it useful to show diagrammatically the split of £68 (see paragraph
4.58). We estimated that just under £12 of every £68 went on distribution costs,
including the cost of capital. £17 related to tax on profits, and the remaining £39 was
economic profit. This is shown in Figure 4.2.
94
FIGURE 4.2
Split of £68 profit share and commission to distributor (2006)
Fully-allocated selling costs
Tax on pre-tax profit
Capital charge
Economic profit
Source: CC based on data provided by the parties.
4.81 We found that these levels of profitability had persisted over at least the last four
years. No distributor provided evidence to suggest that the picture had been different
prior to this period although several parties made comments regarding the relationship between claims ratios, profitability and the health of the wider economy: one
distributor (Abbey) said that we should consider PPI returns over the full insurance
cycle as a short period might not be representative; another distributor (Nationwide)
said that the overall economic cycle impacted on default and delinquency rates and
that these factors also affected the profitability of PPI products since rising
unemployment would produce higher claims costs. One underwriter (Genworth) said
that claims ratios increased dramatically when there were downturns in the economy
and that it was essential to take into account the full economic cycle when assessing
the level of claims and the significance of claims ratios; another underwriter (Aviva)
told us 31 that profitability fluctuated over the course of the economic cycle. However,
we thought that the prices of PPI policies could be adjusted in response to changing
economic conditions. We also thought that policies were of a short-term nature:
although many single-premium policies can have policy terms of five years or more,
we note that a significant proportion of policies are terminated prior to their full term
and that the average length of time a policy is held is therefore shorter than this. 32
4.82 We consider that the evidence suggests that distributors representing a substantial
part of the share of supply of PPI have earned profits that were persistently and
substantially in excess of the cost of capital.
Profitability and pricing of the bundle of PPI and credit
4.83 As noted in paragraph 3.95 and Appendix 3.1, many parties put it to us that PPI and
credit were fundamentally interlinked in that they were sold together, priced interdependently, run as an integrated business, and shared a common cost base. As
31
In the context of a discussion on deal construction. See Table 1 in Appendix 4.3, which shows the breakdown of terminations for PPI (2002 to 2006) as a percentage of policies at the start of the year. 52 per cent of PLPPI policies were terminated early, of which 33 per cent were due to early settlement of loan. 32
95
noted in Section 3, we consider that PPI is in a separate product market to credit. We
therefore consider that evidence on the profitability of PPI as distinct from the credit
product is most relevant to our assessment of competition in that product market.
However, to understand further the relationship between PPI and credit and the
effects on consumers as a result of this relationship, we examined evidence on the
profitability of PPI distribution in conjunction with the credit product, and the effect of
PPI income on lending policy and pricing. This analysis can be found in Appendix 4.5.
4.84 We concentrated on the three main types of PPI policies (personal loan, first-charge
mortgage and credit card). As noted in paragraph 4.65, these products comprised a
substantial proportion of the share of supply in terms of GWP.
4.85 We looked at the profitability of PPI in combination with the underlying credit product
using a model of market economic profitability supplied by one significant provider; a
range of board papers, strategy documents and analysts’ reports; and financial data
provided by the parties. We also created two models of market profitability, one for
personal loans and one for credit cards. The results were generally consistent. The
results of our analysis on the profitability of the credit and PPI bundle suggest that:
• The personal loans business has suffered from declining profits in recent years to
the point where in 2006 it appears to have been loss making before taking into
account income from PPI. With PPI included, the sector appeared to have been
marginally profitable. This appears to be a recent phenomenon: the evidence
suggests that prior to 2005 the personal loans sector was profitable, even without
PPI income.
• The credit card and mortgage sectors do not appear to have been as reliant on
income from PPI in recent years. PPI penetration has historically been lower and
income from PPI generally less significant than for personal loans. The evidence
that we have examined suggests that both sectors have been profitable over the
last five years, even before taking into account income from PPI.
4.86 We also considered the reasons for declining margins (and recent losses) on
personal loans, and in particular whether this was as a result of a policy to crosssubsidize the sale of credit with income from PPI. Distributors appeared to have
operated a range of different pricing strategies. Some distributors (eg []) suggested
that there was deliberate cross-subsidy. One distributor told us that the declining
margins were a result of intense competition in the personal loans market and that
distributors had lowered APRs to the point that they were unsustainable without PPI
income ([]). Another distributor ([]) told us that cross-subsidy was occurring as a
consequence of market forces driving down loan APR. Two distributors ([])
suggested that the phenomenon was more to do with mis-pricing, ie impairment
losses had been higher than expected, and that, in hindsight, APRs had been set too
low. In particular, we heard from various parties (including []) that impairment
losses had been higher than anticipated due to several factors, including:
• changes to bankruptcy laws in 2004 which had the effect of making it easier to file
for bankruptcy than previously and consequentially there has been an increase in
the number of people filing for bankruptcy;
• the introduction of individual voluntary arrangements (IVAs) and the emergence of
specialist firms marketing IVAs to consumers; and
• macroeconomic factors including over-indebtedness, and higher interest rates.
96
4.87 From the profitability evidence alone, it is impossible to determine the extent to which
declining margins (and in some cases losses) in personal loans result from ex-ante
strategies to compete aggressively on APR and/or cross-subsidize APRs with PPI
income or from unexpectedly high impairment rates. We therefore supplemented our
analysis of profitability with an examination of evidence on credit pricing and lending
strategies.
4.88 The evidence suggests that:
(a) PPI income has influenced pricing decisions on personal loans because it has
been taken into account in the loan pricing strategies and pricing models of
lenders.
(b) Anticipated income from PPI has influenced lenders’ determinations of creditscore cut-offs.
(c) Some segments of consumers falling within low credit score bands would be
unprofitable were it not for PPI income earned within that segment. This suggests
that unprotected consumers at lower credit scores may, in particular, be crosssubsidized by consumers purchasing PPI.
4.89 This evidence, in conjunction with the evidence on declining profitability and, in some
cases, losses on personal loans, leads us to conclude that prices for personal loans
have been distorted, to some extent, by the ability of distributors to anticipate PPI
income on those loans. We conclude that this distortion has caused loan prices to be
lower than they would otherwise have been were it not for PPI. There is some
evidence to suggest that this distortion is greater at lower credit scores where PPI
penetration tends to be higher, impairment losses tend to be more significant, and
therefore the importance of PPI income is greater.
4.90 The evidence that we have seen suggests that PPI income has had significantly less
influence on the pricing decisions on credit cards and mortgages. This is consistent
with our analysis of profitability of the credit card and mortgage sectors which
suggested that both sectors had been profitable over the last few years, even before
taking PPI income into account. On balance we conclude that there is little evidence
of distortion of credit card or mortgage prices as a result of PPI income.
4.91 It was suggested to us that that we should analyse the profitability of credit sold with
PPI and credit sold without PPI, and the extent of cross-subsidy within those groups.
We assess the extent to which PPI profits result in lower credit prices in more detail
in paragraphs 10.442 to 10.464.
Conclusions on the extent of competition between distributors
4.92 All the evidence on PLPPI suggested to us that there is little competition on PLPPI
between distributors. The sale of PLPPI is highly profitable (paragraph 4.73). Whilst
there is some price variation over time (paragraph 4.8), we considered it to be very
limited, and we found no correlation between price and quality across PLPPI
products (paragraph 4.31). We found that PLPPI products are difficult to compare
(paragraph 5.40), and that there is limited search for PLPPI (paragraph 4.33) and a
lack of advertising of PLPPI prior to the point of sale of the credit (paragraph 4.21).
4.93 The evidence on CCPPI also suggested that there is little or no competition on
CCPPI between distributors. The sale of CCPPI is highly profitable (paragraph 4.73).
We saw minimal price variation over time (paragraph 4.8), and we found no
correlation between price and quality across CCPPI products (paragraph 4.31). We
97
found that CCPPI products are difficult to compare (paragraph 5.40), with a large
variation in benefits payable for products of very similar price. We found that there is
limited search for CCPPI (paragraph 4.33) and a lack of advertising of CCPPI prior to
the point of sale of the credit (paragraph 4.21).
4.94 The evidence on SMPPI led us to conclude that there is little competition on SMPPI
between distributors. The sale of SMPPI is highly profitable (paragraph 4.73). We
saw no price variation at all over a five-year period in the products we looked at
(paragraph 4.8). We found that it was difficult to compare products (paragraph 5.40),
and that there was limited search for SMPPI (paragraph 4.31) and a lack of
advertising of SMPPI prior to the point of sale of the credit (paragraph 4.21).
4.95 The evidence on MPPI included some indications that there may be more
competition on MPPI between distributors than there is for other PPI products, but
the evidence overall indicated to us that there is no significant degree of competition
on MPPI as between distributors and intermediaries. We found high profitability for
distributors (paragraph 4.73), though we did not look at the profitability of selling
MPPI for intermediaries. We found virtually no price variation over time (paragraph
4.8). We did find some correlation of price with quality, though there was a wide
dispersion within this trend (paragraph 4.31). We noted that there was some
searching carried out by consumers for MPPI (paragraph 4.33), though we noted that
most search takes place on combinations of MPPI with mortgages and that Internet
comparison websites are not configured to allow consumers to search for such
combinations.
4.96 The evidence led us to conclude that there is not effective competition in the supply
of PLPPI, MPPI, SMPPI or CCPPI, and that prices for PPI are higher than they would
be in a well-functioning market. Moreover, given the lack of competition, the
consequent lack of advertising and low levels of consumer search, consumers at the
credit point of sale have little awareness of the alternative policies that may be
available and we conclude that there is less choice for consumers than there would
be in a well-functioning market.
Coordination in the distribution of PPI
4.97 We considered whether it was likely that there would be adverse effects on
competition as a result of coordination between the distributors of PPI. However, as
we found that there was little competitive interaction between distributors, and that in
effect each distributor had a monopoly over the supply of PPI to its own credit
customers, we concluded that coordination in the distribution of PPI was unlikely.
Distributors can already charge high prices without the need to coordinate; there is
therefore very little incentive for distributors to coordinate their behaviour. We
therefore decided not to conduct an analysis of whether PPI markets were
susceptible to coordinated effects.
98
5. Factors affecting the nature and extent of competition in the supply of
PPI
5.1 Having established that there is little competition between providers of any type of
PPI policy sold at the point of sale, we looked at why this is the case, and what
factors are involved. The purpose of this was to see if there are any features which
are preventing, restricting or distorting competition in the supply of PPI.
5.2 We focused on five factors which might contribute to the lack of competition:
• the extent to which distributors actively seek to compete with each other;
• whether comparing PPI products is beneficial and whether there are barriers to
such searching;
• having taken out PPI, whether consumers switch between PPI products, and
whether there are barriers to switching;
• whether there are barriers to entry into the supply of PPI by new providers, or
barriers to expansion by suppliers already in the market; and
• whether there is an advantage for distributors to selling PPI at the credit point of
sale (the point-of-sale advantage) and whether this contributes to the lack of
competition we found.
5.3 We then looked at whether some stand-alone PPI and short-term IP products which
have recently been introduced to the market might impose a competitive constraint in
the short to medium term.
5.4 Finally, we looked at the effects on consumers of the level of competition in the
market.
5.5 For the reasons set out below we reached the following conclusions:
(a) Most distributors and intermediaries do not actively seek to win customers by
using the price of their PPI policies as a competitive variable; and that this is a
feature of relevant markets that prevents, restricts or distorts competition in the
supply of PPI (see paragraphs 5.6 to 5.8).
(b) There are significant barriers which hinder consumers who try to search for PPI,
which limit consumer choice and act as a barrier to expansion for other PPI
providers, in particular providers of stand-alone PPI; and that this is a feature of
relevant markets which prevents, restricts or distorts competition in the supply of
PPI (see paragraphs 5.16 to 5.57).
(c) There are significant barriers to switching between PPI policies, which limit
consumer choice and act as a barrier to expansion for other PPI providers, in
particular providers of stand-alone PPI; and that this is a feature of relevant
markets which prevents, restricts or distorts competition in the supply of PPI (see
paragraphs 5.58 to 5.87).
(d) The sale of PPI at the point of sale of the credit product further restricts the extent
to which other PPI providers can compete effectively and is a feature of relevant
markets which prevents, restricts or distorts competition in the supply of PPI (see
paragraphs 5.88 to 5.119).
99
Competition between PPI providers
5.6 In competitive markets we would expect firms to seek to win business by offering
desirable products (based on a combination of price and quality) and proactively to
advertise their products to customers. In a well-functioning market we would expect
to see a range of activities associated with active price and non-price competition
between suppliers. These could include: offering price reductions, discounts and
other sales promotions; advertising based on price; and comparative advertising
along with other efforts to simplify customers’ choices.
5.7 However, we have not observed them in relation to PPI. In particular:
• We saw little in the way of price competition. We saw very little evidence of price
reductions in the period 2002 to 2006 (see paragraphs 4.8), and little evidence of
discounts on PPI price or sales promotions (see paragraph 4.20).
• We saw little in the way of competition in relation to non-price competition (see
paragraph 4.23).
• We saw very little advertising of PPI, and little evidence of distributors seeking to
win business through the price of their PPI offer (see paragraphs 4.20 to 4.21).
We also saw very little in the way of comparative advertising (see paragraph
4.20).
5.8 We concluded, therefore, that most distributors and intermediaries fail actively to
seek to win credit and/or PPI customers by using the price or quality of their PPI
product as a competitive variable. We conclude that this is a feature or relevant
markets which prevents, restricts or distorts competition in the supply of PPI. We
thought that this feature applies to all PPI products.
Search
5.9 We looked at whether customers benefit from searching for the best PPI product, and
whether there are barriers to effective searching. Further details of our analysis on
these issues can be found in Appendix 5.1.
Benefits to search
5.10 We were told by many parties that customers search for the best-value credit
product, and several parties told us that customers search on credit APR. The
implication of this is that, in practice, customers find it worthwhile searching on credit
to find the best-value credit product.
5.11 However, despite these potential gains from searching, the evidence of the extent of
search showed that consumers are not actively comparing products. We found that
the extent of search through comparing a PPI policy with one or more stand-alone
policies was low (see paragraph 3.61) and the limited extent to which customers
compared bundles of credit and PPI did not indicate that there were significant levels
of search for PLPPI and SMPPI. We found that there was a greater degree of
searching for MPPI, (see paragraph 3.113); however, given that there would appear
to be a wide dispersion in MPPI prices, there was less searching than we might
expect. The relatively low level of searching, in the light of significant potential gains
from searching therefore suggested that there were significant barriers to search,
which were preventing or deterring customers from realizing these gains.
100
5.12 We looked at whether low PPI prices were offset by high credit prices, such that
searching on the cost of a bundle of PPI and credit would be a good way of finding
the best-value PPI product. Figures 5.1 and 5.2 show the dispersion of the cost of
credit, PPI, and the bundle of credit for personal loans and mortgages. 1 These are
drawn as box plots. 2
FIGURE 5.1
Dispersion of cost of credit and PLPPI in March 2006
(for 129 PPI/loan product combinations)
£ per month for a £2,000 loan over 12 months
80
60
40
20
0
Cost of credit only
Cost of PPI only
Total cost of credit and PPI
Source: CC calculations based on data from the parties.
1
For personal loans, we used data on PLPPI premiums and APR requested from the 12 largest credit providers for a loan
repaid over one year. This data was provided by 12 major credit providers for 80 loan products, in combination with one or
more PLPPI products for five years from 2002 to 2006. For mortgages, we used data on MPPI premiums and APR for a
£100,000 mortgage, repaid over 25 years for 90 products, in combination with one or more MPPI products for five years from
2002 to 2006.
2
A box plot shows the median (the line inside the box) and the inter-quartile range (the box); and the lines extending from the
box represent the furthest point from the inter-quartile range that is within 1.5 times that of the inter-quartile range. Points
outside of this range are marked individually.
101
FIGURE 5.2
Dispersion of cost of credit and MPPI in March 2006
(for 90 MPPI/mortgage product combinations)
£ per month for a £100,000 mortgage over 25 years
500
400
300
200
100
0
Cost of credit only
Cost of MPPI only
Cost of credit and PPI
Source: CC calculations based on data from the parties.
5.13 We found price variation between PLPPI policies, for both PLPPI on a stand-alone
basis and the credit-PLPPI combination. Price dispersion for MPPI appears to be
less pronounced than that of PLPPI. However, stand-alone providers were not
contained within the dataset and their inclusion may increase dispersion. As shown in
Figure 1 of Appendix 5.1, MPPI prices do still show price dispersion, with the highest
price more than three times the lowest price. For PLPPI the data also confirmed the
evidence obtained from Which? data (paragraph 2.71) about the relative costs of
PLPPI and personal loan interest. Figure 5.1 shows that the median cost of personal
loan interest is below the median cost of PLPPI. For the majority of PLPPI policies
the cost of PPI is greater than the interest and fee cost of the underlying personal
loan, and in some cases up to twice the cost (see paragraph 2.73). Figure 5.1 also
shows that high PLPPI costs are not cancelled out by low personal loan interest
prices.
5.14 For CCPPI, we observed minimal price variation between policies, but we observed
significant variation in benefit calculations between policies suggesting that consumers may still benefit from search. We did not have data from which we could
analyse price dispersion for SMPPI.
5.15 The evidence led us to conclude that there are benefits to be gained for customers in
searching for the best-value PPI policy, but that searching on price alone will not
guarantee that the best-value policy is purchased.
102
Barriers to search
5.16 Whilst we found that there are benefits to search, as we noted in paragraph 5.11, we
thought that there were significant barriers to search. We considered what these
might be.
5.17 We focused on four possible barriers to search:
• the availability of, and ease of access to, information;
• the ease of using information to compare products;
• the relationship with the credit provider; and
• the availability of PPI on a stand-alone basis.
The availability of, and ease of access to, information
5.18 As we found in paragraphs 3.46 to 3.47, a significant proportion of consumers do not
realize that they can buy PPI from anyone other than their credit provider. For those
who do, we looked at what information on products is available and the ease of
access to this information.
5.19 We thought that there were three ways customers could compare policies: looking at
firms’ advertised rates; getting individual quotes from providers; and using Internet
resources such as Internet comparison websites. We considered the ease of access
to these forms of information.
5.20 In terms of advertising, some respondents said that PPI was not always advertised,
and we found that the majority of PPI advertising is in conjunction with the underlying
credit. Where PPI has been advertised separately, it has tended to be ‘below-the-line’
advertising (such as brochures available in-branch and direct mailing).
5.21 As noted in paragraph 4.20, we found only a few examples of companies providing
information to consumers comparing their PPI offer with those of competitors.
Further, the evidence we received from parties on advertising spend, and data
obtained from Nielsen, suggested that advertising spend on PPI alone was very low
compared with advertising spend on credit, and to other insurance policies.
5.22 However, the extent to which the lack of advertising is a barrier to search depends on
the extent to which advertising is successful in encouraging people to buy the PPI
policy advertised. The evidence we received suggested that advertising PPI has not
been very successful at generating sales growth at current price levels. We received
evidence from companies which sell, or have sold, stand-alone PPI, that very large
advertising campaigns would be required to generate significant sales of policies,
with evidence of extremely low response rates to direct mail campaigns. 3 Similarly, a
provider of PPI sold in combination with credit told us that when it used direct mail to
3
Aviva told us that for its stand-alone flexible MPPI policy direct mail ‘stand alone’ campaigns generated response rates of less
than [] per cent. HBOS told us that it piloted stand-alone PPI insurance with direct mail in a significant volume. []
103
advertise the benefits of its PPI it saw no material uplift in awareness of the benefits
of PPI and take-up rates were no better than for other direct mail promotions. 4
5.23 The evidence suggested that there is very limited advertising of PPI but that
advertising of PPI on its own is not successful in significantly increasing PPI sales in
the current market structure and at current prices.
5.24 We found in paragraph 4.21 that providers tend to advertise the credit on its own,
and introduce the PPI when the consumer has looked into the credit. Advertising of
PPI therefore primarily takes place at the point of sale.
5.25 In terms of getting quotes from individual providers, there are two ways of getting
quotes: by getting an informal quote; or getting a full quote by going through the full
sales process.
5.26 For PLPPI, SMPPI and MPPI it is possible to get informal quotes from some
company websites, which are based on the credit’s typical APR. As such, the quote
should be fairly accurate for two-thirds of customers. Some banks explained that
existing customers can obtain an accurate quick quote over the telephone or in a
branch based on the information already held on them. With some banks, however,
an accurate quote can only be obtained by going through a complete loan plus PPI
sales process, which leaves a credit footprint.
5.27 In order to get an accurate quote for the price of PPI, a consumer has to get a quote
for credit as well, as the sales process for credit and PPI is integrated (as noted in
paragraph 44 of Appendix 5.1, for some of the largest distributors the only way to get
a quote for PPI is to go through the entire credit application process). The sales
processes for credit vary in length but in our view are relatively time-consuming in
order to get a quote for PPI (we received evidence that for a personal loan the
application process might take about an hour). 5 We considered that the length of
time required to get an accurate quote is a cost to searching that would have to be
borne by consumers. Other costs can include time and money spent looking on the
Internet and making telephone calls, but we thought the incremental costs were
relatively low.
5.28 For credit cards, it appears that quotes are based on a fee per £100 of statement
balance; therefore they do not tend to vary between illustration and credit application
quote.
5.29 As noted in paragraph 4.22, when PPI prices are provided to consumers by
distributors and intermediaries, they are provided in a way which does not facilitate
easy comparison of the relative costs of the PPI and the interest payable on the
credit. PPI is presented as a small component of the total monthly repayment. If this
comparison were easier, consumers might find it easier to find the overall best-value
combination of PPI and credit (either as a bundle or by choosing the best-value credit
on its own and the best-value stand-alone PPI to accompany it).
5.30 The CC GfK NOP 2008 in-home survey found that for those consumers who
compared PLPPI, MPPI and SMPPI products, comparison websites, search engines
and providers’ websites were popular sources of information. As noted in paragraphs
4
More formally, the Dorfman-Steiner condition (Dorfman-Steiner (American Economic Review, 44, 826–36, 1954), implies that
for a profit maximizing monopolist the advertising-to-sales ratio is equal to the ratio of the elasticity of demand with respect to
advertising relative to the elasticity of demand with respect to price. A low advertising sales ratio therefore implies that demand
is relatively unresponsive to advertising expenditure.
5
Summary of post-Emerging Thinking hearing with Lloyds TSB, paragraph 24.
104
3.117 to 3.118, Internet comparison websites allow consumers to search for different
PPI products to differing degrees. The comparisons possible on the main Internet
comparison websites and the extent to which consumers use these websites is
analysed in Appendix 3.8
5.31 Price comparison websites can be useful tools for searching; however even with the
recent launch of the FSA PPI comparison tables it is still not always possible to
search on combinations of credit and PPI, or to compare the prices of the PPI
component of several different bundles of credit and PPI. It is also difficult to search
on combinations of price and quality; the websites focus on price comparisons. The
Internet comparison website operators we spoke to told us of no significant new
developments which would alter this situation.
5.32 On the basis of the evidence we have seen on the availability and ease of access to
information, we believe that consumers face significant hurdles when trying to obtain
systematized information on PPI, and that hinders their ability to compare PPI
products. These hurdles may affect their behaviour in terms of which PPI policy, and
sometimes which combination of PPI and credit, they buy.
The ease of using information to compare products
5.33 The second possible barrier to search we considered was the ease of using
information to compare products. We looked at the clarity and complexity of information provided, customers’ understanding and their financial skills, and the role of
financial advice.
• Clarity and complexity of information provided
5.34 The pricing structures of different PPI policies are variable and make comparisons
difficult. We noted several ways in which PPI policies’ prices and terms can make
comparisons difficult:
• PPI prices are not a good guide of the product quality (paragraph 4.31).
• We found (as did Defaqto 6 ) that comparisons of PPI policies on a like-for-like
basis are not easy, as they each have differing pricing methods (see Appendix
2.1). We found the pricing of single-premium PPI to be particularly complex (see
paragraphs 2.84 to 2.89), making like-for-like comparisons especially difficult both
with regular-premium PPI and other single-premium policies.
• For personal loans, mortgages and second-charge mortgages, the price of the
PPI will depend on the amount being borrowed and the term of the loan, and this
may not be readily available without an individual quote that generates a credit
footprint (for credit cards, PPI is typically quoted as a fee per £100 balance). The
need for customers to obtain an individual quote was greatest for single-premium
PPI, as a result of the lack of any standard metric (or ‘common currency’) by
which the price of single-premium PPI may be compared between providers (see
paragraph 2.84 to 2.89)
• Variations in policy terms and coverage make price comparisons difficult as
consumers are not comparing like with like.
6
Defaqto, Payment Protection Insurance 2008, The Party’s Over, February 2008.
105
5.35 In addition, the way in which PPI price is shown relative to the price of interest on the
underlying credit product does not make clear the relative magnitudes of the two.
5.36 For example, we looked at the cost of taking a £5,000 loan over three years on
moneysupermarket.com. 7 The information provided for the credit offer at the top of
the list 8 was:
Loan amount
APR
36 monthly payments of
Total charge for credit
Total amount payable
Amount per month with optional PPI
Total amount payable with optional PPI
£5,000
6.9%
£153.68
£532.48
£5,532.48
£175.44
£6,315.84
5.37 From this, a consumer can deduce that the monthly cost of PPI is £21.76, which
looks relatively small in comparison with the overall monthly payment as it adds only
14 per cent to monthly repayments. This is comparing the PPI cost with the cost of
repaying both interest and capital.
5.38 This is typical of the way the PPI price is shown to consumers. We noted that most
consumers focus their search on credit, rather than on the combination of credit and
PPI (see, for example, Figures 3, 4 and 5 of Appendix 3.8), and that searchers for
credit appear to focus on APR. Further, we noted that the price of taking out PPI is
often more than the cost of interest and can vary significantly (see, for example,
Figures 2.1 and 2.2 and paragraph 2.73). Searching for PPI can therefore have a
significant impact on the value for money consumers achieve on their protected
credit products. However, as the cost of PPI is typically shown relative to the total
payment on credit, as opposed to the interest cost, consumers may not be fully
aware of the relative importance of PPI. We were concerned, therefore, that consumers are not receiving the information they need in a suitable format for them to
make a reasoned choice as to the best-value product. In particular we thought that
information which allowed an easy comparison between the cost of the PPI on credit
and the cost of interest on the credit would help consumers search for the best-value
credit product, PPI product, or combination.
5.39 We considered the non-price characteristics of PPI policies whether sold alongside
credit or not, in order to understand the complexity of PPI products. We noted that
there is considerable scope for variations between products, such that comparisons
are difficult:
• The types of cover offered by policies can vary considerably (as noted in
paragraph 2.3, PPI usually offers ASU and sometimes L cover; however, some
products offer CI or hospitalization (H) cover as well).
• The period of time for which a policy pays out varies considerably by policy,
depending on whether claims are paid out on a daily or monthly basis, and
whether they are paid retrospectively or have an excess period at the start of a
claim. Figure 5.3 shows the proportion of policies that pay out using the different
combinations, according to Defaqto. The policies likely to pay out the most over
7
Information is provided in different formats by different providers. However, every format we have seen shares with moneysupermarket.com the basic principle of not showing the monthly cost of PPI in comparison with the monthly cost of interest. 8
Your personal loan.co.uk, on 27 April 2008. 106
the length of a claim (policies that pay out for each day of a claim, going back to
the first day of the claim) are at the bottom of each bar; 9
FIGURE 5.3
Defaqto: benefit types by policy
percentage of policies
100
2%
2%
80
60
6%
25%
44%
72%
41%
40
20
50%
4%
32%
22%
0
MPPI
PLPPI
Retrospective, daily
Excess, daily
CCPPI
Retrospective, monthly
Excess, monthly
Source: Defaqto, Payment Protection Insurance 2008, The party’s over, op cit.
• Policies also have initial exclusion periods during which a policy cannot be
claimed on (for example, unemployment claims typically cannot be made within
the first 30 to 120 days, depending on the policy).
• The amount of benefit paid out can vary considerably. The number of payments
made varies—for example, an accident and sickness claim can pay out for
12 months, or for the whole period of the loan. With credit cards the amount paid
each month of a claim can vary from 3 to 15 per cent of the outstanding balance.
• The exclusions for pre-existing conditions can vary considerably. Different policies
have different approaches to claims for backache, stress and chronic conditions.
5.40 This evidence led us to conclude that PPI policies are sufficiently complex to make
comparisons between policies difficult. This was in line with the OFT’s analysis 10 that
policy terms and conditions varied between providers.
• Customers’ understanding and financial skills
5.41 The report of the CC BMRB 2007 survey concluded ‘Findings indicate that overall
knowledge and awareness is very limited and there is a considerable amount of
confusion on product features and benefits’. 11
5.42 We considered whether the ability to search effectively was affected by the extent to
which consumers had the ability to understand the information they received or
found. PPI consumers appear not to be very different demographically from non-PPI
customers; there is certainly nothing about the make-up of the group of consumers
9
Defaqto, Payment Protection Insurance 2008, The party’s over, op cit; the most favourable method for the customer of calculation of benefits is ‘retrospective, daily’. 10
OFT Annexe E, Summary report of mystery shopping exercise, October 2006, paragraph 3.5. 11
Payment Protection Insurance Market Investigation Research Report, BMRB, February 2008, p15. 107
that purchase PPI which suggests that they are able to search more effectively than
a representative sample of the population.
5.43 A study undertaken in early 2006 by the FSA 12 looked at the financial capability of
consumers, including the ability to make informed choices about financial products.
The FSA found that, generally, people scored poorly on their ability to choose a
financial product. They often did not understand the risks they were exposing
themselves to, and did not use insurance in situations where it might benefit them to
do so. The FSA did find that people learnt from their experiences; in general, the
more products they had bought, the better they were at making appropriate choices.
Experience was in fact the strongest indicator of capability, and was far stronger than
income. 13
• The role of financial advice
5.44 In order to overcome this, some consumers choose to use financial advisers to help
them make an appropriate choice. However, for MPPI (the only product sold in any
volume by intermediaries, we found (paragraph 3.73) that intermediaries tend to offer
a static and narrow range of MPPI products, suggesting a limited amount of search
on the part of intermediaries.
• Conclusion on the ease of using information
5.45 We found that the variation in pricing structure, terms and conditions and the way
information is provided by firms mean that information is not available in a sufficiently
usable format that enough consumers successfully compare products to drive
competition between providers. This may restrict the ability of consumers to search
effectively and to make an informed choice, and we concluded that this is a
significant barrier to effective search.
Relationship with the credit provider
5.46 The third possible barrier to search we looked at was whether the relationship with
the credit provider was a barrier to searching effectively.
5.47 Customers value their existing relationships with financial providers; survey evidence
from the parties reveals that the existing relationship with a provider plays an
important role in the choice of lender.
5.48 We considered whether the link between credit and PPI was perceived as a barrier to
search. It was put to us that some consumers believe that search will impact on
subsequent credit applications, thus limiting their desire or motivation to search. 14 As
noted in paragraph 5.27, in order to get an accurate quote for some products,
12
The FSA study also considered whether, and how, people keep abreast of financial developments. The primary source of
information for people is newspapers generally (41 per cent), followed by radio or television programmes (39 per cent). Only
19 per cent of people keep up to date by reading the financial pages of newspapers; and only 7 per cent tune in to specialist
programmes on radio or television. The FSA said that the results indicated that most people absorb financial information while
reading, watching or listening to other things which interest them. The FSA found that there was a strong correlation between
keeping well informed about financial matters and both income and general levels of education; and men significantly
outperform women on staying informed, even when factors such as income were taken out.
www.fsa.gov.uk/pubs/other/fincap_baseline.pdf.
13
The FSA in its 2001 review of the mortgage buying process found that there was some evidence that many individuals and
households had only a limited involvement with long-term financial planning. This research is quite old; nonetheless it stated
that there was a small but growing amount of literature around the extent of financial literacy and there had been a number of
studies in the past indicating the problems people had with concepts integral to financial choices such as, for example, APR.
FSA Consultation paper 28, Comparative information for financial services, October 1999, paragraph 4.11, found that 40 per
cent of consumers said that they generally found it difficult to understand leaflets about financial products.
14
HBOS response to Emerging Thinking, p5.
108
customers must complete a full credit application, which is recorded at credit
bureaux. All other things being equal, financial institutions may consider more carefully applications for credit from consumers whose credit record indicates that they
have been applying for credit from a number of different providers in a relatively short
period of time, because that may be a sign of a consumer’s financial distress or of
fraud. We were told by RBSG that making multiple applications had less effect on the
likelihood that someone would be offered credit than it used to, but we recognized
that making multiple applications is still a factor taken into account. 15
5.49 Survey evidence did show that some consumers think that taking out PPI with the
credit will positively affect their chances of being granted credit, even though we are
not aware of any credit provider which requires a PPI policy as a condition of granting
credit. The CC BMRB 2007 survey reported that 25 per cent of MPPI consumers who
decided to take out PPI at the point of sale of their credit product believed that this
made the credit application more likely to be accepted (based on 440 responses).
Similarly, 29 per cent of CCPPI consumers (based on 576 responses), 41 per cent of
PLPPI consumers (based on 569 responses) and 47 per cent of SMPPI consumers
(based on 590 responses) believed this. The survey also showed that, of those
respondents who believed their credit application was more or less likely to be
accepted if they took out PPI, 16 per cent of MPPI consumers said that they were
specifically told by their credit provider that their application was dependent on their
purchase of PPI, and 80 per cent said they had that perception; 13 per cent of
SMPPI consumers said they were told this, and 82 per cent said they had that
perception. For PLPPI, 20 per cent of consumers said that they were told their credit
application was dependent upon taking PPI, and 75 per cent said they had that
perception. The figures for CCPPI consumers are 9 per cent and 90 per cent
respectively. 16 Details of further surveys we received from the parties on this can be
found in Appendix 5.1. Similarly, the OFT GfK NOP 2006 survey found that 27 per
cent of respondents thought that PPI take-up would help with their application for
credit, 17 and the FSA ICOB review 18 found that 44 per cent of respondents felt that a
loan application was more likely to be approved if they agreed to take out PPI at the
same time.
5.50 Whilst FSA evidence we received showed that people are being told that PPI is an
optional product, between one-quarter and a half of consumers (depending on the
product) still believe that PPI take-up will positively influence the credit application
process, and some believe that PPI is mandatory. We concluded that such beliefs,
allied to the importance of an existing relationship, limit search for PPI on a standalone basis.
Availability of PPI on a stand-alone basis
5.51 The final potential barrier to search we looked at was whether a lack of PPI available
on a stand-alone basis created a barrier to search on a stand-alone basis.
15
The proportions of PPI consumers interviewed in-home in our CC GfK NOP 2008 survey who indicated that they made multiple credit applications were 21 per cent of 76 MPPI consumers; 55 per cent of 64 SMPPI consumers; and 25 per cent of 60 PLPPI consumers. Neither our surveys nor the OFT survey asked questions about the perceived effect of multiple applications on the consumer’s creditworthiness. 16
Payment Protection Insurance Market Investigation Research Report, BMRB, February 2008, pp20&21. 17
Based on 952 respondents who had a PPI policy at the time of the survey: 68 per cent no; 19 per cent yes, I assumed; 4 per cent yes, hinted at by lender; 4 per cent yes, stated by lender; and 4 per cent don’t know. Data does not differ significantly
between SMPPI, MPPI, CCPPI and store card PPI OFT survey Q13. 18
FSA ICOB review Interim report: consumer experiences and outcomes in general insurance markets, Omnibus research, January 2007. 109
5.52 Large distributors tend not to offer PPI for sale independently of a credit product (as
noted in paragraph 3.36, the only PPI products sold at the point of sale which are
also available on a stand-alone basis are four MPPI products). The small size of the
stand-alone market limits the scope for search for stand-alone providers.
5.53 We also found that many consumers do not know they can buy PPI from anyone
other than the credit provider. Survey evidence we saw showed that between 31 and
65 per cent of consumers did not believe they could have gone to a different PPI
provider (see paragraph 3.47).
Conclusions on barriers to search
5.54 We concluded that there are some substantive barriers to searching for PPI:
(a) It is time-consuming to obtain accurate quotes when PPI is provided alongside
credit, with some firms only providing accurate price illustrations for PPI by going
through a full credit application.
(b) Price comparison websites, whilst useful, are limited in their offer of comparisons.
(c) PPI policies are complex. Variations in policy terms and conditions, and hence
the manner in which information is provided by firms, mean that the cost of PPI is
not presented in a way that is easy to make comparisons and have a detrimental
effect on consumers’ ability to understand that information. The pricing structure
of single-premium PPI is an additional source of complexity in relation to PLPPI
and SMPPI. The complexity of PPI policies restricts consumer search and the
ability of consumers to make informed choices.
(d) When PPI is sold alongside credit, some consumers perceive either that PPI
take-up will have a positive influence on their credit application process or that it
is a condition of taking the credit.
(e) The relatively low level of stand-alone provision restricts consumers’ ability to
search for PPI on a stand-alone basis (although it is still possible to search for the
bundled product).
5.55 We thought that all these barriers were present when a customer is taking out a
personal loan, mortgage or secured loan and deciding whether to take out PLPPI,
MPPI or SMPPI (respectively) from the credit provider or from someone else. We
thought that all these barriers, except for barrier (a), were present when a customer
is taking out a credit card and deciding whether to take out CCPPI from the credit
provider or from someone else.
Conclusions on search
5.56 We found that there are benefits for consumers in searching, but that customer
search is not sufficiently widespread to constrain prices to a competitive level. We
found that there are some significant barriers to search.
5.57 We concluded that those consumers who did want to compare products were
hindered by product complexity (both the variations in pricing structures and terms
and conditions, and the way information on PPI is presented to consumers); the
perception that taking PPI would increase their chances of being given credit; the
bundling of PPI and credit; and the relatively low level of stand-alone provision. We
conclude that these obstacles to allowing consumers to search effectively act as
110
barriers to search for all types of PPI policies. In addition, the time taken to obtain
accurate price information is a barrier in relation to the provision of PLPPI, MPPI and
SMPPI. These barriers to search impede the ability of consumers to make comparisons and therefore effective choices between PPI policies. They also, therefore,
act as barriers to expansion for other PPI providers, in particular providers of standalone PPI. These barriers to effective search are features of relevant markets which
prevent, restrict or distort competition in the supply of PPI. If there were a greater
degree of product comparison being carried out by consumers, distributors would
have a stronger incentive to compete with each other to show consumers that their
product offer was the best value and this would also provide greater opportunities for
stand-alone providers to compete.
Switching
Barriers to switching
5.58 We considered whether there were any barriers to consumers switching, in addition
to the search barriers we have previously identified. We looked at: providers’ access
to relevant information; initial exclusion periods; pre-existing condition exclusions. We
then looked at the rebate structures for single-premium policies to see if this acted as
a barrier to switching.
Providers’ access to relevant information
5.59 For CCPPI we thought that access to consumers’ credit information restricted the
extent to which customers switch CCPPI on a stand-alone basis. Although standalone CCPPI provision exists, these products are not linked to credit balances.
Customers choose a level of insurance, which means that the level of insurance is
not tied to the actual amount owing on a card, and hence the customer is more likely
to be over- or under-insured in any given month than insured for the exact amount
owing (and hence the coverage is of lower quality than one which tracks the balance
exactly). We considered this to be a barrier to switching to stand-alone CCPPI
policies.
5.60 We noted that it is very easy physically to switch future credit needs to CCPPI
policies associated with other credit cards because customers can transfer balances
between cards. We also noted that outstanding balances can be switched to new
credit cards, with associated CCPPI, taking advantage of special balance transfer
offers.
Initial exclusion periods
5.61 We noted that, for all types of PPI, the initial exclusion period at the start of a policy
before a consumer can claim limited switching. We received evidence from CML, 19
who told us that customers might be deterred from switching MPPI because it could
leave them without cover for a 60-day period. Genworth 20 also highlighted this as a
potential barrier. Further, the Post Office 21 said that non-coverage (particularly on
unemployment) when switching PPI policies acted as a deterrent to switching. Its
Lifestyle product had a qualification period for an unemployment claim to combat the
risk of anti-selection. We recognized that for stand-alone policies a customer could
19
CML response to issues statement. Genworth said that the exclusion period at the start of a policy might act as a barrier to switching. Summary of post-issues statement hearing with Genworth.
21
Summary of post-issues statement hearing with the Post Office. 20
111
overlap policies, retaining the old policy until the exclusion period on the new policy
had expired, but doing so incurred an additional cost of having two policies running in
parallel for a time. The evidence led us to the view that, for all types of PPI, the initial
exclusion period can act as a deterrent to switching, as customers are not fully
covered during the initial period of a new policy.
Exclusions for pre-existing conditions
5.62 Similarly, we thought that the exclusions in policies for pre-existing conditions acted
as a barrier to switching. We received evidence from AXA, Cardif Pinnacle 22 and
CML 23 that, if a consumer switched PPI, any medical conditions they had developed
since they initially purchased PPI would become pre-existing under the new policy,
and therefore might not be covered, which might deter switching. We agreed that, for
all types of PPI, for someone who has developed a condition which would be
excluded from coverage as a pre-existing condition under a new policy, the existence
of exclusions for pre-existing conditions constitutes a barrier to switching PPI
policies.
The effect of single-premium policies on switching
5.63 As noted in paragraphs 2.77 to 2.78, PPI can be paid for with a single premium or a
regular premium. Nearly all PLPPI policies and about two-thirds of SMPPI policies
are paid for by a single premium, paid up front (although, as noted in paragraph
10.247, towards the end of our inquiry several parties took the decision to stop selling
single-premium PLPPI policies for reasons unrelated to our inquiry). 24 The barriers to
switching we have considered so far apply equally to policies sold with regular
premiums and single premiums. We considered whether there were any barriers to
switching associated with PPI policies being sold as single-premium policies. Our full
analysis of this is set out in Appendix 5.2. Our purpose in looking at single-premium
policies, and in particular the refund paid if a policy is cancelled before term, was not
to take a view on the fairness of the rebate offered. The purpose was to see whether
the size of the rebate was a barrier to switching.
5.64 As set out in paragraphs 2.79 and 2.80 and Appendix 2.10, we were told that single
premium policies presented advantages for customers. We also received evidence
that single-premium policies presented advantages for distributors, including
significantly greater income, and that settlement of insurance is more likely with
monthly-premium policies (for example, we saw an internal document from one
provider which noted that a key feature of a proposed regular-premium policy was
‘regular premium so easy to switch with no refund issue’).
5.65 We received mixed views on how easy it is to switch away from a single-premium
policy; some parties told us that it was easier to switch away from a regular-premium
policy than from a single-premium policy, whilst others disagreed.
5.66 We noted that single-premium policies do not in practice prevent customers settling
their loans with PPI ahead of term; we were, however, unable to separate out those
who were switching to another provider from those who were paying off their debt.
22
Summary of post-issues statement hearing with Cardif Pinnacle. CML told us that customers might be deterred from purchasing MPPI because they might not be covered for pre-existing conditions. CML response to issues statement. 24
Based on an analysis of policies sold by the 12 largest distributors of PPI, 93.1 per cent of PLPPI policies sold are paid for with a single premium; 64.6 per cent of SMPPI policies are single-premium policies.
23
112
5.67 We noted that rebates are now given on all single-premium policies. Customers do
not get a refund when they cancel a premium paid on a monthly basis). We found
that the largest distributors calculated refunds based on the ‘rule of 78’ (see
Figure 5.4 for the profile of the rebate given if a policy is cancelled) or an actuarial
method which produces a similar refund profile. We have no reason to think that
other distributors use significantly different refund profiles. We were told that the rule
of 78 reflected the risk taken on by the lender—there is a higher risk at the start of a
policy reflecting the greater amount owing at that time. We also received evidence
that most claims are made in the first year or 18 months of a policy’s life, and this
was reflected in the rebates offered by those companies which had developed their
own actuarial rebate profiles.
FIGURE 5.4
Profile of rebate for a three-year loan, calculated using the rule of 78
100
Rebate as % of initial premium
90
80
70
60
50
40
30
20
10
0
1
3
5
7
9
11
13
15 17
19
21
23 25
27
29
31 33
35
Months elapsed
Source: CC.
5.68 We looked at whether the costs associated with cancelling a single-premium policy
were different from those associated with cancelling a regular-premium policy.
5.69 In terms of time costs, we found that it can take longer to cancel single-premium
policies than regular-premium policies. However, some regular-premium policies
have notice periods, and overall we concluded that there was not a substantive time
difference either way; however, our conclusion on whether single-premium policies
are a barrier to switching did not rely on this assessment.
5.70 We also looked at the financial cost of switching. We wanted to understand whether
the rebate from a single-premium policy was sufficient to take out a comparable
replacement (either regular-premium or single-premium) policy to cover the
remaining credit payments. Our analysis of this is set out in Appendix 5.2.
5.71 In order to estimate a ‘pure’ switching cost, we needed to be able to model switching
between identical policies. As we were not aware of policies offered by different
providers with identical terms and conditions, we looked at the rebate received when
a policy is cancelled and then how much it would cost to take out the identical policy
from the same provider the same day, as either a single-premium policy or as a
hypothetical regular-premium policy (where the total cost of the insurance over the
remaining life of the loan is divided into equal monthly amounts). Whilst we
113
recognized that this is an artificial situation, it did allow us, for each of the largest
distributors of single-premium PPI, to see the difference between the rebate provided
and the amount required to insure the remaining credit repayments with exactly the
same level of cover. We thought, therefore, that this was a reasonable way of
estimating switching costs in the absence of real-life identical policies.
5.72 We found that the cost of switching in each case followed an inverted ‘u’-shape when
plotted against the duration of the loan, and that at all times there was a positive
switching cost—that is, the rebate received upon cancelling a single-premium policy
is insufficient to take out an identical PPI policy for the remainder of the credit
repayments. Given the shape of the rebate curve (shown in Figure 5.4), this was to
be expected.
5.73 The level of shortfall varied over time and between providers, with the maximum
shortfall occurring usually up to halfway through the loan period. The cost of
switching was also higher for those that charged an administration fee for cancelling
the policy. When looking at switching to a hypothetical regular-premium policy, the
maximum switching cost varied by distributor and it was between £146.91 and
£429.77 for a loan of £5,000 over 36 months (the maximum switching cost for nine of
the providers we looked at varied between 53 and 79.5 per cent of the premium paid
for the replacement regular-premium policy). There was one exception to this: the
maximum switching cost for one provider was significantly higher (over £700). We
thought that the main driver for this was the significantly higher APR charged on
credit by that provider, as its offer was to sub-prime customers. Our analysis of
switching to an identical single-premium policy was based on more limited data; this
element of the switching costs calculated varied between £81 and £405.
5.74 We recognized that the rebate profiles used by distributors are based on a wellknown rule or on actuarial rules which reflected risk. We have not taken a view on the
fairness of rebates offered; our focus is on the effects of actual rebates on competition. We found that the effect of using these rebate profiles for single-premium
policies is that there is a switching cost associated with cancelling the PPI policy and
switching to an alternative policy. That switching cost can be large, amounting to a
significant percentage of the cost of purchasing an identical replacement policy.
5.75 We concluded that the rebate profiles used for single-premium policies were
significant barriers to switching.
Conclusions on switching
5.76 In addition to the costs of search already found (paragraph 5.54), we conclude that
there are four substantive barriers to switching:
(a) For CCPPI, the inability of stand-alone providers to offer insurance which tracks
the outstanding balance on a credit card means that CCPPI policy holders cannot
find a perfect substitute for their policy; instead they can only find a policy which
insures a set amount, leading them to be over- or under-insured in any given
month.
(b) Initial exclusion periods on PPI policies, which mean that someone switching is
not insured against certain eventualities at the start of the new policy.
(c) Exclusions for pre-existing conditions, though only for those customers who have
developed conditions which are exempt from cover after they took out their
original PPI policy but before switching.
114
(d) The rebate profiles used for calculating rebates when cancelling a singlepremium policy. Whilst the profiles are based on well-established principles or
actuarial methods based on claims experience, the effect is that customers
receive often significantly less than they need to purchase identical replacement
cover.
5.77 We found that barriers (b), (c) and (d) applied to switching PLPPI and SMPPI
policies, barriers (b) and (c) applied to switching MPPI policies, and barriers (a), (b)
and (c) applied to switching of CCPPI policies.
5.78 These barriers to switching limit consumer choice. They also, therefore act as
barriers to expansion for other PPI providers, in particular providers of stand-alone
PPI. We conclude that the existence of these barriers to switching between PPI
policies is a feature of relevant markets which prevents, restricts or distorts
competition in the supply of PPI.
Barriers to entry and expansion
5.79 We considered whether there are additional barriers to entry or expansion for either
suppliers selling stand-alone PPI, or PPI alongside credit. Details of our analysis of
all the possible barriers are included in Appendix 5.3.
5.80 We looked first at the sale of stand-alone PPI and short-term IP sold away from the
credit point of sale (paragraphs 5.81 to 5.85) and then sales of PPI alongside credit
(paragraph 5.86).
5.81 Four large underwriters we spoke to raised with us problems associated with adverse
selection (see paragraphs 20 to 24 of Appendix 5.3). 25 Evidence from Aviva on its
experience of selling stand-alone MPPI between 2002 and 2003 showed that it paid
out over [] more in claims and expenses than it received in premiums. A business
sustaining such losses would not normally be considered to be viable over a
prolonged period. Equivalent figures for the main party insurers from whom we
collected data show payout rates of claims and expenses (ie after distributors have
received their share of premium as commission) averaging around 90 per cent—that
is, paying out in claims and expenses less than the premiums received.
5.82 The evidence provided suggested that stand-alone PPI providers suffered from an
increased risk of adverse selection, because consumers who search for this type of
insurance are more likely to make a claim than customers who purchase PPI when
applying for credit. We thought this was exacerbated by PPI being sold alongside
credit—we thought that PPI when sold with credit may experience some reduced risk
of adverse selection as distributors will have access to all credit purchasers to try and
sell PPI to rather than just those who approach them for PPI, and because a
consumer who is facing a potential loss of income may be less likely to take on a new
credit product. We concluded that adverse selection was a further barrier to entry and
expansion faced by stand-alone PPI providers whilst distributors sell PPI alongside
credit.
5.83 We considered whether customer awareness of PPI and the options available
constituted a barrier to entry and expansion. As noted in Appendix 5.3, we were told
by some parties that customers generally had a low awareness of the offers of
25
Adverse selection, sometimes referred to as anti-selection, occurs as a result of information asymmetries between the insurer
and the customer. Customers who know they face a large risk of needing to make a claim on an insurance policy are more
likely to take out insurance than those who face small risks; however, the insurer does not know which customers fall into which
category and so cannot price its insurance premium accordingly.
115
different credit providers. The CC GfK NOP 2008 survey on search behaviour found
that 48 per cent of PLPPI customers (based on 3,156 responses), 49 per cent of
SMPPI customers (based on 2,387 responses) and 61 per cent of MPPI customers
(based on 1,257 responses) had considered taking out insurance before approaching
the lender. 26 Moreover, the same survey found that many PPI customers who did not
compare products from different providers believed that they could only buy the PPI
product from their credit provider—see paragraphs 3.46 and 3.47. We concluded on
the basis of the evidence that low customer awareness of PPI, and poor
understanding of the options available, restrict the ability of stand-alone providers
successfully to enter and expand into PPI markets.
5.84 In order to overcome the poor awareness of the availability of stand-alone options,
providers must market their products. The evidence we received on marketing costs
to attract customers for a stand-alone policy was that they were significant. Most
notably, evidence from Aviva was that the cost of acquiring customers for a flexible
mortgage PPI product, even when cross-selling the product to existing customers of
household and motor insurance, was high. Each customer gained cost £[0–200], 27
[]. We concluded that, with the current market structure and at current prices, the
cost of marketing stand-alone PPI to consumers is large if PPI products are not sold
in combination with a credit product.
5.85 We noted that providers of stand-alone PPI and short-term IP have not achieved
significant sales, and that some large underwriters (Axa, Aviva) have exited the
stand-alone market as a result of their experience. As noted in paragraph 3.43, there
has been some new entry recently. However, we concluded in paragraph 5.134 that
we do not expect these new entrants to exert a competitive constraint in the short to
medium term.
5.86 We then considered whether there are barriers to entry and expansion for companies
wishing to sell PPI alongside credit. We did not receive any evidence that it was
difficult for an existing supplier of credit to start distributing PPI along with its credit
products. We found only three credit suppliers which did not also sell PPI, one of
which (Nationwide) had only temporarily put on hold its sales of PLPPI and CCPPI.
As noted in paragraph 5.82, at the point of sale a distributor is more likely to experience a positive selection bias than adverse selection, and by talking to customers
during the credit sale (either on an advised or non-advised basis) distributors are
able to overcome the lack of customer awareness without incurring high marketing
costs. We did not, therefore, think that there were any substantive barriers to entry or
expansion for companies wishing to sell PPI alongside credit.
5.87 We conclude that there are significant barriers to entry and expansion for stand-alone
PPI providers seeking to sell PPI products without access to customers at the credit
point of sale, due to adverse selection, poor consumer awareness and high
marketing costs. These issues are discussed in more depth in the context of the
point-of-sale advantage (paragraphs 5.88 to 5.119).
The point-of-sale advantage
5.88 Most parties told us that there was an advantage associated with selling PPI at the
point of sale of credit—though there were markedly different views on how important
26
PPI search behaviour: a research report for the Competition Commission, GfK NOP, April 2008, p10. Capital One thought that if it had to offer a stand-alone PPI product to new customers, the costs of reaching them would make selling only PPI unviable, as generating significant volumes of new accounts using direct marketing strategies could be very
expensive; in some cases the cost of acquiring a new customer could be greater than £[0–200] per customer. 27
116
that advantage was. We therefore considered whether there was an advantage for
distributors in selling PPI alongside credit.
What is the point-of-sale advantage
5.89 The point-of-sale advantage is the benefit which distributors enjoy from having sole
access to customers at the time that they are buying their credit.
5.90 It has been put to us that the point of sale is the time when customers are likely to
focus their attention on the needs which PPI is designed to meet, 28 and therefore it is
an opportune time for distributors to attempt to sell PPI to customers.
5.91 Given this and the factors discussed in paragraphs 5.100 to 5.105, the ability to offer
their PPI at the point of sale (either on an advised or non-advised basis), and be the
only provider to do so, gives distributors an advantage over those who cannot do so.
Evidence for the existence of a point-of-sale advantage
5.92 We considered whether there is evidence of a point-of-sale advantage.
5.93 The CC GfK NOP 2008 survey 29 found that around one-half to three-quarters of
customers had not considered PPI before approaching their lender. In fact, 49 per
cent of MPPI holders had not considered PPI before the point of sale, and the figures
for SMPPI, PLPPI and CCPPI were 66, 70 and 76 per cent respectively. Whilst some
parties told us that at least some customers considered PPI before the point of sale,
and indeed our survey showed evidence of this, we received evidence from several
parties that customers do not generally consider PPI prior to the point of sale. 30
5.94 Another piece of evidence of the existence of a point-of-sale advantage is that some
people think that the chances of their credit application being approved will be
enhanced by taking out PPI along with the credit (see paragraph 5.49).
5.95 We also saw evidence that a large number of PPI holders believed that they could
not have bought PPI from a provider other than their credit provider. The CC GfK
NOP 2008 survey found that 31 per cent of MPPI holders, 61 per cent of SMPPI
28
Responses to Emerging Thinking from ABI, Cattles, HBOS, RBSG.
CC, Payment Protection Insurance Market Investigation Research Project Report, February 2007, prepared by BMRB.
30
For example, AXA told us that it was unlikely that customers would have considered purchasing PPI before reaching the point
of sale of the credit product; and Cattles said that customers tended not to give thought to the risk of borrowing or what they
would do if they were not able to meet loan repayments in the event of accident, sickness, unemployment or death, except at
the time that they took the decision to purchase a credit product. Abbey told us that the majority of the public were not well
versed in PPI before they made an application for credit; Aviva told us that most people did not think about making provision for
an unexpected drop in income until they were asked to confront the risk when they applied for a loan; British Insurance told us
that when consumers decided to apply for credit they were focused on securing that credit, they were not thinking about
insuring themselves; HBOS said that the point of sale was the only time that consumers had a high likelihood of identifying and
doing something about their income protection needs (though it said that this did not support a conclusion that customers do
not consider PPI prior to the point of sale, and noted there was a distinction between identifying a need for insurance and taking
the opportunity to fulfil that need). Lloyds TSB said that the majority of customers did not engage with PPI until they had made
a choice based on loan APR because that was where the competition was. The PPI was introduced when the lender sat down
with the customer to discuss what cover they might need; MBNA told us that the driver for customers to contact them was the
credit product, not the PPI product. The point of sale of the credit product provided an opportunity for MBNA to talk to
customers about a service (such as PPI) which it could provide to them. However, MBNA also said that less than half of
customers took out CCPPI at the point of sale of the credit card. In its hearing, HSBC said that it was ‘… essential that PPI can
be offered and sold at the same time that the customer is purchasing the credit product. Does this give the seller of the credit
product some advantage in respect of the sale of PPI? Yes it does. But not an advantage which is overwhelming. The low
take-up of PPI compared to other general insurance products demonstrates that customers exercise choice at the point of sale’.
HSBC noted, however, that customers could have considered PPI before the point of sale, but might not purchase PPI if it was
not sold at that time.
29
117
holders, 60 per cent of PLPPI holders and 65 per cent of CCPPI believed this to be
the case.
5.96 The FSA said that many firms did not make it clear to the mystery shopper that the
loan product and the PPI product were separate, such that the mystery shoppers
were left with the impression that the PPI ‘comes with the loan’. The FSA conducted
a mystery shopping exercise which showed that:
• 77 per cent of advisers discussed the PPI product with the customer before the
credit was offered;
• 40 per cent of customers were not told that the loan and the PPI were separate;
and
• 55 per cent were given the monthly figure including PPI as the first quote.
5.97
[]
5.98 Despite this evidence (both of actual sales and of customer perceptions about the
role PPI plays in credit applications) pointing to a point-of-sale advantage, the
providers’ ability to encourage customers to take PPI does appear to vary with sales
channel. The penetration of PPI policies sold over the Internet is significantly lower
than the penetration of PPI policies sold face to face at a branch or over the
telephone (see Appendix 5.3). This evidence suggests that credit customers are
significantly more likely to buy PPI if they are sold it by a salesperson (face to face or
over the telephone) than if they are left to decide whether to purchase it without
prompting. 31
The causes of the point-of-sale advantage
5.99 We looked at a range of possible causes of the point-of-sale advantage.
5.100 The evidence suggested to us that customer behaviour plays a part. As noted in
paragraph 5.93, many customers did not consider PPI before arriving at the credit
point of sale, some thought that taking out PPI would enhance the chances of being
granted credit, some thought it was a condition of taking out the credit, and many
thought that they could only obtain PPI from their credit provider.
5.101 We also noted that searching for information about PPI is not easy, primarily because
there is limited information available prior to the point of sale. As noted in paragraphs
5.39 and 5.40, PPI policies are complex, with many different terms and conditions; as
noted in paragraph 4.31, price is not a good indicator of product quality; and, as
noted in paragraph 3.118 and Appendix 3.8, while there are some Internet
comparison websites, search facilities are inevitably limited and the evidence
suggested that they are only used to a limited extent to search for PPI.
5.102 Furthermore, distributors appear to focus their advertising on credit products. While
the availability of PPI is included in some credit literature, or is advertised in leaflets
31
That is not to say that distributors do not attempt to point out to customers the risks associated with not taking out PPI, even if
there is no person-to-person contact. For example, on 30 May 2008 the NatWest website personal loan quick quote calculator
provided quotes for a loan with PPI first. If the customer clicked on the tab to see the quote without PPI, the customer saw the
following text above the quotes: ‘These quotes do not include repayment protection. An estimated 30 million working days were
lost in 2006 due to work-related ill health and work place injury (Source: Health & Safety Executive). Our Personal Loan
Protector could offer the cover you’re looking for if you are unable to work through accident, sickness or redundancy.’
118
in branches or on websites, there is little specific advertising of the PPI products prior
to the point of sale.
5.103 The lack of advertising prior to the point of sale, allied to customers’ lack of proactive
search prior to the point of sale (as noted in paragraph 5.93), means that the majority
of consumers who purchase their PPI at the point of sale are reliant upon the
information which is provided to them by the distributor at that time.
5.104 It is also significant that no distributors offer competing PPI products at the point of
sale (though some mortgage intermediaries offer a choice of policies from different
underwriters), which restricts customer choice at that point when they are most likely
to be focused on their insurance needs.
5.105 Finally, we noted that many distributors sell PPI on an advised basis. This is
permitted by the FSA and is entirely legitimate. However, distributors advise on the
suitability of their PPI product to meet a customer’s needs. They do not (and are not
required to) advise on whether there are PPI policies sold by others which might
better meet that customer’s needs, whether there are alternative products which
might better meet that customer’s needs (for example, life insurance), or whether the
product offers good value for money for a customer. It is also the case that such
alternative policies and products are not offered to a customer when the sale is on a
non-advised basis. We do not suggest that PPI distributors are in any way breaching
FSA rules by doing any of the above. However, customers will not always buy the
right product for them if they rely on the limited offering available at the point of sale,
or rely on the advised sale process to deliver the best product for them.
The advantages gained by distributors from selling PPI at the point of sale
5.106 We considered the ways in which distributors gain an advantage from selling PPI at
the point of sale.
5.107 First, we looked at whether there are economies of scope in the provision of both
credit and PPI. Distributors benefit from shared facilities when selling both products
at the point of sale and also from reduced marketing costs, suggesting that there are
some economies of scope to be enjoyed:
• There are cost efficiencies associated with an integrated sales process (for
example, the information which distributors need from a customer before providing
a PPI policy is much the same as is required when making an application for
credit, and only has to be obtained once). We heard evidence from one distributor
([]) who told us that, for the sales element of channel costs, according to a
predefined sales script, the time spent on selling PPI and the credit through a
telephone contact centre could be split one-third PPI, one-third credit and onethird time common to both products. It also said that in theory this split could also
be applied to a branch sales process even though this was not a predefined sales
script. There are no additional customer acquisition costs associated with selling
PPI that are not incurred when acquiring the credit customer because, by virtue of
making an application for credit, the customer has presented themselves to the
PPI distributor as a potential consumer of PPI.
• There are cost efficiencies in marketing, as credit suppliers use their credit
advertising to attract customers, to whom they try to sell both credit and PPI. If
they did not sell PPI, they would probably need to engage in the same level of
advertising just to gain the credit customers. In addition, as discussed further
119
below, any marketing of PPI seems to be much more effective when delivered at
the point of sale of the credit product than at any other time.
• The PPI policy can more easily be tailored to the credit product—something which
is particularly useful for CCPPI, as only those policies sold by the credit provider
can track the outstanding balance on a monthly basis).
• When PPI purchases are associated with the purchase of a credit product, there is
a reduced risk of adverse selection compared with when a policy is sold on a
stand-alone basis (see paragraph 5.82). Advertising prior to the point of sale could
indeed increase adverse selection. In fact, many PPI policies are only offered for
purchase by the credit supplier at the point of sale; so customers who do not take
that opportunity to buy PPI will not be able to purchase it at all from their credit
supplier, or will not proactively be offered it by the distributor. For example, MBNA
told us that, when it offered personal loans, its PLPPI policy could only be bought
at the beginning of the loan, it could not be added at a later date; RBSG said that
if a customer decided to buy PPI post-loan purchase, the loan would have to be
restructured to incorporate the PPI element into it; HSBC said that, before it
changed to offering LifeChoices to credit customers (see Appendix 2.3), it did not
follow up personal loan customers with PPI marketing after the point of sale
(though it did with credit card and retail PPI); Cattles told us that it did not try to
sell PPI to customers other than at the point of sale. We found that this further
reduces the risk of adverse selection for distributors.
5.108 Bearing in mind the fully-allocated cost we estimated of selling a PPI policy relative to
the average revenue earned on a PPI policy (see paragraphs 29 and 54 of Appendix
4.4), we thought that the economies of scope from the integrated sales process and
shared marketing costs might be relatively small.
5.109 In addition to the economies of scope discussed above, the PPI distributor has the
crucial advantage of access to the customer at the optimal time for making a sale.
Unless the customer has actively researched PPI in advance, the credit provider will
have first access to the customer to sell the benefits of PPI. As noted in paragraph
5.107, some only offer PPI at the point of sale, or only proactively offer it then, and if
a customer has not compared PPI policies before arriving at the point of sale there is
an opportunity cost—meaning that customers who have not researched PPI
beforehand can feel that it is ‘now or never’ for this PPI policy, and may feel it safest
to buy it. Furthermore, we found in paragraphs 3.46 and 3.47 that many customers
thought it could be obtained only from the credit distributor, and we also found that
few customers who did compare policies revisited their purchase decisions during the
cooling-off period. 32
5.110 In contrast, stand-alone providers, even if they are able to identify holders of credit
products (for example, through current account relationships), are likely to be offering
PPI to a customer who already has or has been offered PPI by the credit provider
and declined it.
5.111 Moreover, in so far as there can be an advantage following the point at which the
customer enters into the credit agreement, such advantage sits with the credit
provider. Once the distributor has established its relationship with the customers to
whom it has provided credit, it has continued access to the customer to communicate
the availability of PPI (for example, when it sends out a statement each month).
32
The CC GfK NOP 2008 survey indicated that between one-quarter and one-third of PPI customers review their PPI policies
during the cooling-off period. It should be noted that this finding is based on a small number of in-home interviews (76 MPPI,
64 SMPPI and 60 PLPPI customers).
120
The advantages gained by customers from PPI being sold at the point of sale
5.112 It was put to us that selling PPI at the point of sale is beneficial to customers, for a
number of reasons, such as:
(a) If the cost advantage which distributors enjoy as a result of selling the credit and
PPI at the point of sale is passed on to consumers, this should translate to a
customer benefit in the form of lower prices.
(b) Survey evidence showed that customers cite ‘peace of mind’ as being the
primary reason they purchase PPI. Given that over 95 per cent of the PPI sold is
sold at the point of sale, if customers were not able to buy it at that time, they
might not be aware of it at all, and thereby the opportunity for that peace of mind
would not be available to them.
(c) Customers’ perception of need and risk seem to change at different points in the
credit application process, therefore the point of sale might be of psychological
importance to customers.
(d) Customers buying PPI at the point of sale do not have to go through another
interview to buy insurance to cover the credit.
5.113 We noted these possible benefits. We recognized that (d) was a benefit from buying
at the point of sale and that there may be a psychological importance for customers
in buying insurance at the same time as credit (the CC’s GfK NOP 2008 quantitative
survey found that, for those consumers who had not shopped around for their PPI,
the main barriers to shopping were the convenience of buying credit and PPI at the
same time from the same provider, their preference for dealing with a company that
they already knew, and a lack of time to shop around). 33
5.114 We agreed that, in principle, the cost advantage distributors enjoy could translate to
lower prices for consumers. However, our analysis of profitability of PPI (paragraphs
4.60 to 4.82) suggested that PPI is highly profitable for distributors, and, as such, we
did not think it likely that customers were paying low prices for PPI. We also agreed
that ‘peace of mind’ is the most cited reason for taking out PPI; however, it was not
clear to us that consumers’ peace-of-mind needs could be met only through sale of
PPI at the point of sale.
5.115 Overall, the evidence led us to conclude that distributors gain significant advantages
from selling PPI at the credit point of sale. We also thought that customers gain
advantages from buying PPI at the point of sale. However, we did not think that all of
the possible benefits put to us were felt by customers in practice (in particular, lower
prices); also we were not clear that customers could not still achieve some of the
benefits away from the point of sale.
How the point of sale affects competition
5.116 The evidence led us to conclude that access to customers at the point of sale of
credit gives distributors and intermediaries an advantage over companies that do not
have access to the customers at that time. The key advantage is that distributors and
intermediaries can advertise their PPI offer to customers when they are most
susceptible to the advertising, whilst other PPI providers are unable to. The sale of
33
PPI search behaviour: a research report for the Competition Commission, GfK NOP, April 2008, p22.
121
PPI at point of sale of credit therefore acts as a barrier to other PPI providers
competing for customers of any given distributor or intermediary.
5.117 The magnitude of that advantage varies according to whether the point of sale
involves the customer meeting a salesperson, speaking to them on the telephone, or
does not involve contact with a salesperson, 34 and whether the sale is conducted on
an advised basis or a non-advised basis. 35 These factors are, however, largely
controlled by the distributor or intermediary who can choose to change from selling
on a non-advised to basis to an advised basis (or vice versa) and can change the
focus of its selling to maximize the use of salespeople and enhance PPI uptake.
5.118 The result of distributors and intermediaries selling PPI at the point of sale is that
sales of PPI by companies without access to customers at the point of sale are very
small, and, as noted in paragraph 5.87, there are substantial barriers to companies
successfully entering and expanding into the stand-alone market.
5.119 The evidence led us to conclude that the sale of PPI by distributors and intermediaries at the credit point of sale further restricts the extent to which other PPI
providers can compete effectively. We concluded that this is a significant restriction
and is therefore a feature of relevant markets which prevents, restricts and distorts
competition in the supply of PPI. In this respect we noted that there are some
economies of scope associated with selling PPI at the point of sale of credit;
however, it did not appear that these were passed on to customers in the form of
lower prices (see paragraphs 10.380 to 10.390.
The possible competitive constraint from new stand-alone products
5.120 In paragraphs 3.40 to 3.42 we found that sales of stand-alone PPI and short-term IP
were small and do not currently act as a competitive constraint on distributors or
intermediaries selling any PPI product at the point of sale, and are therefore not
within the same market as PPI sold at the point of sale. As noted in paragraph 3.38,
some new products have come on to the market since the start of this investigation,
and it was put to us that these products were imposing an increasing constraint on
PPI sold at the point of sale. 36
5.121 We considered, therefore, the extent to which we thought new products would
constrain sales of PPI at the point of sale over the short to medium term. We focused
on four products which have been brought to our attention: Barclays’ Plan B, which
offers Barclaycard holders the chance to insure other credit cards; and three policies
which enable customers to specify the level of income they would receive from the
insurance policy in the event of a claim—HSBC’s LifeChoices short-term IP product,
Churchill’s IP policy, and the Post Office’s Lifestyle Protection policy.
5.122 The Barclays Plan B product is targeted to both existing and new customers.
However, new customers need to take a Barclaycard in order to purchase the
Barclaycard CCPPI. The product was launched in September 2007 and Barclays had
sold approximately [0–2,500] policies covering cards other than a Barclaycard by
mid-2008.
34
EFS told us that sales over the Internet had a lower penetration rate, [] (see summary of post-Emerging Thinking hearing with EFS). [] told us that its highest penetration rate was achieved in branches [].
See paragraph 2.68. 36
As shown in Table 3.3, one existing stand-alone provider ([]) has seen very significant growth in its PPI sales during 2007 and the first half of 2008. However, we found in Section 3 that this was not sufficient to exercise a competitive constraint. 35
122
5.123 HSBC’s LifeChoices product initially offered 12 (and currently offers 14) different
combinations of cover based around accident, sickness, unemployment, life and
critical illness cover (see Appendix 2.3 for further details of the offer and the
background to the product launch). Between January and June 2008, HSBC sold
[] LifeChoices policies with some element of A, S or U cover, of which [] were
sold within 120 days of the sale of a credit product.
5.124 The Post Office launched its Lifestyle Protection product in June 2006. The rationale
behind the product was to produce a stand-alone proposition that was both more
value for money compared with traditional PPI and a more flexible alternative in that
the cover could either be tailored to credit commitments or provide financial certainty
even where there were no credit commitments. Research conducted by AXA showed
that direct switching from PPI to Lifestyle protection might occur with the launch of its
new product, which suggested that it was targeting existing PPI customers. The Post
Office sold [] Lifestyle Protection policies in 2007.
5.125 RBSG launched its Churchill IP product in May 2007. The target market comprised
[]. The target age group was [] although the product would be mass marketed.
As at February 2008 there were [] live policies.
5.126 It was put to us 37 that if large banks such as Barclays, HSBC and RBSG were
introducing new stand-alone policies, they must be confident of success, and we
should take account of that in considering the extent to which these policies can be
expected to constrain PPI sold at the point of sale.
5.127 As noted in Table 3.3, sales to date of these new policies have been extremely small.
This is perhaps unsurprising as they have only recently been launched and have had
limited exposure to customers. We were clear that at this point in time they do not
constrain PPI policies sold at the point of sale.
5.128 We considered that it was relevant to look at how others had fared in selling PPI on a
stand-alone basis, to give us a feel for how sales of these products might develop
over time. We received evidence from three large underwriters that had had
experience of selling PPI on a stand-alone basis.
5.129 We received evidence from Aviva regarding its sales of stand-alone MPPI through its
Norwich Union brand between 2002 and 2004. Between 2002 and 2004, Aviva sold a
total of [fewer than 10,000] PPI policies on a stand-alone basis. We noted that Aviva
had been unable to create a sustainable business—see paragraph 5.81—and had
withdrawn from the market in 2004.
5.130 AXA also sold a very limited number of stand-alone PPI policies. It sold [fewer than
500] policies of its AXA Direct Creditor product between 2002 and 2004, when it
withdrew from the market.
5.131 We also noted that Cardif Pinnacle had not achieved significant sales. It sold
[0–2,500] stand-alone MPPI policies in 2007, and at the end of 2007 had a total of
[10,000–15,000] MPPI policies in force (with a GWP of £2.9 million).
5.132 We noted that Barclays and HSBC products in particular might benefit from crossselling opportunities to customers of those banks; however, we thought that the large
underwriters had similarly had large customer bases to whom they could cross-sell
products, but had failed to achieve significant sales.
37
Barclays response to Emerging Thinking (LECG report), paragraphs 2.10–2.18.
123
5.133 This evidence indicated to us that there were significant difficulties associated with
creating a successful stand-alone policy in the current market structure. The failure of
the largest underwriter in the country, and two other significant underwriters, successfully to develop stand-alone policies suggested to us that the success of new
products was in no way assured. We noted that one existing stand-alone provider
([]) has seen very significant growth in its PPI sales during 2007 and the first half
of 2008, though stand-alone sales still make up a small fraction of all PPI sales.
However, we saw nothing to make us believe that point-of-sale providers have
changed their behaviour in response to this growth.
5.134 The evidence we received on past experience, and the limited sales of the new
products achieved to date, indicated to us that these products were unlikely to offer a
competitive constraint on PPI sold at the point of sale in the short to medium term.
The effects on consumers of the level of competition in the market
5.135 Having found features that prevent, restrict or distort competition in the sale of PPI
products, we considered what the effects on consumers would be. First, as we found
that distributors have high margins when selling PPI, we thought that consumers
were paying prices higher than they would in a well-functioning market.
5.136 We also thought that the high margins on PPI available to businesses gave them the
incentive to maximize the uptake of PPI among their credit customers. The effect of
this would be higher levels of sales than might otherwise be the case and that the
quality of sales would not always be the paramount consideration for businesses. We
noted in this respect that the FSA has been active in trying to improve sales of PPI
products, and has taken action against 20 firms over poor PPI selling practices.
5.137 We noted that the high prices for PPI might have a countervailing downwards
pressure on sales, resulting in lower sales than might otherwise be the case, for two
reasons. First, some people who would buy PPI do not do so because the high cost
is prohibitive to them. Second, high prices, in combination with regulatory action
against mis-selling, have resulted in adverse media coverage, which we would
expect to discourage some customers from buying PPI.
5.138 The overall effect on demand of the upwards and downwards pressures in paragraphs 5.136 and 5.137 is ambiguous. Nevertheless we believe that these pressures
result in a distortion of demand for PPI.
5.139 We noted that competition between credit providers is largely on the APR of the loan.
The APR is a way of showing the interest repayable on taking out a loan. It does not
take into account the cost of optional products, such as PPI. As such, we thought that
there was an incentive for distributors to sell credit at prices lower than would otherwise be the case, using the high margins on PPI to maintain an adequate rate of
return. In this respect we noted that some distributors told us that they did crosssubsidize personal loan APRs with PPI (whilst others said that this was not the case).
We also found evidence from loan pricing models and strategy documentation that
suggested PPI income had influenced APRs. On balance, we thought that an effect
of the level of competition in PLPPI markets was that personal loan APRs were lower
than they would be otherwise. We found less evidence that PPI income had had a
significant influence on the pricing of credit cards or mortgages.
5.140 Having concluded that PLPPI prices were higher than they would be in a wellfunctioning market, and that as a result credit prices were lower, we thought that
consumers’ spending patterns in personal loans were distorted as a result of the
state of competition in PPI markets.
124
5.141 Finally, we thought that the net effect of this was that there may be a transfer of
welfare between consumers who purchase PLPPI and those who do not, with PLPPI
purchasers subsidizing consumers who purchase credit without APR. As noted in
paragraphs 4.88 and 4.89, we found limited evidence to show that this cross-subsidy
is primarily within credit bands—that is, that PLPPI purchasers are cross-subsidizing
non-PLPPI purchasers of similar risk. We did not think it was the case that poorer
customers (who are more likely to take out PPI) were subsidizing better-off
customers (who are less likely to take out PPI). However, we concluded that this
cross-subsidy meant that non-PPI purchasers paid less for personal loans than
would be the case were it not for the PPI income earned from PPI purchasers within
that credit band.
5.142 We recognize that there is an argument that overall this balances out and that if
distributors do not make excessive profits when selling credit and PPI, we should not
be concerned whether, within a market for credit and PPI, some customers received
a better deal than others. We considered whether there was a relevant customer
benefit of lower credit prices to consumers arising from this cross-subsidy in
paragraphs 10.442 to 10.463.
5.143 We also found that barriers to searching and switching limit consumer choice (see
paragraphs 5.57 and 5.76 to 5.78), and concluded that consumers had less choice of
PPI policies than would be expected in a well-functioning market. Given that there is
no clear evidence of PPI distributors or intermediaries seeking to win sales from each
other by competing on non-price factors such as quality innovation or choice, we
consider that it is unlikely that there is vigorous competition on non-price factors
because these factors are particularly difficult for consumers to observe (paragraph
4.2). Given this lack of competition, we consider it is possible that there is less
innovation than would be expected in a well-functioning market.
Conclusions on factors affecting the nature and extent of competition
5.144 We conclude that there are features of relevant markets which, alone or in
combination, prevent, restrict or distort competition in the supply of PPI to nonbusiness customers in the UK:
(a) Distributors and intermediaries fail actively to seek to win customers by using the
price or quality of their PPI policies as a competitive variable.
(b) Consumers who want to compare PPI policies (including PPI combined with
credit), stand-alone PPI or short-term IP insurance policies are hindered in doing
so. Product complexity (the variations in pricing structures (in particular in relation
to single-premium policies) and in terms and conditions, the way information on
PPI is presented to customers), the perception that taking PPI would increase
their chances of being given credit, the bundling of PPI with credit, and the limited
scale of stand-alone provision act as barriers to search for all types of PPI
policies. In addition, the time taken to obtain accurate price information is a
barrier in relation to the provision of PLPPI, MPPI and SMPPI. These barriers to
search impede the ability of consumers to make comparisons, and therefore
effective choices, between PPI policies. They also, therefore, act as barriers to
expansion for other PPI providers, in particular providers of stand-alone PPI.
(c) Consumers who want to switch PPI policies to alternative providers or to
alternative types of insurance policies are hindered in doing so. Terms which
make switching expensive (in the case of single-premium policies) act as barriers
to switching for PLPPI and SMPPI policies. Terms which risk leaving consumers
uninsured (for a short period of time or in case they suffer a recurrence of a
125
problem) act as barriers to switching for all types of PPI policies. In addition, the
lack of access to consumers’ balance information acts as a barrier for switching
for CCPPI as it renders stand-alone providers unable to offer equivalent CCPPI
policies. These barriers to switching limit consumer choice. They also therefore
act as barriers to expansion for other PPI providers, in particular providers of
stand-alone PPI.
(d) The sale of PPI at the point of sale by credit providers further restricts the extent
to which other providers can compete effectively.
5.145 We have found, pursuant to section 134(1) of the Act, that there are features of
relevant markets, which alone or in combination, prevent, restrict or distort
competition in the supply of PPI and accordingly there is an AEC within the meaning
of section 134(2) of the Act. The features are those that we have identified in this
section and which are summarized in paragraph 5.143.
5.146 We concluded that the detrimental effects on consumers of these features were
higher prices for, and less choice in, PPI policies than would be expected in wellfunctioning market. We also concluded that demand for PPI was distorted. We also
concluded that it was possible that there was less innovation than would be expected
in a well-functioning market.
126
6.
Retail PPI
Introduction
6.1 In this section we consider the state of competition in the supply of retail PPI.
6.2 We concluded that the type of retail finance sold in conjunction with the PPI defines
the type of PPI policy. We found two types of retail finance: 1
• ‘Personal loan’ retail finance, which is a personal loan provided by or on behalf of
the retailer granting credit to the consumer to purchase a specific good (or goods,
usually of high value) sold by the retailer of that good.
• ‘Credit account’ retail finance, which is a running account credit facility provided by
the retailer to the consumer to purchase goods from the retailer. The consumer
typically buys goods at different times, such that their credit balance rises and falls
over time.
6.3 We concluded that PPI sold alongside ‘personal loan’ retail finance was a form of
PLPPI, in that the PPI is typically sold to insure one-off purchases (similar to motor
retail finance PPI—see paragraph 2.10), and the PPI is not bundled with a product
such as merchandise cover. Therefore our conclusions on PLPPI in Section 5 apply
to ‘personal loan’ retail PPI and we do not consider ‘personal loan’ retail PPI any
further in this section.
6.4 We concluded that PPI sold alongside ‘credit account’ retail finance, although sharing
some of the characteristics of CCPPI, required separate investigation into the state of
competition for ‘credit account’ retail PPI.
Retail PPI
What is retail PPI
6.5 Although credit account retail finance may be offered by any type of retailer, our
investigation found that retail PPI was only offered by home-shopping retailers. We
found that the underwriters we consulted provided PPI for retailers who offered
personal loan retail finance and to home-shopping firms that distributed retail PPI for
credit account retail finance, but not to any other type of retailer. Many retailers may
distribute PPI for store card retail finance, but this is outside the scope of our
investigation. This section focuses on retail PPI for credit account retail finance (retail
PPI).
6.6 Home-shopping retailers generally do not have a high street presence and sell
mainly clothing, gifts, furniture and electrical goods through biannual catalogues. 2
Orders are placed by telephone, post or, increasingly, over the Internet. 3 We were
told that it is not uncommon for customers to have maintained a catalogue account
for 20 years or more, and that a substantial number of customers maintain accounts
with one or more catalogue provider. Goods can be paid for by cash, credit card, or
by using a credit account facility offered by the retailer. The credit account may be
opened at the point of purchase of the goods and offers a variety of methods to pay:
1
A third type of retail finance may be provided by store cards, but these and store card PPI are outside the scope of this investigation.
2
The use of catalogues may be complemented by other merchandise promotions during the course of the year.
3
During the period between May and October 2008 SDGFS recruited [] new credit customers (setting up a credit account) and [over half] of these customers placed their first order via the Internet. 127
either spreading payments over a fixed period, for example 20 weeks, or making
minimum payments on the account, which may be interest-free or may involve some
interest or paying the entire balance. Table 6.1 shows the proportion of homeshopping sales by method of payment for three of the largest retail PPI providers. It
shows that the clear majority of home-shopping purchases are made using a retail
credit account.
TABLE 6.1 Home shopping sales in 2007 by method of payment
per cent
Retailer
Cash/
credit card
SDGFS
J D Williams
Express Gifts
[]
[]
[]
Retail credit
Uninsured
Insured
[]
[]
[]
[]
[]
[]
Source: Information provided by retail PPI distributors.
6.7 Retail PPI insures the outstanding balance on a home-shopping credit account with a
retailer. The policy pays a benefit which is typically the minimum monthly payment
due each month on the credit account, which may vary from 5 per cent up to 33 per
cent of the outstanding balance. Payments are made on a monthly basis until the
claimable event ceases or the account balance is cleared, whichever is sooner.
These policies insure against some or all of A, S, U, incapacity (I), death (L), H and
CI. The premium is paid monthly (either on a calendar or four-weekly basis) and is
calculated as a percentage of the monthly outstanding balance.
6.8 Average retail credit account balances for customers with retail PPI are typically
between close to £100 and £400 per month. SDGFS told us that average monthly
premiums were between £[] and £[] per month. This is consistent with average
balances of £[] to £[] per month. Other distributors told us that their average
balances were lower and consequently their average premium income was lower.
JD Williams Limited (JD Williams) receives on average £[] per year in premium
income per customer—consistent with average balances of around £[]. Otto UK
(Otto) receives on average £[] to £[] per year—consistent with average balances
of between £[] and £[]. Finally, Express Gifts Limited (Express Gifts) receives on
average £[] per year—consistent with average balances of around £[].
6.9 All the distributors we received evidence from sold retail PPI bundled with merchan­
dise cover. This is insurance that covers the goods purchased for accidental damage
and theft. It is bundled with the retail PPI so that the retail PPI is not available without
the merchandise cover. 4 Merchandise cover was outside of our terms of reference.
We therefore did not consider merchandise cover, except where it affected the extent
of competition for retail PPI.
Who sells retail PPI?
6.10 We identified four home-shopping retailers that achieved a GWP for retail PPI in
excess of £1 million in 2007. These were SDGFS, 5 JD Williams, Otto and Express
Gifts, whose main retail brands are shown in Table 6.2. 6 Retail PPI is a small part of
4
One distributor, Otto, sold a retail PPI policy for which the merchandise cover was optional for a short period in 2007. It subsequently exited the supply of retail PPI. 5
Formerly Everyday Financial Services. 6
In November 2007, Express Gifts suspended sales of retail PPI, except to online customers, and Otto ceased selling retail PPI entirely. 128
these retailers’ overall business, contributing between 1 and 5 per cent of total
turnover. 7
TABLE 6.2 Selection of main retail brands selling retail PPI
Distributor
SDGFS
JD Williams
Otto
Express Gifts
Littlewoods, Littlewoods Direct, Great Universal, Kays, Abound, Additions, Marshall Ward
J D Williams, Ambrose Wilson, Simply Be, Premier Man, Fashion World, Fifty Plus
Grattan, Freemans, Kaleidoscope, Look Again, Montage
Studio, Ace
Source: Information provided by retail PPI distributors.
Possible alternative policies
6.11 In Section 2, we considered possible alternative products to PPI. 8 For retail PPI,
there were particular potential alternatives for customers and retailers that were
linked to alternative types of credit available and which shared similar characteristics
to a credit account and retail PPI. First, customers could purchase goods from the
retailer using a credit card with the associated CCPPI product (see Sections 3 to 5).
Second, retailers could offer a store card credit facility with an associated store card
PPI policy. Store card PPI is outside the scope of this investigation.
Sales of retail PPI
6.12 Table 6.3 sets out the total GWP 9 and number of customers for retail PPI distributed
by the four distributors for the six-year period 2002 to 2007. These distributors sold
very small volumes of retail PPI compared with the distributors of other forms of PPI.
One distributor, SDGFS, sold over half of all retail PPI by GWP over most of this
period. GWP has been declining—falling by 12 per cent in 2006 and by a further
[0–10] per cent in 2007. This mirrors the trend in other forms of PPI. 10 There were
around 2 million retail PPI customers for most of this period, although numbers
declined to around 1.4 million by 2007. In November 2007, Express Gifts stopped
selling retail PPI, except to online customers, and Otto stopped selling retail PPI
entirely (see paragraphs 6.139 and 6.140).
7
We consider profitability of retail PPI from paragraph 6.99. See paragraphs 2.15–2.22. 9
GWP is for retail PPI only, although the distributors reported GWP to us for retail PPI and merchandise cover combined (with the exception of Express Gifts), and all sold combined policies. We calculated GWP for retail PPI, excluding merchandise cover, based on the price per £100 outstanding balance which related to retail PPI only as reported to us by each party. 10
See paragraph 2.24. 8
129
TABLE 6.3 Retail PPI GWP and number of customers split between distributors
GWP
Distributor
2002
Total
Total (excluding SDGFS
other brands)
Growth (%)
SDGFS: Littlewoods
SDGFS: other brands†
JD Williams
Otto
Express Gifts
2003
2004
2005
2006
£ million
Share of
supply (%)
2007
2007 100%
73
72
71
69
61
73 73
72
71
69
61
[]
[]
[]
[]
[]
[]
0
[]
[]
[]
[]
[]
–3
[]
[]
[]
[]
[]
–3
[]
[]
[]
[]
[]
–12
[]
[]
[]
[]
[]
[–10–0]*
[]
[]
[]
[]
[]
Customers‡
>50
20–50
<10
<10
’000
Total
1,857
2,016
2,121
2,064
1,687
1,441
9
5
–11§
–18
–15
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
Growth (%)
SDGFS
JD Williams
Otto¶
Express Gifts
[]
[]
[]
[]
Source: CC calculations based on information provided by retail PPI distributors.
*Growth rate excludes SDGFS: other brands.
†SDGFS does not hold data on other brands prior to 2007/08.
‡This is the number of accounts at the end of each year. Retailers differ in how they manage accounts that hold nil balances and are therefore not contributing retail PPI premiums. This excludes SDGFS: other brands. §Growth rate figure excludes sales of Otto. ¶Otto does not hold this data prior to 2005. 6.13 Table 6.4 shows the penetration rates of retail PPI for each of the distributors. The
penetration rate for [] and [] is much higher at over 25 per cent, and up to 50 per
cent respectively, than for [] and [] at below 16 per cent. However, penetration
rates fell by over one-quarter for both [] and [] between 2005 and 2007.
TABLE 6.4 Penetration rates for retail PPI
% of credit
Distributor
[]
[]
[]
[]
2002
2003
2004
2005
2006
2007
[10–20]
[40–50]
[10–20]
[40–50]
[10–20]
[40–50]
[0–10]
[0–10]
[0–10]
[10–20]
[30–40]
[40–50]
[10–20]
[10–20]
[30–40]
[40–50]
[0–10]
[10–20]
[20–30]
[30–40]
[0–10]
Source: CC calculations based on information provided by retail PPI distributors.
Note: The penetration rate used here, for all except [], is the value of purchases made using a credit account with retail PPI
as a proportion of the value of all purchases made using a credit account. [] provided the proportion of insured balances with
retail PPI. Other measures could also have been used such as the proportion of customers with retail PPI.
Organization of the retail PPI value chain
Underwriters
6.14 All distributors of retail PPI are vertically separated from the underwriter of the
insurance. We did not find that there was a competition problem in underwriting for
130
other forms of PPI, 11 and, as underwriting for retail PPI follows the same principles
as for other forms of PPI, we had no reason to believe that there was a competition
problem for retail PPI. Therefore, we found that there is a sufficient level of
competition between underwriters of retail PPI.
How retail PPI is sold
Sales process for retail PPI
6.15 Retail PPI is offered by distributors to new and existing customers with a credit
account. Unlike the other types of PPI considered in Sections 3 to 5, retail PPI is not
necessarily sold at the opening of the credit account. Customers of Otto were, and
customers of JD Williams are, called some weeks after opening an account and
holding a balance. Sales are usually initiated by the distributor, either during the
sales process for goods and a credit account, from a telemarketing call or by
providing a tick-box for online customers. 12 The payment method for retail PPI
policies is always ‘regular premium’. 13 The method of sale differs for each of the retail
PPI distributors and is shown in Table 6.5.
TABLE 6.5 Sales processes for retail PPI
Distributor
SDGFS Retail PPI is offered at the point of sale of the retail good and credit account when a customer first telephones
to place an order. It is offered at [] intervals to existing customers without retail PPI, and is available at any
time for Internet orders. Sales have been on a non-advised basis since June 2007.
JD Williams After their third statement, customers are sent information about retail PPI and are telephoned by an out­
sourced telemarketing bureau. Sales are non-advised.
Otto Sales were suspended in 2007. Prior to this, customers were telephoned once they had opened an account
and held a balance. This was some weeks after opening the account. Sales were on an advised basis.
Express Gifts Retail PPI is available only to customers who open or use a credit account online. A tick-box is provided for
customers to accept retail PPI. Direct telephone marketing was suspended in 2007. Sales are non-advised.
Source: Information provided by retail PPI distributors.
The price of retail PPI
6.16 The price of retail PPI is quoted on a monthly basis as pounds per £100 of
outstanding balance, similar to CCPPI. We found that prices were in the range £1.95
to £1.99. The cost of merchandise cover is included in this price; SDGFS told us that
retail PPI made up £[] of the total premium, Otto briefly sold a retail PPI policy
without merchandise cover for £[], JD Williams told us that retail PPI made up
£[] of the total premium, and Express Gifts told us that the retail PPI element was
£[]. In all cases the amount of the premium which related to retail PPI was more
than £1, significantly so in some cases. In comparison, we found that the price for
CCPPI was around £0.79.
11
See Section 7. SDGFS said that sales ‘initiated’ by it were restricted to specific marketing materials sent to existing customers, any inbound
calls from customers ordering goods and outbound sales calls to existing customers. It said that there was no automatic prompt to take out PPI when an existing customer placed a subsequent order over the Internet. 13
That is a monthly premium is paid to maintain the policy, which may vary from month to month depending upon the balance of the account. See paragraph 2.78.
12
131
Market definition
6.17 In this section we define the relevant market for the distribution of retail PPI. As noted
in our guidance, and in Section 3, the key to market definition is substitutability—the
extent to which customers can readily switch between products, or suppliers can
readily switch their facilities between the supply of alternative products. 14 In
assessing the likelihood of substitutability, as in Section 3, we apply the SSNIP
framework in our analysis regarding the relevant market.
6.18 We identified the following substitution possibilities that could, potentially, render
unprofitable a SSNIP (above competitive levels) by a hypothetical monopolist of
sales of retail PPI to customers of a particular retailer:
• customers could substitute to another method of insuring their retail credit
balances (for example, by taking out stand-alone PPI or a short-term IP policy);
• customers could substitute to an alternative method of financing and insuring their
purchases (for example, by substituting to a credit card with CCPPI); or
• customers could substitute to another retailer that offers a retail credit account and
retail PPI.
6.19 As in Section 3, we started with the narrowest conceivable market—that of the sale
of retail PPI by a distributor to its own retail customers. We then considered whether
the substitutability of retail PPI with any of the alternatives listed in paragraph 6.18
was sufficient to justify a wider market definition, including one or more of the
alternatives outlined in paragraph 6.18. We assess the constraint on retail PPI
distributors from (a) the possibility of consumers substituting to alternative types of
credit insurance in paragraphs 6.22 to 6.32. We consider the constraint on retail PPI
distributors from (b) the possibility that customers may substitute to alternative
combinations of credit and PPI in paragraphs 6.33 to 6.56. Finally, we consider the
constraint on retail PPI distributors from (c) the possibility that customers may
substitute to alternative bundles of goods, credit and PPI in paragraphs 6.57 to 6.70.
6.20 SDGFS said that this was not an appropriate way of analysing the relevant market. It
said that we should use as a starting point the product with which the undertaking
addresses the market, which it said was a bundle of goods, credit and retail PPI. We
did not agree, for the reasons set out in paragraph 8 of Appendix 3.7. In addition we
noted that retail PPI is optional, that SDGFS offers PPI to existing customers during
inbound and outbound calls (see footnote to paragraph 6.15),and that JD Williams
only offers retail PPI to customers some weeks after opening an account and holding
a balance (see paragraph 6.15). We were content, therefore, that it was appropriate
to consider all of the possibilities regarding market definition for retail PPI.
6.21 SDGFS told us that in addition to substitutability between retail PPI and other
products, we should also consider the effects of any change in the costs of
production as the quantity produced decreased and the margin earned on each unit
sold on the profitability of a SSNIP. However, we noted that where there is little
substitutability between retail PPI and other products then the change in quantity as a
result of a price rise will also be small. If this is the case there will be a small change
in the costs of production. This factor would therefore have little impact on the
profitability of a SSNIP in this instance. Furthermore, to the extent that there were
changes to the costs of production as a result of lost sales, these changes would
14
CC3, and paragraph 3.10.
132
result in cost savings and would tend to result in higher profits as a result of a SSNIP.
Taking these cost savings into account, as suggested by SDGFS, would therefore
simply strengthen the case for a narrow market definition and would not lead to a
wider market definition. We noted that where margins are high, the critical value of
sales loss that would render a price rise unprofitable is smaller than is the case when
profit margins are low. However, as noted in paragraph 3.12, in a market
investigation, the direct application of the SSNIP test is complicated by the
cellophane fallacy which may lead to the erroneous conclusion that markets are wide
if current prices are in excess of competitive levels. We consider indicators of the
general level of competition and consequently whether retail PPI prices are likely to
be in excess of competitive levels in paragraphs 6.78 to 6.147. In the context of this
evidence we were satisfied that, in this case, a finding that there was limited
substitutability was sufficient to define a narrow market for retail PPI.
Substitutability of retail PPI with short-term IP and stand-alone CCPPI
6.22 In order to assess the extent to which retail PPI distributors are constrained by the
potential for their customers to substitute to other types of credit insurance, we
looked at the following:
• the availability and suitability of alternatives to retail PPI; and
• the responsiveness of retail PPI demand (in terms of the penetration rate) to past
increases in retail PPI prices.
6.23 In a wider market, where customers can substitute easily between retail PPI and
alternative types of credit insurance, we would expect to see a range of alternative
policies which are close substitutes for retail PPI. We would also expect to see the
result of customer substitution to these alternatives, namely significant sales of these
alternative products.
6.24 In addition, in a wider market we would expect to see that increases in retail PPI
prices cause customers to substitute away from retail PPI to these alternative
products. Changes in retail PPI prices should therefore result in a reduction in the
proportion of retail credit balances that are insured using retail PPI.
The availability and suitability of alternatives to retail PPI
6.25 We found that it was not possible to buy retail PPI from one distributor to cover a
credit account with another retailer. However, a customer could use stand-alone PPI
or short-term IP, which was an alternative to all forms of PPI including CCPPI.
6.26 In Section 3 we found only two providers offering stand-alone substitutes for CCPPI
sold at the point of sale: Paymentcare and, as of 2007, Barclays (which is available
to customers who hold a Barclaycard and want to insure other credit cards as well). 15
We similarly concluded that these were the only stand-alone products available as
possible substitutes for retail PPI. However, only [0–2,500] of these products were
sold in a year 16 which represents a very small proportion of the 1.4 million retail PPI
customers in 2007. Similarly, in Section 3, we found that sales of short-term IP were
very low, with sales of 10,000 to 15,000 sold in a year. 17 []
15
See paragraph 3.39.
See paragraph 3.40 and Table 3.3.
17
See paragraph 3.40.
16
133
6.27 We noted that these policies are poor substitutes for retail PPI, which will limit the
extent to which retail PPI customers can substitute away from retail PPI in response
to a price rise. First of all, short-term IP and stand-alone CCPPI policies do not track
the balance of the credit account. Therefore a customer will be over- or underinsured for their balance at any time. In addition, no alternative short-term IP or
stand-alone CCPPI policy that we are aware of includes merchandise cover, so
consumers cannot switch to an alternative policy offering comparable cover. We
therefore concluded that substitution to stand-alone CCPPI and/or short-term IP in
response to a retail PPI price rise would be unlikely.
6.28 One retailer (JD Williams) said that a further alternative that should be considered
was the possibility that customers could self-insure, and that this might place a
constraint on the prices retailers could charge for retail PPI. We noted that if selfinsurance were a constraint, then this should be represented in the general level of
responsiveness of retail PPI demand to retail PPI prices. This is because if
customers are willing to self-insure as a close (substitutable) alternative to retail PPI,
then changes in retail PPI prices should lead to relatively large changes in retail PPI
demand. We consider responsiveness of retail PPI demand to retail PPI prices in
paragraphs 6.29 to 6.31.
The responsiveness of retail PPI demand to past increases in retail PPI prices
6.29 In addition to assessing the availability and suitability of alternatives to retail PPI, we
also looked at the retailers’ experiences following increases in retail PPI prices in the
past. If retailers are constrained by the potential for customers to substitute to
alternatives, then we would expect an increase in the price of retail PPI to result in a
reduction in the penetration rate, as customers substitute away from retail PPI to
these alternatives.
6.30 As noted in paragraph 6.6, we were told that it is not uncommon for customers to
have maintained a catalogue account for 20 years or more. This might indicate a
limited degree of responsiveness to price. We asked the distributors for evidence on
the responsiveness of retail PPI demand to price changes. There have been only
three price changes for retail PPI in the period 2002 to 2007 by Otto and JD Williams;
SDGFS and Express Gifts did not change their price at all over the period. 18 In each
case, the price of the combined retail PPI and merchandise cover policy was
increased:
• Otto increased prices by 22 per cent in late 2002, at the same time as increasing
merchandise cover from 12 months to 24 months, and by 9 per cent in summer
2004, whilst increasing merchandise cover to 30 months. Otto told us that whilst
limited relevant historical data was available in relation to these improved levels of
cover and the product’s consequent marketability, ‘it is not apparent that the price
increases significantly impacted the marketability of the retail PPI product’. We
considered that the impact on sales may have been offset by the increase in
merchandise insurance, but it was not possible to separate these two factors. We
can therefore infer very little from the Otto price rises.
• JD Williams increased prices by 14 per cent in June 2003. JD Williams told us
that:
18
SDGFS purchased the Empire brand in July 2008, and reduced the price of shopping insurance to Empire customers to the
level paid by its other customers (see paragraph 6.53 for further details).
134
Initially there was no material impact on sales after the retail PPI
premium increase, but sales penetration did begin to fall towards the end
of that year (penetration was unchanged until October before falling). At
around the time of the decline in penetration rates a new telemarketing
script in line with GISC19 standards was introduced.
6.31 Given the timings of the price rise and the script change, we thought the decline in
penetration rates was more likely to be related to the new script than the price rise
four months previously. The lack of impact on sales in the four months following the
price rise implied to us that the fall in sales from October onwards was not related to
the price change. SDGFS said that it could have taken four months for price
increases to flow through to penetration rates and we could not assume that the price
increase did not influence sales. However, given the time lag and the new sales
script, our judgement was that the experience of JD Williams in June 2003 was more
likely to be evidence that customers did not react to a retail PPI price change by
substituting to alternative types of credit insurance. This evidence was therefore
consistent with a narrow market which does not include alternative types of credit
insurance.
Conclusion on the substitutability of retail PPI with other types of credit insurance
6.32 We found that the lack of suitable alternatives to retail PPI, the limited sales of those
alternatives, their limited effectiveness as substitutes and the limited impact that they
appeared to have on retail PPI sales following a price rise indicated that the prospect
of customer substitution to alternative types of credit insurance was a relatively weak
constraint on retail PPI suppliers. 20
The substitutability of the retail credit and retail PPI combination with alternative
combinations of credit and PPI
6.33 As outlined in paragraph 6.18, in addition to other means of insuring their retail credit
balances directly, it is also potentially possible for consumers to react to retail PPI
prices by substituting to alternative methods of financing their purchases, on which
PPI is available.
6.34 It is not possible to take out a retail credit account with one retailer to pay for
purchases of goods at another retailer. We therefore concluded that the main
potential source of substitution of this type, in response to a rise in retail PPI prices,
would be for retail PPI customers to substitute to a credit card with CCPPI to pay for
their purchases of goods instead of using a retail credit account and retail PPI.
6.35 An important factor in the assessment of the constraint imposed on retail PPI
distributors from substitution to alternative combinations of credit and PPI is the
extent to which customers consider retail PPI when deciding on their choice of credit
supplier.
6.36 A change in the price of retail PPI will affect the price of a combination of retail credit
and retail PPI from that retailer too. This in turn may induce some customers to
respond to changes in the price of retail PPI, in the same way that they would
19
General Insurance Standards Council.
SDGFS said that we should consider the prospect for supply-side substitution in our assessment of the relevant market. We
considered that there was no prospect of supply-side substitution into the supply of retail PPI to a retailer’s credit account
customers in response to a SSNIP. There are a number of factors that would prevent supply-side substitution. These factors
are discussed in the context of barriers to entry in paragraphs 6.171–6.176 and also in the context of the point of sale
advantage in paragraphs 6.177–6.184.
20
135
respond to an equivalent change in the price of retail credit by substituting to
alternative insured credit products. In other words, customers may base their
decisions on the whole cost of credit and PPI, rather than just on the cost of PPI
alone and as a result they may substitute between combinations of credit and PPI.
Those customers that do not consider the whole cost in this way will not substitute
between combinations of credit and PPI.
6.37 Where such substitution in response to a change in retail PPI price is sufficient, it
may warrant a wider market definition to include alternative combinations of credit
and PPI. However, if few customers respond in this way to changes in retail PPI
prices, then this constraint will be weak and a narrow market definition would be
appropriate.
6.38 In order to assess the constraint on retail PPI distributors from this potential source of
substitution, we assess the available evidence on the propensity of retail PPI
consumers to consider retail PPI prices when making their choice of the method of
financing their home-shopping purchases. In other words, we assess the propensity
of retail PPI consumers to consider the whole cost of the combination of credit and
PPI.
The extent to which retail PPI customers consider the whole cost of the combination
of credit and PPI
6.39 In order to consider the extent to which retail PPI customers consider the whole cost
of the combination of credit and PPI, we looked at the following:
• whether retail PPI exhibited any differences, when compared with other types of
PPI (and in particular CCPPI), that would make it more likely that retail PPI
customers would consider the whole cost of the combination; and
• the evidence supplied by the parties regarding the extent to which their customers
consider and react to the whole cost of the combination.
• Evidence from other forms of PPI
6.40 In Section 3 we concluded that for PLPPI, MPPI and CCPPI the relevant product
market was no wider than the sale of PPI by a distributor to its own credit customers,
and in particular that customers in these markets did not consider the whole cost of
the combination to a sufficient degree to warrant a wider market definition. 21 In this
section, we consider whether retail PPI customers are more likely to consider the
whole cost of the combination than customers of these other forms of PPI and
therefore whether a wider market definition is appropriate.
6.41 If retail PPI customers do consider the whole cost of the combination, then there
must be some differences between how customers shop for retail PPI compared with
other forms of PPI, where we have already found that few customers consider the
whole cost. These differences may be because (a) customers are more likely to
consider the whole cost of the combination when shopping for retail PPI than for
other forms of PPI, or (b) different customers or types of customers are involved and
that these types of customer are more likely to consider the whole cost of the
combination. We considered each of these two possibilities.
21
See Section 3.
136
6.42 First, we looked at the similarities and differences between retail PPI and other types
of PPI to assess whether customers would be more likely to consider the whole cost
of a combination for retail PPI than is the case for other types of PPI. We identified
the following similarities and differences:
• CCPPI and retail PPI polices have similar characteristics. They both insure a
variable balance; premiums are calculated as a percentage of the monthly
balance and are zero when a nil balance is held; they provide cover for similar
types of risk; prices, terms, conditions and benefits are expressed in the same
way.
• Unlike personal loans and credit cards and their associated PPI policies, retail PPI
and retail credit is available only on the purchases made with a particular retailer.
• Unlike other forms of PPI, merchandise cover is included with retail PPI policies. 22
• The level of balances insured by retail PPI is generally much smaller than average
credit card balances, personal loans or mortgages. 23
6.43 We considered that many of the key differences would appear to make it more
difficult for customers to consider the whole cost of a combination of PPI and credit
for retail PPI when compared with other forms of PPI. For example, although retail
PPI and CCPPI are similar, the bundling of retail PPI with merchandise cover, 24 and
the differences in terms of the credit between retail credit accounts and credit cards,
would seem to make comparing a combination including retail PPI, with an
alternative combination of credit and PPI, more difficult. In addition, because the
characteristics of a retail PPI and retail credit combination are markedly different from
the characteristics of a credit card with CCPPI, customers will be less likely to react
to changes in retail PPI prices by substituting to credit cards with CCPPI from retail
PPI than they would be to substitute between insured credit cards.
6.44 We also found that as the average balance on a retail credit account is small, the
benefits to search for retail PPI would appear to customers to be smaller than for
other types of PPI. We did not think that the fact that retail credit and retail PPI was
available only on purchases from one retailer would affect substitution between retail
PPI and insured credit cards. However, this factor does prevent substitution between
combinations of retail credit and retail PPI offered by different distributors. We
therefore formed the view that it is unlikely that customers would have a higher
propensity to consider the whole cost of a combination of credit and PPI in retail PPI
than in other forms of PPI.
6.45 Second, we considered whether the customers of retail PPI are a separate group
from customers of other types of PPI. SDGFS conducted a survey of its customers in
2007 25 which reported that only [<25] per cent of its retail PPI customers had CCPPI.
This suggests that there may be only a limited overlap between CCPPI customers
22
A similar practice was found to exist in relation to store cards. In the Store Cards market investigation the CC found that the
bundling of multiple types of cover contributed to a lack of transparency, given that the price of the individual elements of the
package is not available to customers. See Store cards market investigation, 7 March 2006 paragraph 8.149.
23
See paragraph 6.8. In 2006, the average balance across all credit cards, calculated using data from Datamonitor, was £940.
24
SDGFS said that the inclusion of merchandise cover did not make it more difficult for customers to compare the price of a
bundle of credit and PPI, as merchandise cover was an additional element of cover provided to home shopping customers. We
noted that the inclusion of merchandise cover in the bundle of PPI and retail credit clearly did not make comparison with a
bundle of CCPPI and credit card credit easier. In our view it makes it harder, as it introduces a significant element that is not
included in CCPPI and is unrelated to insuring the credit.
25
Source: Harris Interactive Customer Research 2007.
137
and retail PPI customers. 26 However, it was not clear to us that a limited overlap
would make it more or less likely that retail PPI customers would consider the whole
cost of the combination of credit and PPI.
6.46 We also looked at the demographic profiles of retail PPI customers compared with
the profiles of customers for other forms of PPI. Systematic information on the
demographics of retail PPI customers was not available. However, all four of the
largest distributors of retail PPI told us that the majority of their customers were
female. For example, [] told us that 82 per cent of its home-shopping customers
were female in [] (though PPI penetration rates were about the same for its male
and female customers). [] told us that 91 per cent of its retail PPI customers were
female in []. This contrasts with other forms of PPI where the gender mix of
customers is more even. 27
6.47 However, it is not clear that any differences in the demographics of retail PPI
customers, when compared with customers of other types of PPI, mean that it is
more or less likely that retail PPI customers will consider the whole cost of the
combination of credit and PPI. Retail PPI customers may have lower average
incomes than customers of other types of PPI, which may mean that they are more
price sensitive than customers of other types of PPI. However, we have no evidence
that this is the case. Furthermore, even if this were the case, we concluded that this
would not outweigh the additional difficulties in considering the whole cost of a
combination including retail PPI outlined in paragraph 6.43.
6.48 In Section 3 we found a consistent pattern across other types of PPI that we studied,
that PPI customers do not consider the whole cost of the combination of credit and
PPI when making their choice of supplier in the markets for credit. We considered
whether there are material differences in the characteristics or behaviour of retail PPI
customers that would make them significantly more likely to consider the whole cost
of the combination than customers of other forms of PPI. We concluded that it is
more difficult for retail PPI customers to substitute a combination including retail PPI
for an alternative combination of credit and PPI than is the case for CCPPI customers
substituting between different insured credit cards.
6.49 SDGFS said that it was not sufficient for us to rely on inferences based on work
undertaken which did not include any analysis on examination of retail PPI. We noted
SDGFS’s comment. However, our consideration of data on other products was
conducted based on products very similar to retail PPI (that is to say other PPI
products) and we considered possible differences between policies and customers
as part of our analysis. We were content therefore that we were able to make the
inferences we made from the data.
26
JD Williams said that it believed that other bundles of credit (including retail credit) and PPI would be regarded by its
customers as genuinely substitutable for JD Williams’ Retail PPI offer. It said there was a wide variety of credit and PPI options
available to retail credit customers, including from other mail order operations offering retail credit and PPI as well as store and
credit cards and personal loans. JD Williams said that the typical JD Williams customer had at least one other mail order
account and more than one credit card. We did not agree with JD Williams. The evidence from SDGFS shows that only [less
than 25 per cent] of its customers had a credit card with CCPPI—a level it considered to be substantial. We did not receive
survey evidence from JD Williams regarding the proportion of its customers that held CCPPI or retail PPI with another provider,
we therefore concluded that only a limited number of retail PPI consumers have CCPPI. Furthermore, even if there were a
larger incidence of retail PPI customers holding CCPPI or alternative retail PPI accounts with other companies, this does not
constitute evidence that consumers would substitute between them in response to a SSNIP for retail PPI, nor does it constitute
evidence that retail PPI consumers are more likely to shop on the whole bundle including retail PPI than purchasers of other
types of PPI.
27
See Appendix 2.6, Tables 1&2.
138
• Evidence from the parties
6.50 We asked the providers for evidence relating to whether retail PPI customers
consider the whole cost of the combination of credit and PPI. In particular, we asked
for any available primary evidence on this point. We asked:
• whether customers take retail PPI prices into account when choosing a credit
supplier; and
• for any evidence or analysis of the impact that an increase in retail PPI prices
would have on sales of credit.
6.51
[] told us that it had no evidence to suggest that customers took retail PPI prices or
other terms and conditions into account when making their choice of retailer or credit.
It said that customers were attracted to the product offering rather than the credit
account or PPI product that might subsequently be offered to them. It also said that,
as far as it was aware, changes in prices of its retail PPI had no impact on its sales of
credit or its goods. 28
6.52
[] told us that although it had no specific evidence with which to answer this
question, given that the sale of PPI was not directly linked to a retail purchase or the
opening of a credit account, the impact was likely to be minimal.
6.53 In response to our questions, SDGFS told us that its customers purchased a bundle
of retail goods, financed by credit and protected by shopping insurance. It said that
competition took place at the level of the bundled product, including credit and PPI,
but that was not to say that competition did not also take place at the level of PPI
alone. It also said that any increase in PPI price would have an effect on the sale of
goods and by virtue of this the sale of credit. It told us that when it purchased the
Empire brand in July 2008 it reduced the price of shopping insurance to Empire
customers to the level paid by its other customers. It told us that the rationale for this
was that it considered Empire’s price to be too high and would have a knock-on
effect on the propensity of customers to purchase goods from Empire. It also said
that the PPI price was too high compared with the prices charged by Empire’s
competitors and this element on its own would be uncompetitive. SDGFS did not
provide any documentary evidence to confirm its views.
Conclusion on the substitutability of the retail credit and retail PPI combination with
alternative combinations of credit and PPI
6.54 We found that it is more difficult for retail PPI customers to substitute a combination
including retail PPI for an alternative combination of credit and PPI than is the case
for CCPPI customers substituting between different insured credit cards. Given that
we found, in Section 3, that substitution between insured credit cards, in response to
a change in CCPPI prices, was low, 29 we concluded that it was unlikely that substi­
tution to alternative combinations of credit and PPI would be sufficient to constrain
retail PPI suppliers.
6.55 We sought evidence from the parties regarding the extent to which retail PPI
customers would or do substitute to alternative combinations of retail PPI in practice.
The responses that we received suggested that such substitution is unlikely. Two of
28
However, these changes in price have, in the past, been concurrent with a change in terms and conditions. See para­
graph 6.30. 29
See paragraph 3.27. 139
the distributors ([]) told us that they were not aware of any effective substitutes for
their retail PPI offering. Two distributors ([]) said that customers might substitute to
alternative forms of PPI. Only one provider ([]) suggested that customers might
substitute to other retail credit providers’ combinations of credit and PPI. However,
none of the parties was able to provide documentary, survey or any other primary
evidence that customers would or do substitute to alternative combinations of credit
and PPI.
6.56 We therefore concluded that the possibility of customer substitution to alternative
combinations of credit and PPI, in response to an increase in the price of retail PPI,
would be a relatively weak constraint on retail PPI suppliers.
The substitutability of the goods, retail credit and retail PPI bundle with alternative
bundles of goods, credit and PPI
6.57 As outlined in paragraph 6.18, in addition to substitution to alternative types of
insurance, and substitution to alternative bundles of PPI and credit, it is also possible
that retail PPI consumers could respond to an increase in retail PPI prices by
substituting to another retailer and purchasing a bundle of goods, retail credit and
retail PPI from that retailer instead.
6.58 As discussed in relation to substitution between bundles of credit and PPI above, a
key question in this regard is the extent to which retail PPI customers consider retail
PPI when making their choice of retailer.
6.59 We note that there are a number of practical reasons why substitution between
retailers in response to a change in retail PPI prices may be expected to be limited.
First, home-shopping catalogues are differentiated products. They have different
ranges of products, they may focus on different segments of customers, and they
operate under different brands. Customers may therefore be deterred from substitut­
ing between retailers because the specific product they are interested in is not
available or because of a preference for one retailer over another.
6.60 In addition, in order to compare the bundles on offer from competing retailers, retail
PPI customers must compare not only the differences in retail PPI offer, but also the
differences in retail credit account terms and the differences in the home-shopping
offer itself. 30 This makes the task of comparison more difficult.
6.61 The difficulty of this task is further compounded by the difficulty customers face in
assessing and comparing the whole cost of the bundle of goods, retail credit and
retail PPI. The price of goods is readily available within the home-shopping
catalogues produced by retailers. However, there is no easy way for customers to
compare the whole cost of a bundle and no readily available measure of the
combined price. Similarly, we are not aware of any comparison tools, such as
comparison websites, that cover retail PPI.
6.62 Finally, we note that, with the exit of Otto and Express Gifts from the supply of retail
PPI in 2007, there are only two large retailers for customers to choose from in order
to purchase a bundle of goods, retail credit and retail PPI.
30
SDGFS said that this was not difficult, and told us that [30–50] per cent of its customers had made purchases from at least
one other home-shopping operator in the preceding six months. We noted its comment; however, we did not think that the fact
that some customers buy goods from more than one catalogue gave us any evidence that customers would substitute between
them in response to changes in retail PPI prices, nor did it provide any evidence about the difficulty of doing so. Given that
there is a large variation in the benefits paid from a policy (see Table 6.8), and that information on terms and conditions is not
easy to find (see paragraphs 6.157–6.161), we disagreed with SDGFS.
140
The extent to which retail PPI customers consider the whole cost of the bundle of
goods, credit and PPI
6.63 Notwithstanding the difficulties outlined in paragraphs 6.59 and 6.62, we asked
distributors of retail PPI directly for evidence on whether retail PPI customers
consider the whole cost of the bundle of goods, credit and PPI. In particular, we
asked for any available primary evidence on this issue.
6.64 First, we asked distributors to provide evidence that their customers take retail PPI
prices or other terms and conditions into account when making their choice of
retailer. Otto and Express Gifts told us that they had no such evidence. SDGFS and
JD Williams gave contradictory evidence in response to this question (although
neither was able to provide us with primary evidence):
• SDGFS told us that ‘any increase in the price of its shopping insurance would
have a material impact on Littlewoods sales of goods. Littlewoods home shopping
customers will have the catalogues for the other home shopping providers and will
actively compare prices for the total package offered by each provider’. As noted
in paragraph 6.53, when it purchased the Empire brand in July 2008, it reduced
the price of shopping insurance to Empire customers to the level paid by its other
customers, telling us that the rationale for this was that it considered Empire’s
price to be too high and would have a knock-on effect on the propensity of
customers to purchase goods from Empire.
• JD Williams told us that:
We have no evidence to suggest that customers take retail PPI prices or
other terms and conditions into account when making their choice of
retailer or credit. With regard to our particular mail order catalogue
experience, the customers are attracted by the product offering rather
than the credit account or PPI product we may subsequently offer them.
6.65 Second, we asked for any evidence and analysis of the extent to which their
customers would or do switch to any alternative products in response to a small but
significant change in the price of retail PPI. We also asked whether they had any
evidence of the actual or likely impact of changes to retail PPI prices on sales of retail
PPI, retail credit or sales of goods. Such evidence would directly inform us about the
level of competition for retail PPI and, if goods or credit sales were affected, whether
customers were considering the whole cost of the goods, credit and retail PPI bundle.
None of the distributors was able to provide analysis of the likely impact of a change
in prices of retail PPI. SDGFS, JD Williams and Otto gave contradictory evidence in
response to this question:
• SDGFS told us:
Although there is no specific evidence or analysis in relation to Home
Shopping Retail PPI, it is well known that home shopping customers do
compare total retail prices (including PPI) across different retailers. The
increased participation of the internet also means it is becoming even
easier for customers to compare home shopping against other providers.
Customers will switch their primary retail purchasing decisions between
different home shopping providers if the total cost of the goods plus any
required PPI is too high.
• JD Williams told us:
141
We have no evidence to suggest that a change in retail PPI price has any
impact on sales of retail PPI. We recently made an increase in premium
charge from £1.98 per £100 of outstanding balance to £1.99. This had no
impact, however it was accompanied by improvements to the PPI product
overall. As far as we are aware changes of this type also have no impact
on sales of retail credit, or product.
• Otto told us ‘we have no evidence or analysis with which to answer this question.
However, given that the sale of PPI is not directly linked to a retail purchase or the
opening of a retail credit account the impact is likely to be minimal’.
6.66 Finally, we asked for examples of any reports, analyses, comparisons or other
documents for the purpose of monitoring competitors in the provision of retail PPI
and retail credit. If customers are considering the whole cost of the bundle, then
providers will need to ensure that their whole offer, including the prices of goods,
credit and retail PPI, is competitive. All the providers told us that they monitored
competitors’ catalogues and other retail promotions. However, we received no
documentary evidence of such analysis and comparisons from SDGFS or Express
Gifts.
6.67 JD Williams produced analyses of competitors’ retail PPI offerings, including prices
and benefits. These analyses were conducted in March and June 2003 and in April
2005.
6.68 Otto provided an analysis of its typical retail customer and their shopping behaviour,
and an analysis of customers’ retail credit offers. Neither analysis looked at retail PPI.
Conclusion on the substitutability of the goods, retail credit and retail PPI bundle with
alternative bundles of products
6.69 We identified a number of practical reasons why, in the absence of other contra­
dictory evidence, we would consider it unlikely that retail PPI customers would
respond to an increase in retail PPI prices by substituting to other retailers offering
retail credit accounts and retail PPI. These reasons include the differentiation
between home-shopping providers’ retail offers, the complexity of the task of
comparing bundles of goods, retail credit and retail PPI and the limited number of
retailers to choose from.
6.70 We received conflicting evidence from the retailers regarding whether, in practice,
these obstacles were overcome by consumers and that substitution between
retailers, in response to changes in retail PPI prices, did in fact occur. We actively
sought, and were unable to find, primary evidence to support the view that
substitution between retailers, in response to retail PPI prices, did occur. We
therefore concluded that, given the practical reasons why substitution may be
expected to be limited (see paragraphs 6.59 to 6.62), the constraint on retail PPI
distributors from the possibility that their retail PPI customers might substitute to
other retailers in response to an increase in retail PPI prices was weak.
Geographic market for retail PPI
6.71 Retail PPI is typically distributed via catalogues and call centres on a national basis.
We have received no evidence that would indicate that there are any regional
differences in prices or policy offerings which may suggest narrower geographic
markets. We therefore consider that it is appropriate as a framework for the analysis
of competitive constraints to define the relevant geographic market as the supply of
142
retail PPI by distributors to their customers within the UK. For providers that do not
sell across the whole of the UK, the relevant market is the area within the UK within
which they do sell retail PPI.
Conclusion on market definition
6.72 We identified three ways in which retail PPI customers could potentially react to a
SSNIP in retail PPI and which could result in that SSNIP being unprofitable. These
were that customers could substitute to other insurance policies, customers could
substitute to other forms of insured credit and customers could substitute to other
retailers that offered a credit account and retail PPI.
6.73 We found that a lack of suitable alternatives to retail PPI, the limited sales of possible
alternatives, their limited effectiveness as substitutes and the limited impact that they
appeared to have on retail PPI sales following a price rise indicated that the prospect
of customer substitution to alternative types of credit insurance was a weak con­
straint on retail PPI suppliers.
6.74 With regard to substitution between combinations of credit and PPI we found that it
was more difficult (and certainly no easier) for retail PPI consumers to consider the
whole cost of a combination including retail PPI when compared with other types of
PPI. We saw no clear evidence to suggest that retail PPI customers were intrinsically
more likely to substitute between combinations of credit and PPI than customers of
other types of PPI. We also received conflicting views from the retailers regarding the
extent to which their customers substitute between combinations in response to
changes in retail PPI prices and terms. We concluded that the possibility that retail
PPI customers may substitute to alternative combinations of credit and PPI, in
response to an increase in retail PPI prices, was a weak constraint on retail PPI
suppliers.
6.75 We also found that it was likely to be difficult for retail PPI customers to substitute
between retailers offering retail credit accounts and retail PPI. We received con­
flicting views from the retailers regarding the extent to which their customers would
substitute in this way in practice and none of the parties was able to provide primary
evidence that customers substitute between retailers in response to changes in retail
PPI prices. We therefore concluded that, given the practical reasons why substitution
may be expected to be limited, the constraint on retail PPI distributors from the
possibility that their retail PPI customers might substitute to other retailers in
response to an increase in retail PPI prices was weak.
6.76 Because the constraint on retail PPI distributors from each of the three potential
sources of substitution was weak, we concluded that a narrow market definition was
appropriate, and that it would be inappropriate to include other types of credit
insurance, or other combinations of credit and PPI, or other bundles of goods, credit
and PPI within the relevant market. We therefore defined the relevant product market
as the supply of retail PPI by a distributor to its own retail customers.
6.77
We concluded that the relevant geographic market for each product market was the
geographic area within the UK in which the distributor is active.
143
Indicators of the extent of competition between retail PPI providers in the
supply of retail PPI
6.78 In this section, we assess a series of indicators of the level of competition between
retail PPI distributors. 31
Variation of retail PPI prices over time
6.79 We looked for price changes of retail PPI by the four distributors, to see if this
showed evidence of competition between providers. Table 6.6 shows the price
changes during the period 2002 to 2007. There were only three significant price
changes over a period of six years. These were all increases; two of them were
accompanied by a major change to merchandise cover, although not the retail PPI
offering. There were some price-led promotions, for instance one month of free
cover. 32 SDGFS said that cover levels had increased over this time and so price had
effectively decreased. However, as noted in paragraph 4.10, we considered that the
need to treat customers fairly, and try to address negative views on this point, were
more likely to be the primary drivers for many of the policy changes we saw in PPI
policies, rather than trying to provide better value for money to gain a competitive
edge and we concluded that this was likely also to be the case in relation to retail
PPI.
TABLE 6.6 Price changes between 2002 and 2007
Distributor
SDGFS PPI prices have remained constant*
JD Williams 2003: Increased from 1.73% to 1.98%.
Otto 2002: Increase from 1.50% to 1.83% (accompanied by increase to merchandise cover)
2004: Further increase to 1.99% (accompanied by increase to merchandise cover)
Express Gifts No changes made.
Source: Information provided by retail PPI distributors.
*SDGFS told us that in July 2008 it acquired the right to use the Empire brand, and all Empire customers who were previously
charged 2.25 per cent received a reduced price of 1.98 per cent.
6.80 In a competitive market we would normally expect to see price variations over time
as firms seek to win business from each other. The level of price variation over time
that we saw was consistent with there being few significant competitive pressures on
retail PPI providers.
Indicators of non-price competition
6.81 In order to make an assessment of the extent to which retail PPI distributors engaged
in non-price competition, we considered evidence supplied to us to identify any new
retail PPI policies which had been introduced and whether there had been any
innovations within existing policies. Individual providers will sometimes find product
changes profitable absent any competition and so we also considered evidence that
the rationale for product change or innovation was to win customers from competi­
tors.
31
SDGFS said that competition primarily took place at the level of the bundle of merchandise, credit and PPI. However, we
found that the relevant economic market was the supply of PPI separate from merchandise and credit (see paragraph 6.74).
32
Otto gave the first month free between 2006 and 2007. JD Williams was running a similar offering on its website on 11 August
2008.
144
6.82 We were provided with evidence of one new product being introduced by Otto 33
between 2002 and 2007. This product was more flexible and could be tailored to suit
different customers, for instance with different policies and prices available to over
65s or unemployed customers, as well as an increase in the monthly benefit. It was
launched in November 2007; however, sales were halted after the first week and Otto
no longer offers retail PPI. We look at Otto’s decision to stop selling retail PPI in
paragraph 6.140.
6.83 Table 6.7 shows the innovations in cover offered by the four distributors. There were
many changes to terms and conditions of policies which increased the number of
customers that were eligible for retail PPI, such as increasing unemployment cover to
75 or removing pre-existing exclusions. There were also increases in the range of
circumstances for which a claim could be made, such as having to leave employment
to care for a relative, and the introduction of merchandise insurance. One distributor,
Express Gifts, made no changes to its policies.
6.84 We saw little evidence of price competition (paragraph 6.80). In a competitive market,
we may see changes in the monthly benefits of policies as firms seek to win cus­
tomers from each other using non-price element of their policy offer. We therefore
looked for evidence of non-price competition. However, for customers who were
always eligible under these policies, the actual benefits received in the event of a
claim were not changed within the period 2002 to 2007. 34
TABLE 6.7 Changes to policies between 2002 and 2007
Distributor
SDGFS
•
•
•
•
•
•
JD Williams
• Mar 2003: Personal accident cover, hospitalization qualifying period and death age limit all increased. CI
cover introduced.
• Jan 2005: Death age limit reduced. Hospitalization benefit increased.
• May 2008: L, CI. Accidental cover age limit increased. Exclusions to CI, L, S & U extended, now easier to
claim. Pet care included with hospitalization claims.
• No changes to monthly benefit.
Otto
•
•
•
•
•
Express Gifts
No significant changes.
Standardization of terms and conditions to ABI creditor claims levels.
Introduction of Carer cover (if you have to leave work to become a carer).
Increased unemployment cover to age 75.
Removal of pre-existing exclusions and other conditions.
Reduced waiting period before claims can be made.
Extension of merchandise cover (from 2 years to 3 years after purchase).
Increases in merchandise cover from 12 months to 24 months in 2002, and later to 30 months in 2004.
Other changes to waiting times and enhanced benefits for hospitalization cover.
Removal of claims cap for merchandise cover.
Permit claims for mental illness.
No changes to monthly benefit (except with introduction of Complete Cover product).
Source: Information provided by retail PPI distributors.
6.85 We considered the reasons for introducing new products and amended retail PPI
policies. First, we looked at internal documents provided by the distributors for
evidence that they were monitoring, and reacting to changes to, their competitors’
retail PPI policies. Second, we considered the explanations and internal documents
provided by the distributors outlining the reasons for the changes in Table 6.7.
33
Otto introduced ‘Complete Cover’ to replace ‘Product and Payment Protection’. SDGFS increased monthly benefits in 2000 and Otto increased the benefits with its new policy, but withdrew the product a week later. 34
145
6.86 Only one distributor provided examples of monitoring competitors’ retail PPI policies.
JD Williams compiled a comparison of retail PPI policies in spring 2003 and 2005,
including prices and benefits. Other distributors told us that they monitored competi­
tors’ catalogues and retailer promotions on an ongoing or ad-hoc basis, but not
specifically in relation to retail PPI. Otto told us that it would ‘be aware of any price
changes, but not necessarily changes to Terms and Conditions’.
6.87 The distributors provided us with explanations of the motivation behind changes to
their retail PPI policies:
• Otto extensively developed a new product that was launched in November 2007
(see paragraph 6.82). It told us that the motivation for this new product was to
ensure that it was offering and selling a product in line with all current and future
regulatory requirements.
• JD Williams explained that its policy changes in 2003 were to ‘improve the scope
of the product offering to our customers’. In 2005, policy changes were made due
to a change of underwriters. Finally, changes in 2008 were made when regulatory
changes to life cover took place and following the FSA TCF Initiative. Changes
were also made to ‘make it easier for customers to claim’.
• SDGFS told us that many of its changes to terms and conditions were due to
competition with other home-shopping brands in an attempt to differentiate the
product offering. Furthermore, it had an objective to achieve a Defaqto five-star
product rating. However, SDGFS was unable to provide documentary evidence of
the motivation for these changes. In addition, we were told that the monthly
benefit received in the event of a claim was significantly higher than that described
in policy documents available to consumers. SDGFS told us that this was an
internal decision for its own administrative convenience. However, in our view,
having made the decision to improve the monthly benefit, not making this known
to customers was inconsistent with competition on non-price elements of the
product—we would expect SDGFS to publicize these higher benefits in order to
win or retain customers if it was competing to win customers based on non-price
factors.
6.88 We received mixed views on the extent to which changes were due to competition.
SDGFS said all enhancements to cover it had introduced over the years were made
in response to competitive pressure from other retailers. JD Williams said that, of the
changes it had made to terms and conditions of its policies, the only changes it had
made as a result of regulatory requirements or changes were to bring the life age
limit down (January 2005) and to change the method of calculating the monthly
premium (May 2008) (an increase in the life age limit was made at the same time in
response to regulatory changes). The British Retail Consortium (BRC) accepted that
many changes had been compliance-driven, but said that great care should be taken
to avoid the conclusion that these changes would not have happened without the
compliance or regulatory pressures.
6.89 We concluded that there had not been significant changes to non-price factors made
in response to competition. We found that many changes to non-price factors were
due to regulatory or compliance issues. Further, there were no material changes to
the actual monthly benefit paid to claimants in the period 2002 to 2007, 35 and we
found no evidence that retailers attempted to win customers by promoting these
35
SDGFS changed monthly benefits in 2000 and Otto changed benefits in 2007, but withdrew retail PPI a week later.
146
changes. We therefore formed the view that distributors do not compete to any
significant degree on the non-price aspects of retail PPI.
Advertising
6.90 We looked at evidence of how retail PPI distributors advertised and marketed their
policies. We found that there were three main channels used to promote retail PPI:
(a) Retail PPI distributors have regular contact with their customers through cata­
logues and continued purchasing of the retail goods and these points of contact
can be used to advertise their retail PPI policies. SDGFS advertises, and Otto
advertised, retail PPI products in their catalogues. Express Gifts advertised twice
in its catalogues between 2002 and 2007. For SDGFS and Otto, these advertise­
ments are, or were, full-page advertisements, equivalent to advertisements for
other insurance products.
(b) Distributors can also market their products alongside the monthly statements they
send out for their credit account customers. JD Williams sends all of its credit
account customers information about retail PPI and the opportunity to purchase it
as part of their third credit account statement.
(c) All the retail PPI distributors, except Otto, use their websites to advertise retail
PPI. This is usually found in the financial services part of the website, alongside
advertisements for other financial products such as general insurance and loans.
Links to retail PPI information are also provided for customers shopping online
during the payment process.
6.91 We found that the degree and nature of advertising of retail PPI was different from
other forms of PPI. First, the distributors do not rely on advertising of credit, and
credit prices in particular, to sell both credit and retail PPI and subsequently market
retail PPI at the credit point of sale. Instead, they must attract customers to their retail
offering more generally and not by any single price. Second, they generally give an
equivalent amount of advertising space to their retail PPI products as to other
financial products, for example alongside other insurance products in the back of
catalogues.
6.92 The cost of retail PPI in the advertisements is usually stated in a comparable format
as a price per £100 balance, although it is sometimes stated as a price per £1
balance or accompanied by price related offers, such as one month of free cover.
6.93 We did not find any comparative advertising for retail PPI or advertising of the
changes made to the scope of policies. 36 Given that all retail PPI policies (bundled
with merchandise cover) are sold at around the same price (see Table 6.6, where
prices are clustered at 1.98 to 1.99 per cent), advertisements are only useful for
customers to compare retail PPI policies if they give information about the different
benefits, such as the level of monthly payments or life payment.
6.94 We found some variation between different distributors in the accessibility of
advertising and information about prices. For instance, JD Williams provides infor­
mation about retail PPI in the ‘Terms and Conditions’ part of its website, rather than
36
SDGFS said that the numerous innovations in the cover it offered had been fully advertised in the catalogues. SDGFS does
advertise its shopping insurance policies. However, we were not satisfied that it advertised changes to its policies in a way
which highlighted improvements with a view to gaining customers on non-price features; as noted in paragraph 6.87, it did not
alert customers that the benefits paid out had significantly improved.
147
with other financial services. 37 In contrast, SDGFS provides information about retail
PPI, including pricing information, as part of its ‘Your money’ section of the
Littlewoods website. We also found that sometimes the merchandise cover part of
the product was given greater emphasis within advertisements than the retail PPI
element.
6.95 We found that the level of advertising of retail PPI policies was comparable to other
financial services and to the credit offering from the retailers. We also found that
relevant price information was included with these advertisements. We therefore
found that the degree and nature of advertising of retail PPI did not indicate a lack of
competition.
Price dispersion
6.96 We found (see paragraphs 6.79 and 6.93) that prices of individual retail PPI policies
did not vary much over time and that they all had about the same price. In a
competitive market where consumers search effectively, it might be expected that
such similarities in price of differentiated products could be explained by similarities
in non-price features of the products, but in this case the evidence showed material
variations in the non-price features. Table 6.8 shows the variation in benefit that is
received for each of the policies for four different claim events as given in the policy
summary documents. The higher-than-advertised level of monthly benefit for
unemployment or sickness under the SDGFS policy is not given in the policy
documents, but we have been told that this is the current level of monthly benefit.
TABLE 6.8 Benefits for claim events where a £500 balance is held at the time of event
Distributor
SDGFS
JD Williams
Otto
Express Gifts
Price per £100
of outstanding
balance
£
1.98
1.99
1.99†
1.95
Monthly benefit for
unemployment or
sickness
£
100–167*
30
42
52
Not working and
hospitalized for 3
weeks
£
167
1,000
500
52
Life
£
Merchandise
cover
(years)
500
500
500
500
3
5
3
2
Source: CC calculations, based on information provided by retail PPI distributors.
*This range depends upon whether an item was bought from Littlewoods or Littlewoods Direct.
†The Complete Cover product had different prices for customers who are over 65 and/or unemployed.
6.97 There were trade-offs between some events and policies. For instance, the Otto
policy had a higher monthly benefit but lower hospitalization benefit compared with
the JD Williams policy. In other cases, there were policies that were clearly of a
higher quality, for instance the SDGFS policy has higher benefits than the Express
Gifts policy. Whilst we are only able to compare for four products, there did appear to
be a lack of correlation between price and quality between policies.
6.98 This is reflected in the relatively large differentials in claims ratios among retail PPI
underwriters. We found that the claims ratios for two of the three largest underwriters
of retail PPI in 2006 were 9 per cent for [] which underwrites [], and 30.6 per
cent for [] which underwrites [].
37
Using the website: www.jdwilliams.co.uk/policies?decoration=true&finalTarget=policies_pomp on 11 August 08.
148
Profitability
6.99 The CC guidelines 38 state that it is normally helpful to consider the effectiveness of
competition by examining the outcome of the competitive processes in the particular
market. One of these outcomes is profitability. The section in the CC guidelines on
profitability states that a situation where, persistently, profits are substantially in
excess of the cost of capital for firms that represent a substantial part of the market
could be an indication of limitations in the competitive process. 39
How we examined profitability of retail PPI distribution
6.100 We examined the profitability of retail PPI in the same way as we did for other types
of PPI in Section 4, by identifying the revenues and costs incurred in the distribution
of retail PPI, as distinct from those incurred in the sale of the goods or the retail
credit. As described in paragraph 6.15, we noted that retail PPI was sold separately
from the sale of the goods and the retail credit as well as at the opening of a credit
account. Retail PPI distributors were generally able to separate out revenues and
costs relating to retail PPI distribution, although in some cases revenues of retail PPI
were combined with merchandise cover, and costs were shared with other activities
of the distributor.
6.101 We focused on the economic profitability of retail PPI (the post-tax profits less a
charge to reflect the cost of capital required to support this business line), as well as
calculating return on equity capital, being post-tax profits divided by equity capital.
We thought that the use of a combination of economic profits and return on equity
capital afforded an economically meaningful picture of both the absolute and relative
(in relation to the cost of capital) level of profits earned from retail PPI distribution.
6.102 We used the revised Market Economics Model as in Section 4, collecting data from
the four largest distributors of retail PPI. We considered the model suitable for our
analysis of the profitability of retail PPI. We considered the following inputs for the
revised model:
• the income earned by distributors;
• the operating costs associated with the distribution of retail PPI; and
• the appropriate capital base and cost of capital for the distribution of retail PPI.
6.103 As retail PPI is all regular premium, there is generally no interest income earned on
retail PPI premiums although there may be immaterial amounts earned when
balances are not paid in full.
• The income earned by distributors
6.104 As explained in more detail in Section 4, the distributor retains commission,
expressed as a percentage of GWP, on each policy sold. The distributor may also be
entitled to a profit share payment in the event that the amount paid out in claims is
less than expected. 40
38
CC3, paragraph 3.78. CC3, paragraph 3.82. As explained in more detail in Appendix 7.1 (paragraphs 70&71), the distributor is not usually liable for any losses in the event that the amount paid out in claims is more than expected (subject to contractual changes). 39
40
149
6.105 We examined data on the GWP each distributor collected from the sale of retail PPI
policies between 2002 and 2007. We found that retail PPI is always sold with
merchandise cover by all the distributors we looked at (see paragraph 6.9), 41 and
that the figures the distributors had provided, with the exception of [], included
merchandise cover. In order to examine the profitability of retail PPI, we needed to
isolate the GWP received from retail PPI only. Distributors told us how much GWP
per £100 outstanding balance per month related to retail PPI and how much to
merchandise cover (see paragraph 6.16). Using this information we estimated, on a
distributor-specific basis, that in 2007 total GWP relating to retail PPI was
approximately £73 million, as shown in Table 6.3. 42
6.106 We collected data on the income (in the form of commission and profit share) that the
distributors received from the sale of retail PPI between 2002 and 2007. This is
shown in Table 6.9. We used the data to calculate the percentage of GWP that the
distributors received as income. We found that distributors received income of
between 52 and 96 per cent of retail PPI GWP, 43 and received income on an
aggregate basis of approximately 70 per cent of retail PPI GWP.
TABLE 6.9 Income from retail PPI as percentage of GWP for each distributor, 2002 to 2007
SDGFS
JD Williams
Otto
Express Gifts
Aggregate
2002
2003
2004
2005
2006
2007
[]
[]
[]
[]
68
[]
[]
[]
[]
67
[]
[]
[]
[]
68
[]
[]
[]
[]
68
[]
[]
[]
[]
70
[]
[]
[]
[]
70
Source: CC based on information provided by the parties.
6.107 We examined the contracts between retail PPI distributors and their underwriters in
order to collect data on commission rates and the share of profit. We found that
commission rates ranged between [50–60] and [70–80] per cent and that profit share
ranged between [60–70] and [90–100] per cent. 44 Details of each contract that we
looked at are set out in Table 6.10. This data is useful as it corroborates the data on
typical distributor income levels shown above.
TABLE 6.10 Summary of financial arrangements
Underwriter
[]
[]
[]
[]
Distributor
Date of contract
2007 GWP
£m
Commission
%
Profit share
%
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
[50–60]
[65–75]
[70–80]
[65–75]
[60–70]
No profit share*
[90–100]
[85–95]
Source: CC based on contracts provided by the parties.
*[]
6.108 For the purposes of the revised Market Economics Model for retail PPI, we used the
aggregate figures as shown in Table 6.9 for the share of income estimate. We noted
that the agreements between distributor and underwriter were structured differently
41
With the exception of Otto which briefly sold retail PPI policies without merchandise cover. This is [60–80] per cent of the GWP originally reported to us. 43
[] told us that owing to timing considerations, any profit share element was not necessarily booked in the same year as the premium was written, and that a coincidence of circumstances could give rise to an apparently high income figure. 44
Commission rates are expressed as a percentage of GWP; profit share rates are expressed as a percentage of the profit, equal to earned premiums less commission paid, claims paid and retention. 42
150
for each contract, and often had complicated arrangements with regard to profit
share payments. We based the model on actual income levels reported to us by
distributors rather than contractual rates to ensure that we used actual income
figures received by each distributor regardless of how the contractual agreements
were arranged.
• The operating costs associated with the distribution of retail PPI
6.109 We asked each distributor for information on costs associated with the distribution of
retail PPI. Even though some costs of supplying retail PPI were shared with the cost
of supplying the goods or retail credit, distributors were generally able to provide us
with their estimates on the costs of supplying retail PPI.
• Evidence from each of the parties
6.110 Otto stopped selling retail PPI in November 2007, but before that date it sold retail
PPI in-house, through telephone channels only (both inbound and outbound). Otto
told us that the direct costs of selling retail PPI were £[] per new sale, and that
ongoing compliance and servicing costs totalled £[] in 2007/08. Based on sales of
[] new policies in 2007, this equated to an approximate total cost of £[40–60] per
sale.
6.111 As with other parties to the investigation, JD Williams told us that it did not have data
solely concerning the profitability of retail PPI, as it was such an integral part of its
customer accounting process and service. However, it did tell us that the costs of
selling retail PPI were broadly similar to those for other forms of PPI. It was able to
provide us with a detailed costs schedule showing costs of sales, regulation,
servicing and claims for the retail PPI business. JD Williams used an external
telemarketing company to make the outbound sales calls so these costs were easy
to identify: based on [] policies sold in 2007 and £[] costs of external
telemarketing, the direct cost per policy sold was £[]. JD Williams then allocated
payroll, marketing, professional and overhead costs, resulting in a selling cost per
policy of £[120–140]. JD Williams included a proportion of customer recruitment
costs, on the grounds that without new customers JD Williams would not have any
customers to sell retail PPI to. We thought that the costs of selling retail PPI should
rightly include some measure of relevant joint and common costs incurred by the
business, but we thought that customer acquisition costs were attributable wholly to
the sale of goods and credit. We did not consider these costs to be related to retail
PPI and there was no compelling conceptual reason for allocating a proportion of
these to retail PPI, thus we deducted these costs from the schedule JD Williams
provided. We wanted to include some measure of the cost of servicing, regulation
and claims handling for the retail PPI business, which were not included in the cost
per policy of £[120–140] calculated by JD Williams above. Adding these costs, we
estimated the total cost of supplying each new retail PPI policy sold in 2007 as
£[130–150]. We noted that it had also included costs of bad debt, equivalent to []
and [] per cent of 2007 and 2008 GWP respectively.
6.112 Express Gifts told us that it sold retail PPI via the telephone, as well as via postal and
Internet orders, at the same time that the customer placed the order for goods and
retail credit, but that it suspended telephone retail PPI in November 2007. It provided
cost data, which showed the cost of selling, compliance and handling enquiries and
claims, as approximately £[] per new policy sold. Express Gifts told us that the cost
data provided only included direct costs, with no overhead allocation.
6.113 SDGFS provided a schedule containing the costs it estimated of selling, compliance
and servicing of retail PPI policies, showing a cost of £[250–300] per sale. We
151
disagreed with its approach for two reasons. First, it allocated categories of cost
which were unrelated to the distribution of retail PPI such as warehousing costs on
the basis that, if the company did not incur costs for distribution, warehousing, and
customer support, it would not be able to sell any of the goods and services shown in
its catalogues and on its website, and without these sales then there would not be
any customer balances on which to offer a retail PPI product. We thought that the
costs of selling retail PPI should rightly include some measure of relevant joint and
common costs incurred by the business. However, we decided that costs such as
warehousing, distribution and bad debt were attributable wholly to the sale of goods
and credit and that there was no compelling conceptual reason for allocating a
proportion of these costs to retail PPI.
6.114 Second, we disagreed with SDGFS’s method of apportioning costs to retail PPI. The
large majority of costs were apportioned on the basis of GWP received on retail PPI
as a proportion of total turnover of the business. 45 We did not believe that an
apportionment based on turnover was appropriate when the price of retail PPI itself is
under examination. This is because, if retail PPI prices were to be excessive, and
hence sales figures inflated, then apportioning costs pro rata to sales could lead to
an excessive proportion of costs being allocated to retail PPI, thus understating retail
PPI profitability. We considered that a more appropriate method would have been to
apportion costs on the basis of activity, ie activity-based costing. An activity-based
costing system seeks to identify cause and effect, in order to assign costs to products
objectively; these causes are called cost drivers. Once costs of the activity have been
identified, the cost of each activity is attributed to each product to the extent that the
product uses the activity. Thus, for example, a product which is not marketed would
not have any marketing costs attributed to it. 46
6.115 Based on this methodology, we attempted to estimate a reasonable fully-allocated
cost of distributing retail PPI for SDGFS. We did not include the costs of ware­
housing, distributing the goods, or support services on the basis that they were not
attributable to the distribution of retail PPI. On the basis that the marketing of retail
PPI was limited to a one-page advertisement in the catalogue and a page on the
website, we also included a proportion (10 per cent) of the £[] million annual costs
of marketing which SDGFS had allocated to retail PPI, which we believe more than
covers this expenditure. 47
6.116 On this methodology, we calculated the cost per new policy sold as £[100–150]. We
regarded this as an upper estimate as it apportions all costs (other than the costs
adjusted for above) on the basis of turnover; we consider that many of the costs
incurred in the sale of retail PPI are minimal compared with the sale of the retail
goods or credit.
6.117 We decided that it was likely that the cost per new retail PPI policy sold was
significantly less than the cost we calculated. First, as mentioned above, we consider
that many of the costs incurred in the sale of retail PPI are minimal compared with
the sale of the retail goods or credit. Second, we compared the cost of acquiring a
home-shopping customer to the cost of a new retail PPI policy. SDGFS told us that
the cost of recruiting a new customer to home shopping was between £[60–100],
based on 2007/08 expenditure on recruitment and marketing (including television
45
The exception to this basis of apportionment were the costs of the SDGFS call centre (for existing customers), based on the
estimated actual time spent making the PPI offer, although SDGFS did not show us the underlying calculations. The costs of
the second contact centre (for customers making their first purchase) were allocated on the basis of turnover, however.
46
The BRC stated that, in relation to the ‘activity based costing’ model, warehousing and distribution costs, as well as an
element of bad debt, should be included in the costs of PPI.
47
We included the same proportion of costs relating to a subsidiary which undertook all the printing operations, mailing and
fulfilment of customer orders (customer statements etc) for the same reasons.
152
advertising and direct mail). Given that retail PPI is sold only to customers already
recruited to home shopping, we thought that the costs of recruiting a new customer to
retail PPI would be lower than the cost of recruiting a customer to home shopping.
Finally, we noted that SDGFS stated that the costs of distributing retail PPI were not
substantially different to other forms of PPI. As noted in Section 4, we estimated the
cost of distributing PPI as £100 per policy. However, SDGFS does not have an
expensive branch network to support, instead operating via call centres and the
Internet.
6.118 Following publication of our provisional findings report, SDGFS provided us with a
report prepared by OXERA assessing our analysis of the profitability of retail PPI.
6.119 It presented an alternative calculation for a fully-allocated cost figure for the distribu­
tion of retail PPI. First, the direct costs for merchandise and financial services
(including insurance) were calculated and used to allocate common costs from the
home-shopping division between these two areas. The direct costs were equal to the
sum of the cost of goods sold and other identified direct operating expenditure. The
cost of goods sold for the insurance products were estimated to be the fees paid to
the underwriter for all insurance products. Common costs for merchandise and
financial services were then allocated based on the relative sizes of their costs of
goods sold and other direct costs.
6.120 Second, the common costs allocated to financial services from the home-shopping
division and the costs from financial services that were common to the distribution of
credit and insurance products were allocated between credit and insurance. The
allocation was based on the direct costs estimated as described in paragraph 6.119.
Third, costs identified as common to all insurance products were allocated based on
the shares of insurance commission income. The methodology employed by Oxera
results in the cost of distributing retail PPI as approximately £[180–220] per policy.
6.121 We decided that this approach overestimates the direct cost of distributing retail PPI.
The distributor acts as agent in the sale of retail PPI. In practice, the distributor
collects the GWP on behalf of the underwriter, retaining the income due to it (in the
way of commission and profit share) and passing on the remainder of the GWP to the
underwriter. In any case it does not necessarily follow that using the proportion of
direct costs should be the basis for allocating all other costs. 48 On the methodology
used by Oxera, common costs are allocated based on a very high proportion of cost
of goods sold, thus tending to overstate the fully-allocated cost of distributing retail
PPI. Our judgement is that our approach gives a fairer result.
• Figure to include in the revised Market Economics Model
6.122 The range of fully-allocated costs of distributing retail PPI was between £[100–150]
and £[100–150] for JD Williams and SDGFS (with £[40–60] for Otto). We took the
upper end of the range, which if too high, would clearly understate profitability, at
£[100–150].
6.123 We were aware that the estimates we noted above included the cost of selling
merchandise cover. However, since retail PPI and merchandise cover are sold as
one product, there is no obvious way to split the costs between them. Given that it
appears that there is no extra effort made in selling retail PPI or merchandise cover,
something approaching a 50:50 allocation would be reasonable. Our judgement was
to split the costs in proportion to GWP received by the distributor for retail PPI and
48
This is because different company structures may result in different costs being classified as direct or indirect.
153
merchandise cover, that is, [60–80] per cent 49 to retail PPI and the remainder to
merchandise cover. Thus, we include a cost of £96 ([60–80] per cent of £[100–150])
in the revised Market Economics Model, which for the reasons set out above, we
consider to be towards the upper end of the range, which, if too high, would clearly
understate profitability. We carry out sensitivity analysis on this figure below.
• The appropriate capital base and cost of capital for the distribution of retail PPI
6.124 We needed to estimate an appropriate capital base and cost of capital in order to
estimate a capital charge for the purposes of assessing economic profit. An
appropriate cost of capital was also needed to compare against the RoE in the
revised Market Economics Model.
6.125 We considered that some capital should be allocated to the activity of retail PPI
distribution. We noted that conceptually the capital requirement of a business should
be irrespective of ownership and that we were attempting to assess the profitability of
the retail PPI offering as a stand-alone product, separate from the retail offering. We
therefore decided to use an estimate of 12 per cent of retail PPI income to reflect
total economic capital in respect of the retail PPI distribution activity, the same figure
as in Section 4. We note that what we call equity capital in our previous analysis of
the profitability of PPI distribution could also be called total capital, as it includes an
allowance, which in our judgement, more than meets the need for intangible assets.
6.126 Similarly, we wanted to estimate a retail PPI-specific cost of capital. We noted that
the cost of capital attached to the nature of the activity, rather than the ownership. In
estimating the cost of capital, we decided to use a figure of 10 per cent, which was
the same figure as in Section 4.
6.127 Oxera’s report prepared on behalf of SDGFS criticized our approach to estimating
the capital base, stating that the appropriate capital base should reflect assets
required for the activity (such as business infrastructure and the customer base),
rather than financial risks, as implied by the use of capital requirements. It presented
an alternative approach to the capital base which showed that profitability was lower
than the CC’s estimates. Under this approach the capital base was defined with
reference to the assets required for the activity of distributing retail PPI, including
capitalized customer acquisition costs and IT infrastructure costs as intangible
assets. The capital costs of software development were capitalized over what was
assumed to be their useful life (five years); revenue costs relating to customer
acquisition costs (including marketing and promotion) were allocated to retail PPI on
the same basis as set out in paragraphs 6.119 and 6.120 and capitalized over what
was estimated to be the average life of a customer account. The capital base
calculated by Oxera amounted to approximately £[] million, equivalent to [150–
200] per cent of SDGFS’ retail PPI revenues in 2007.
6.128 We disagreed with Oxera’s approach to the capital base and its criticisms of our
approach for four reasons. First, our approach to calculating the capital base did
acknowledge the need to include intangible assets. Although taking as a starting
point the capital requirements of PPI distribution by a financial institution, we believe
that our assessment of total capital is sufficient and includes an adequate allowance
for intangible assets. Second, we noted that Oxera’s approach to allocation of
revenue and capital costs was sensitive to assumptions about cost allocation, which
we already discuss at paragraph 6.121, and estimates of useful life. It also included
costs of marketing and promotions which we do not necessarily consider appropriate
49
See paragraph 6.105.
154
to be capitalized; it is not clear whether the expenditure was effective. 50 Third, we
noted that Oxera’s calculations resulted in a capital base equivalent to [150–200] per
cent of SDGFS’ revenues, which we considered to be considerably in excess of any
reasonable estimate; our judgement is that the capital base involved in operating a
retail PPI distribution business is low and that our estimate is a fair estimate of that
capital base. This is on the basis that we believe (a) only a small amount of capital is
needed to operate a PPI distribution business and (b) the risks involved in distributing
retail PPI are low. Finally, in terms of evidence on capital base, in our judgement the
methodology used by a distributor of PPI ([]) in the course of its business was
preferable to an analysis prepared for the purposes of our inquiry.
Results of profitability analysis
6.129 Figure 6.1 is a schematic representation of the revised Market Economics Model for
retail PPI distribution, using the assumptions discussed in further detail above. The
model calculates that the four retail PPI distributors made economic profits of over
£12 million in 2007, representing an RoE of 213 per cent. This is greatly in excess of
any reasonable estimate of the cost of equity.
FIGURE 6.1
Profitability of retail PPI distribution
GWP
Post-tax profit
£ million
x
13
Pre-tax profit
£ million
Revenue
£ million
51
19
£ million
73
Comm/profit share
70
rate %
–
Tax rate
%
30
Economic profit
12.3 –
£ million
Allocated equity
£ million
Capital charge
0.61
£ million
x
Cost
£ million
Cost per policy sold
96
£
32 x
Number of policies
0.337
million
6
x
RoE
213%
Cost of equity
%
10
Source: CC analysis based on the [] Market Economics Model.
6.130 We carried out sensitivity analysis to examine the effect of changes to the cost per
policy. If we allocated the combined cost of selling retail PPI and merchandise cover
£[100–150] the RoE would still be very high, at [45–50] per cent, with economic
profits of over £2 million. For economic profit to be zero, costs per policy sold would
need to increase to £148. We note that Oxera’s calculations show a cost per policy
sold of £[180–220] which in this context would appear unrealistic (see paragraph
6.121).
6.131 Oxera’s report prepared on behalf of SDGFS stated that the profitability estimates
obtained under our approach were highly sensitive to our assumptions underlying the
50
We believe that marketing expenditure that is not effective does not create an asset.
155
cost allocation. 51 We disagreed; as noted above, for the economic profits to be zero,
costs per policy sold would have had to increase by 54 per cent. We believe that our
results are sufficient to draw the robust conclusions we set out below in paragraph
6.134.
6.132 Oxera also suggested that there was an alternative approach to calculating capital
base, as outlined in paragraph 94 of Appendix 4.4: business risk capital was
allocated at [20–25] per cent of costs; and operational risk capital was calculated at
[5–10] per cent of revenues across all distribution activities, including PPI. Using
these figures in the revised model of economic profitability results in a capital base of
£10.5 million (equivalent to 21 per cent of revenues), with economic profits of £11.9
million and RoE of 123 per cent.
6.133 Following publication of our provisional findings report, JD Williams provided us with
calculations it had carried out using our revised model of economic profitability. It
calculated that, for 2007, economic profit was £[] (equivalent to an RoE of 17 per
cent); for 2008, it forecast economic loss to be –£[] (equivalent to an RoE of –54
per cent). We noted that it had included customer acquisition costs in its calculations;
we disagreed with the inclusion of these costs in the distribution of retail PPI as set
out in paragraph 6.111. We also noted that it forecast a 60 per cent increase in the
number of policies sold with only a small increase in GWP. We calculated economic
profit and RoE for JD Williams for 2007 and forecast for 2008 to be 104 per cent and
35 per cent, which, although lower than those found in our revised Market Economics
Model for retail PPI distribution, were still in excess of any reasonable estimate of the
cost of equity.
Conclusions on profitability
6.134 On the basis of the above model, including the assumptions discussed in paragraphs
6.106 to 6.129, we concluded that retail PPI distribution is highly profitable. Commis­
sion levels are high in relation to the costs incurred in selling retail PPI and the capital
requirements are low, reflecting the low-risk nature of the activity.
6.135 Although the RoE we found was extremely high compared with the cost of capital, it
was lower than in Section 4 where we found RoEs of almost 500 per cent. 52 We
considered this to be because distributors’ revenues are directly linked to the level of
GWP per policy (through the fixed percentage commission rate) and these are lower
for retail PPI than for other types of PPI. We found that the average income per
customer per year was no more than £41 53 for all retail PPI distributors, compared
with £160 for a PLPPI policy; 54 however, the costs per policy are not materially
different.
6.136 We looked at commission rates over the last six years and found that there had been
no material change. We had no evidence to suggest that costs had been materially
higher in previous years, and we therefore concluded that these levels of profitability
had persisted over at least the last five years.
51
Oxera carried out a sensitivity analysis which resulted in a range of RoEs of between –9 per cent and 347 per cent. It argued that this showed a high degree of uncertainty in the CC’s estimate of returns. We disagreed that its results implied that con­
clusion because Oxera's range resulted from putting all variables to their extreme values at once. In our view it is extremely unlikely that all variables will be at the end of their ranges all at once. 52
For 2006. 53
Source: based on £[] average GWP per year for [] of which [60–80] per cent is retail PPI, multiplied by [] per cent (commission and profit share as a percentage of GWP). 54
For a £2,000 personal loan repaid over 12 months. See Table 1 in Appendix 5.1. 156
6.137 The BRC challenged the validity of some of the figures used for the calculation of
RoE but told us that, whatever the true figure of profitability of retail PPI distribution
was currently, the economic downturn would ensure that future profitability would be
seriously affected over at least the medium term. JD Williams told us that our
approach to measuring profitability failed to take into account fluctuations in the
economic cycle. It told us that the nature of retail PPI and the risks it insured meant
that profitability would vary significantly according to the economic cycle and that the
CC had assessed profitability in a period of substantial economic growth and high
levels of employment. However, as also stated in paragraph 4.81, we thought that
the prices of PPI policies could be adjusted in response to changing economic
conditions.
6.138 SDGFS told us that retail PPI revenues had been and were still declining
significantly, and that a variable costing approach was not appropriate in the light of
this. SDGFS had told us earlier in the inquiry that its costs were variable. In our
judgement only a small proportion of costs in are likely to be fixed; we believe that
fixed costs are generally only considered to be fixed over the short term and that if
revenues were to continue to decline in the medium to long term a company would
cut its relevant fixed costs.
Recent developments in the supply of retail PPI
6.139 The indicators on price competition, non-price competition and profitability led us to
conclude that there was a lack of competition. However, we also noted that two firms
stopped selling retail PPI since we started our investigation. In November 2007, Otto
stopped selling retail PPI completely and Express Gifts stopped selling retail PPI by
telephone, although it continues to offer retail PPI for online customers. We wanted
to understand why some distributors had stopped selling retail PPI despite the
evidence we had seen that there is a lack of competition and it was highly profitable.
6.140 We asked Express Gifts and Otto for an explanation of why they had stopped selling
retail PPI. We noted that profitability was only one of several important reasons.
• Express Gifts raised the issue of opportunity costs, telling us that it had to balance
the time it took to sell a retail PPI policy with the time taken in offering alternative
products to its customers, as it only had a limited amount of time when speaking
to them on the telephone. It explained that its decision to stop selling retail PPI
was to shorten telephone call lengths to handle more effectively an increased
volume of calls in the run-up to Christmas 2007. [] and retail PPI was replaced
by a non-insurance product. It is currently reviewing its arrangements for retail
PPI. Express Gifts continues to offer retail PPI for Internet customers.
• Otto told us that, []. Sales of retail PPI were not profitable in the short term
because of the costs associated with the selling process relative to the volumes
sold. As a consequence, the company made a pragmatic decision and decided to
focus call centre sales activity on products that had a profitable income stream in
the shorter term []. Otto’s conclusion was that it had not found a way of selling
retail PPI which was both cost-effective in the shorter term, and compliant.
6.141 We took the view that the high-street banks and other providers of PPI which we
examined in Sections 3 to 5 dealt with, and were familiar with, compliance and
regulation as part of their core business, whereas the home-shopping retailers would
regard it as an additional business risk; firms were concerned about increasing
regulation and the risks of mis-selling and that this also had a bearing on their
decision to stop selling retail PPI. Retail PPI is not the home-shopping retailers’ core
157
business and is a minor part of their business with small absolute levels of profit­
ability, despite potentially high profit margins.
6.142 We concluded that there may be differences in the levels of income each firm could
expect from selling retail PPI and that this may have some bearing on whether a firm
would decide to sell retail PPI. Table 6.11 shows the average income earned per
customer in commission from premiums. There is a wide range in income received
from retail PPI, with SDGFS receiving [] times more income per customer than
Express Gifts.
TABLE 6.11 Average income per customer per year
Distributor
SDGFS
JD Williams
Otto
Express Gifts
Average income
per recruited customer
£
[30–40]
[30–40]
[10–20]
[10–20]
Source: CC calculations based on data provided by retail PPI distributors.
Note: This is the average income paid per retail PPI customer over the year. This does not reflect the pattern of premiums paid
over the year, for instance some may pay all in one month, and other spread premiums over the year.
6.143 We looked at how long it would take each distributor to recoup the costs of selling the
retail PPI policy, given the differences in the levels of income each firm could expect
from selling retail PPI. The costs of retail PPI to the distributor generally occur upfront
through selling the product, whilst income and profit accrues over several months
and years as customers pay their premiums. Firms may require profits to be earned
more quickly than this, even if they are at a lower level overall, and may prefer to sell
alternative goods and services that deliver more immediate profits.
6.144 We calculated that, given a combined average cancellation and termination rate of
19 per cent for all the parties, the average life of a policy was just over five years. 55
Based on the estimated selling cost of £96 (see paragraph 6.123), we calculated that
it would take the lengths of time shown in Table 6.12 for each distributor to recoup
the costs of selling the retail PPI policy. 56 If the average life of the policy was shorter
than the time it took to recoup the costs of selling the policy which were incurred up
front then it would not be worthwhile for the distributor to sell a policy. Table 6.12
shows that it is only worthwhile for SDGFS and JD Williams to sell a policy, with a
marginal result for Otto. It would not be worthwhile for Express Gifts to sell a policy.
TABLE 6.12 Payback periods
Distributor
SDGFS
JD Williams
Otto
Express Gifts
Length of time
[<3 years]
[<3 years]
[>4 years]
[>4 years]
Source: CC based on data provided by retail PPI distributors.
55
This assumes that the number of policies stays constant. We noted that Otto’s estimate of the cost of selling a policy was significantly lower (£[40–60]). However, we used £96 as this estimate was based on data provided by the largest providers of retail PPI, and was broadly in line with the cost we estimated of selling a policy which we found for other types of PPI policy (see Appendix 4.4, paragraph 56). 56
158
6.145 We also looked at the average retail PPI customer for each of the four distributors
and by using simple cash-flow assumptions, calculated the internal rate of return
(IRR) 57 for a single retail PPI policy given the same assumptions and figures as we
used above. Table 6.13 shows that, given an assumed cost of capital of 10 per cent,
it would not be a good investment decision for Otto and Express Gifts to sell a policy.
TABLE 6.13 Internal rate of return
Distributor
SDGFS
JD Williams
Otto
Express Gifts
IRR %
[40–50]
[20–30]
[0–10]
[negative]
Source: CC based on data provided by retail PPI distributors.
6.146 We found that this evidence was consistent with Otto and Express Gifts choosing to
cease sales of retail PPI, and did not alter our finding of high average profitability
across the firms that supplied retail PPI. Within any industry we would expect to see
a variety of levels of profitability between firms and over time, due to various factors
such as strategy, efficiency and share of supply. This does not affect our conclusion,
that the supply of retail PPI as a whole was highly profitable and that firms repre­
senting a substantial part of the share of supply had earned profits that were
persistently and significantly in excess of the cost of capital.
Conclusions on the extent of competition between distributors
6.147 The retail PPI evidence suggested that there was little competition between
distributors of retail PPI. We saw minimal price variation over time (paragraph 6.79)
and we found that product innovation was not motivated by competition with rivals
(paragraph 6.89). We found that the sale of retail PPI is highly profitable relative to
the cost of capital (paragraph 6.134). We balanced this evidence against the
withdrawal of two distributors from selling retail PPI. We thought that this behaviour
could be explained by the low absolute level of profit from selling retail PPI, and the
need to earn a more immediate return (paragraphs 6.141 to 6.146). The evidence led
us to conclude that there is not effective competition in the supply of retail PPI, and
that prices for retail PPI are higher than they would be in a well-functioning market.
Factors affecting the nature and extent of competition in the supply of retail
PPI
6.148 Having concluded that there is little competition between distributors of retail PPI
policies, we looked at why this is the case, and what factors are involved. The
purpose of this was to see if there were any features which are preventing, restricting
or distorting competition in the supply of retail PPI.
6.149 In Section 5, we focused on five factors which might contribute to the lack of compe­
tition. 58 We considered each of these factors with respect to retail PPI:
57
An IRR is an indicator of the quality of an investment, being the annualized effective compounded return rate which can be
earned on the invested capital, ie the yield on the investment. A project is a good investment proposition if its IRR is greater
than the rate of return that could be earned by alternate investments. Thus the IRR should primarily be compared with the
appropriate cost of capital (including an appropriate risk premium) and then to alternative projects.
58
See paragraph 5.2.
159
• the extent to which distributors actively seek to compete with each other;
• whether comparing retail PPI products is beneficial and whether there are barriers
to such searching;
• having taken out retail PPI, whether consumers switch between retail PPI
products, and whether there are barriers to switching;
• whether there are barriers to entry into the supply of retail PPI by new providers,
or barriers to expansion by suppliers already in the market; and
• whether there is an advantage for distributors to selling retail PPI and the credit
point of sale (the point-of-sale advantage) and whether this contributes to the lack
of competition we found.
Competition between retail PPI providers
6.150 In competitive markets we would expect firms to seek to win business by offering
desirable products (based on a combination of price and quality) and proactively to
advertise their products to customers. In a well-functioning market we would expect
to see a range of activities associated with active price and non-price competition
between suppliers. These could include: offering price reductions, discounts and
other sales promotions; advertising based on price; and comparative advertising,
along with other efforts, to simplify customers’ choices.
6.151 However, we have not observed these activities in relation to retail PPI. In particular:
• We saw little in the way of price competition. We saw no evidence of price
reductions in the period 2002 to 2007, and little evidence of price changes,
discounts or sales promotions to seek to win customers (see paragraph 6.79).
• We also saw little in the way of competition in relation to quality (see paragraph
6.89).
6.152 We conclude that, as with other forms of PPI, 59 most distributors fail actively to seek
to win retail PPI customers by using the price or quality of their retail PPI product as a
competitive variable. We conclude that this is a feature of the relevant markets which
prevents, restricts or distorts competition in the supply of PPI.
Search
6.153 We looked at whether customers benefit from searching for the best retail PPI
product, and whether there are barriers to effective searching.
Benefits to search
6.154 We observed minimal price variation between retail PPI policies, but we observed
significant variation in benefit calculations between policies suggesting that
consumers would still benefit from search (see paragraph 6.97). This was similar to
the conclusions reached for CCPPI 60 in Section 5.
59
See paragraph 5.8.
See paragraph 5.14.
60
160
Barriers to search
6.155 Whilst we found that there are benefits to search, we thought that there could be
significant barriers to search. We considered what these might be.
6.156 We focused on two possible barriers to search:
• the availability of, and ease of access to, information; and
• the ease of using information to compare products.
The availability of, and ease of access to, information
6.157 We found that although retail PPI distributors advertised their products and included
price information, there are no benefits to search with respect to the price of retail
PPI as all retail PPI policies were priced at around the same level. Instead, the
benefits to search will come from comparing the benefits of policies. Therefore, we
considered the availability of information about the benefits of policies.
6.158 We found that information on benefits, and therefore the differences between
policies, were not explained in advertisements and introductory pages of websites.
The detailed information on benefits, for instance the value of hospitalization cover or
monthly benefits for unemployment, could only be found by reading policy
summaries. To calculate monthly benefits may also require customers to read terms
and conditions for credit accounts to calculate the minimum payment due each
month, and therefore the benefit that would be received. For one distributor, SDGFS,
the monthly benefit was up to five times greater than that given in the policy summary
documents such that customers would not be aware of it even if they had
searched. 61
6.159 The practice of bundling merchandise cover and retail PPI into one combined policy
also makes it difficult for customers to evaluate value-for-money of retail PPI policies,
and contributes to a lack of transparency.
6.160 Finally, we note that Internet-based comparison tools, which are available for other
types of PPI, do not cover retail PPI.
6.161 We therefore believe that consumers face significant difficulties when trying to obtain
systematized information on retail PPI that hinders their ability to compare retail PPI
products.
The ease of using information to compare products
6.162 We found that retail PPI policies were not easier to understand than other regularpremium PPI policies, and that retail PPI customers would not have a higher level of
understanding or financial skill than customers of other types of PPI. Therefore, our
conclusions for other forms of PPI regarding ease of using information to compare
products also apply in relation to retail PPI. 62 We found that the variation in terms
and conditions and the way information is provided by firms mean that information is
61
The monthly benefit for accident, sickness and unemployment is given in the policy summary as ‘the scheduled payment’ due
on the credit account, which is 5 per cent of the balance for Littlewoods Direct customers. However, SDGFS told us that the
actual benefit was 33 per cent of the balance.
62
JD Williams said that it did not accept that customers have difficulty in assessing the value for money represented by retail
PPI policies, noting that cost is shown on monthly statements. We noted this, but in our view cost is only part of the value-for­
money calculation, as cost has to be related to the benefits obtained in order to take a view on value for money.
161
not available in a sufficiently usable format that enough consumers successfully
compare products to drive competition between providers. This may restrict the
ability of consumers to search effectively and to make an informed choice, and we
concluded that this is a significant barrier to effective search for retail PPI, as for
other forms of PPI.
Conclusions on search
6.163 We found that there are benefits for consumers in searching, but that customers face
significant barriers in comparing products. These barriers to effective search are
features of relevant markets which prevent, restrict or distort competition in the
supply of retail PPI. If the benefits of policies were clearer to consumers, and if retail
PPI were available separately from merchandise cover, distributors would have a
stronger incentive to compete with each other to show consumers that their retail PPI
product offer was the best value. This was similar to the conclusions reached for
other forms of PPI in section 5. 63
Switching
Barriers to switching
6.164 We considered whether there were any barriers to consumers switching, in addition
to the search barriers we have previously identified. We looked at: providers’ access
to relevant information; and exclusions.
Providers’ access to relevant information
6.165 For retail PPI, we thought that access to consumers’ credit account information
restricted the extent to which customers switch retail PPI on a stand-alone basis.
Although stand-alone PPI policies exist which could be used to cover outstanding
balances, these products are not linked to credit account balances. Customers
choose a level of insurance, which means that the level of insurance is not tied to the
actual amount owing on a credit account, and hence the customer is more likely to be
over- or under-insured in any given month than insured for the exact amount owing
(and hence the coverage is of lower quality than one which tracks the balance
exactly).
6.166 In addition to this, stand-alone providers do not have access to information on the
goods purchased by retail PPI customers. This makes it more difficult for them to
offer merchandise cover and consequently makes it more difficult for consumers to
substitute to stand-alone, as they have to substitute to a product without merchan­
dise cover.
6.167 We considered this to be a barrier to switching to stand-alone retail PPI policies.
Initial exclusion periods
6.168 We noted that, for retail PPI as for all types of PPI, the initial exclusion period at the
start of a policy before a consumer can claim limited switching. The evidence for
other types of PPI, 64 which also applied to retail PPI, led us to the view that for all
63
See paragraph 5.57.
See paragraph 5.61.
64
162
types of PPI the initial exclusion period can act as a deterrent to switching, as
customers are not fully covered during the initial period of a new policy.
Exclusions for pre-existing conditions
6.169 Similarly, we concluded that the exclusions in policies for pre-existing conditions
acted as a barrier to switching. 65 We agreed that for all types of PPI, including retail
PPI, for someone who has developed a condition which would be excluded from
coverage as a pre-existing condition under a new policy, the existence of exclusions
for pre-existing conditions constitutes a barrier to switching PPI policies.
Conclusions on switching
6.170 In addition to the costs of search already found (paragraph 6.163), we conclude that
there are three barriers to switching:
(a) the inability of stand-alone providers to offer insurance which tracks the
outstanding balance on a credit account, or to offer merchandise cover, means
that retail PPI policy-holders cannot find a perfect substitute for their policy;
instead they can only find a policy which insures a set amount, leading them to
be over- or under-insured in any given month and which does not include
merchandise cover;
(b) initial exclusion periods on retail PPI policies, which mean that someone
switching is not insured against certain eventualities at the start of the new policy;
and
(c) exclusions for pre-existing conditions, though only for those customers who have
developed conditions which are exempt from cover after they took out their
original retail PPI policy but before switching.
Barriers to entry and expansion
6.171 We considered whether there are additional barriers to entry or expansion for
suppliers selling stand-alone retail PPI. We received evidence on the barriers facing
stand-alone PPI suppliers in Section 5, which we concluded also applied to stand­
alone retail PPI suppliers. 66,67
6.172 First, we received evidence from four large underwriters that there are problems with
adverse selection in distributing stand-alone PPI. 68 The evidence provided suggested
that stand-alone PPI providers suffered from an increased risk of adverse selection,
because consumers who search for this type of insurance are more likely to make a
claim than customers who purchase PPI when applying for credit. This was exacer­
bated by retail PPI being sold alongside credit accounts. When retail PPI is sold to
existing credit account holders, distributors may experience some reduced risk of
adverse selection as they have access to all credit purchasers to whom they can try
65
See paragraph 5.62. See paragraphs 5.79–5.87. 67
SDGFS said that we could not take research relating to other products and apply it to retail PPI. We cover this general
argument in paragraph 6.49. In the case of this particular evidence, there are no specific stand-alone retail PPI providers; possible alternative stand-alone policies are the same policies as are possible alternatives for CCPPI.
68
See paragraphs 20–24 of Appendix 5.3. Adverse selection, sometimes referred to as anti-selection, occurs as a result of information asymmetries between the insurer and the customer. Customers who know they face a large risk of needing to make a claim on an insurance policy are more likely to take out insurance than those who face small risks; however, the insurer does
not know which customers fall into which category and so cannot price its insurance premium accordingly.
66
163
to sell retail PPI, rather than just those who approach them for retail PPI. In addition,
a consumer who is facing a potential loss of income may be less likely to take on
more credit. We concluded that adverse selection was a barrier to entry and
expansion faced by stand-alone PPI providers compared with distributors selling
retail PPI alongside credit accounts.
6.173 Second, we considered whether customer awareness of retail PPI and the options
available constituted a barrier to entry and expansion. For other forms of PPI we
were told that customers generally had a low awareness of the offers of different
credit providers. 69 The evidence suggested to us that low customer awareness of
PPI, and poor understanding of the options available, restrict the ability of stand­
alone providers successfully to enter and expand into PPI markets. We concluded
that CCPPI and retail PPI customers would be similar in their awareness of PPI and
the options available such that this would also apply to customers of retail PPI.
6.174 In order to overcome the poor awareness of the availability of stand-alone options,
providers must market their products. The evidence we received on marketing costs
to attract customers for a stand-alone policy was that they were significant. We
concluded that, with the current market structure and at current prices, the cost of
marketing stand-alone PPI to consumers is large if PPI products are not sold in
combination with a credit product, and we concluded that this also applied to retail
PPI.
6.175 We noted that providers of stand-alone PPI and short-term IP have not achieved
significant sales, and that some large underwriters (Axa, Aviva) have exited the
stand-alone market as a result of their experience. We found that only [] policies
for potential stand-alone retail PPI had been sold in a year (see paragraph 6.26), that
sales of short-term IP were under 15,000 per year (see paragraph 3.40) and that the
four largest distributors of retail PPI did not sell retail PPI to customers of other
retailers.
6.176 We concluded that there are significant barriers to entry and expansion for stand­
alone PPI providers seeking to sell retail PPI products without access to customers at
the credit point of sale and to customer account information. These issues are
discussed in more depth in the context of the point-of-sale advantage (paragraphs
6.177 to 6.184).
The point-of-sale advantage
6.177 We found that other forms of PPI were sold at the point of sale of the credit product
which they protected. 70 For retail PPI, there is an equivalent initial point of sale when
the credit account is opened with a retailer and, similar to CCPPI, regular billing when
the distributor will contact the customer.
6.178 However, there are also additional points of sale when the customer purchases other
goods from the retailer, either with or without using their credit account. Different
distributors sell retail PPI at different points in their relationship with the customer.
We considered whether the following different types of selling process restricted
competition:
• selling retail PPI at the initial point of sale of a credit account;
69
See paragraph 5.83.
See paragraphs 5.88–5.119.
70
164
• selling retail PPI to consumers with existing credit accounts at subsequent points
of sale of retail goods; and
• selling retail PPI by subsequently contacting customers with an existing credit
account.
6.179 SDGFS sells around half of its retail PPI policies to telephone customers at the initial
point of sale 71 when customers first contact the retailer to order goods and open a
credit account. Table 6.5 sets out the sales process for each retail PPI distributor. We
found that the penetration rates for [] were between two and three times greater
than for the other distributors (see Table 6.4).
6.180 Sales could be made when customers make a subsequent purchase of goods. In
these cases, customers have already used their credit account to pay for purchases
and so have had the opportunity to consider their protection needs and would not
believe that their credit application would be affected.
6.181 Distributors contact customers directly to sell retail PPI after the initial point of sale.
For instance, JD Williams contacts customers by telephone after they have received
three monthly statements and have an outstanding balance.
6.182 For each of these three selling points for retail PPI, the point-of-sale advantage is
weaker than for other forms of PPI. 72 First, customers are primarily purchasing goods
with a view to spreading the cost of purchase over time, rather than applying for
credit, and therefore are less likely to believe that their ability to purchase goods on
this basis was dependent on taking out retail PPI. 73 This is particularly the case for
sales not at the initial point of sale. We have no reason to think that customers
believed their application for a credit account was enhanced by taking out retail PPI.
Second, retail PPI is advertised in the same catalogues as the goods being pur­
chased and sold alongside other insurance products; consumers are therefore more
likely to be aware of retail PPI before the point of sale than they are with other PPI
policies, for which we found little advertising.
6.183 Despite these differences, we concluded that there remain substantial advantages
from selling retail PPI at the initial point of sale, and subsequently, by virtue of
supplying the credit to which the retail PPI relates. These are similar to the point-of­
sale advantages identified in Section 5 for other forms of PPI. First, the distributor
may more easily offer retail PPI that is tailored to the credit product, in particular the
ability to tailor premiums to the changing balance, as we found was particularly useful
for CCPPI. 74,75 Second, as with other forms of PPI, stand-alone providers, even if
they were able to identify holders of credit products, are likely to be offering retail PPI
to a customer who already has PPI or has been offered retail PPI and declined to
71
Express Gifts sold 86 per cent of its retail PPI policies at the initial point of sale, until it stopped using this approach in 2007.
The BRC said that distance selling took away the point-of-sale pressure and negated some of the advantage that may exist in
a face-to-face environment. We noted its view. We also noted that penetration rates of PPI policies sold by telephone are lower
than those for PPI policies sold face to face (see Appendix 5.3). We thought that selling PPI to customers face to face achieved
a higher degree of success than selling to customers over the telephone; however, we did not think that selling over the tele­
phone meant that there was no point-of-sale advantage, or that it sufficiently reduced the point-of-sale advantage to allay our
concerns.
73
See paragraph 5.94.
74
See paragraph 5.107.
75
JD Williams and SDGFS said that a stand-alone policy did not need to tailor the level of insurance to match the outstanding
balance. However, we thought that a policy that does insure the right balance offers a more effective solution to consumers
than one which either insures either too much or too little.
72
165
take it out. 76,77 Third, once the retailer has established its relationship with the
customers to whom it has provided credit, it has continued exclusive access to the
customer to communicate the availability of retail PPI in connection with its credit
account. 78
6.184 We found that the sale of retail PPI at the initial point of sale and continued exclusive
access to customer accounts restricts the extent to which other PPI providers can
compete effectively, and is therefore a feature of relevant markets which prevents,
restricts and distorts competition in the supply of retail PPI.
The effects on consumers of the level of competition in the market
6.185 Having found features that prevent, restrict or distort competition in the sale of retail
PPI products, we considered what the effects on consumers would be. First, as we
found that distributors have high margins, and enjoy high profitability, when selling
retail PPI, we concluded that consumers were paying prices higher than they would
in a well-functioning market, in line with the findings for other forms of PPI. 79
6.186 However, unlike our findings in relation to some other types of PPI, 80 we did not find
that high prices for retail PPI were significantly distorting prices for credit or retail
goods—all parties we asked told us that the income earned on retail PPI did not
affect the prices set for retail credit or the underlying goods.
Conclusions on factors affecting the nature and extent of competition
6.187 We conclude that there are features of relevant markets which, alone or in
combination, prevent, restrict or distort competition in the supply of retail PPI to nonbusiness customers in the UK. These are set out in paragraphs 6.148 to 6.186 and
may be summarized as follows:
(a) Distributors fail actively to seek to win customers by using the price or quality of
their retail PPI policies as a competitive variable.
(b) Consumers who want to compare retail PPI policies, stand-alone PPI or shortterm IP insurance policies are hindered in doing so. Product complexity (the
variations in terms and conditions, the way information on retail PPI is presented
to customers); the bundling of retail PPI with credit accounts and with merchand­
ise cover, and the limited scale of stand-alone provision act as barriers to search
retail PPI policies. In addition, the time taken to obtain accurate information on
benefits is a barrier in relation to the provision of retail PPI. These barriers to
search impede the ability of consumers to make comparisons, and therefore
effective choices, between retail PPI policies. They also, therefore, act as barriers
76
See paragraph 5.110. SDGFS said that this was not a significant advantage. It said that the existing provider had made the third party’s task easier
because the customer had already been given full information about PPI, so all the third-party provider needed to do was compare the two products and offer its product. It also said that many customers paid their balances at a post office or bank, where they were exposed to PPI advertising, negating the point-of-sale advantage. However, we noted that despite these points there were no third-party suppliers of PPI to cover retail account balances at the time of writing. We therefore disagreed with SDGFS’s view that access for third-party providers was easy and that paying a bill at a post office or bank negates the point-of-sale advantage. 78
See paragraph 5.111. 79
See paragraph 5.135. 80
See paragraph 5.140. 77
166
to expansion for other retail PPI providers, in particular providers of stand-alone
PPI.
(c) Consumers who want to switch retail PPI policies to alternative providers or to
alternative types of insurance policies are hindered in doing so. Terms which risk
leaving consumers uninsured (for a short period of time or in case they suffer a
recurrence of a problem) and the bundling of retail PPI with merchandise cover
act as barriers to switching. In addition, the lack of access to consumers’ balance
information acts as a barrier for switching for retail PPI as it renders stand-alone
providers unable to offer equivalent retail PPI policies. These barriers to switching
limit consumer choice. They also therefore act as barriers to expansion for other
retail PPI providers, in particular providers of stand-alone PPI.
(d) The sale of retail PPI at the initial point of sale, and the continued exclusive
access to credit accounts by distributors further restricts the extent to which other
providers can compete effectively.
6.188 We have found, pursuant to section 134(1) of the Act, that there are features of
relevant markets, which alone or in combination prevent, restrict or distort
competition in the supply of retail PPI, and accordingly there is an adverse effect on
competition within the meaning of section 134(2) of the Act. The features are those
that we have identified in paragraphs 6.148 to 6.186 and which are summarized in
paragraph 6.187.
6.189 We concluded that the detrimental effects on consumers of these features were
higher prices for retail PPI policies than would be expected in a well-functioning
market. We also concluded that it was possible that there was less innovation than
would be expected in a well-functioning market.
167
7. The underwriting of PPI
7.1 We looked at the extent of competition between underwriters of PPI policies. Our
analysis of this is set out below, and in Appendix 7.1.
Market definition
7.2 We considered the appropriate product and geographic market definitions, applying
the methodology set out in our guidelines. 1
Product market
7.3 In looking at the relevant product market we considered the extent to which there
would be demand- or supply-side substitution in response to a SSNIP.
7.4 Demand-side substitution occurs because an increase in price makes a product less
attractive to customers who therefore decide to purchase less of it and more of a
substitute. Supply-side substitution occurs when a price rise prompts other firms to
start supplying, at short notice, an effective substitute to the product in question.
7.5 The evidence we received indicated that contracts between underwriters and
distributors are usually specifically designed and negotiated for a particular credit
product, with significant input from the distributors. 2
7.6 We received mixed views from the parties regarding the relevant product market for
underwriting. One view put to us was that different types of PPI policies (such as
MPPI and PLPPI) represented different risks and in turn were priced differently. 3 This
would suggest that PPI providers could be limited in the degree to which they could
substitute between underwriters from different categories.
7.7 However, several underwriters told us that there was no reason why a firm
underwriting one type of PPI would not be able to underwrite all types of PPI
policies. 4 It was noted that MPPI might be slightly different as it typically would not
include cover for L; however, this was not considered to be a barrier to switching.
7.8 We noted that the larger underwriters typically underwrite several different types of
PPI and most underwriters told us that they were willing to tender for PPI policies
irrespective of whether they currently underwrite those policies. Evidence on tenders
suggested that this was the case. Table 7.1 shows the winners of a number of tender
processes about which we have good information, and how many unsuccessful
bidders there were.
1
CC3. For example, Lloyds TSB post-issues statement hearing summary, paragraph 2, HBOS response to issues statement, p4. Summaries of post-issues statement hearings—Cardif Pinnacle, paragraph 2; London General Insurance, paragraph 19. 4
Summaries of post-issues statement hearings—Aviva, paragraph 1; AXA, paragraph 1; Cardif Pinnacle, paragraph 1,
Genworth, paragraph 3; HSBC, paragraph 1. 2
3
168
TABLE 7.1 Examples of market tenders and parties that have responded
Parties issuing the tender Parties that submitted a bid
Lloyds TSB* Capital One MBNA BMW Financial Services (GB) Ltd
Countrywide Aviva and Lloyds TSB (Scottish Widows)* were successful in bids for different policies;
bids were received from three other underwriters; none of these were verticallyintegrated firms.
AXA was the successful bidder; bids were received from six other underwriters including
one vertically-integrated firm.
HBOS* (St Andrews Group) was the successful bidder; bids were received from six other
underwriters; none of these were vertically-integrated firms.
HBOS* (St Andrews Group) was the successful bidder; bids were received from three
other underwriters, including one vertically-integrated firm.
AXA was the successful bidder; bids were received from five other underwriters,
including two vertically-integrated firms.
Source: CC, based on information provided by the parties.
*Indicates vertically-integrated companies.
7.9 Data we received showed that over the period 2002 to 2005, 45 tender opportunities
arose with a total GWP of £3.32 billion. There were 120 responses to these tenders
submitted.
7.10 Table 7.2 presents a summary of bidding activity by underwriters in the market for
underwriting PPI.
TABLE 7.2 Summary of tendered opportunities for underwriting of PPI in the UK, 2002 to 2006
Aviva
AXA
Barclays
Cardif Pinnacle
Genworth
HBOS (St Andrews)
HSBC
London General Insurance
Lloyds TSB (Scottish Widows)
RBSG
Total (2002–2006)
Vertically-integrated
Independent
Bids submitted
Tenders won
GWP
£m
Success rate*
%
Average GWP
£m
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
120
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
45
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
3,319.7
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
37.5
[]
[]
[]
[]
[]
[]
[]
[]
[]
[]
73.8
26
94
12
33
1,163.9
2,155.8
46.2
35.1
97.0
65.3
Source: CC, based on information provided by the parties.
Notes:
1. Many underwriters were unable to supply complete and accurate information on their bidding activity. This table therefore
represents a summary of the information that is available.
2. Does not include contracts with GWP of less than £5 million. Does not include contracts to underwrite PPI for store cards.
7.11 There does not appear to be any significant barrier to an underwriter expanding its
business from underwriting one type of PPI policy to another. Most large PPI
underwriters provide PPI underwriting services for two or more types of credit, and
have bid in the last five years to provide PPI underwriting services for at least one
credit product they do not currently underwrite, showing a willingness to try to move
into the underwriting of new types of PPI policy. In addition, most parties we spoke to
told us that they thought that the underwriting market covered all types of PPI
policies.
7.12 We considered that no individual incumbent underwriter would be able profitably to
impose a SSNIP above competitive levels, for example when the contract is retendered, because this would induce the distributor to award the contract to another
underwriter. Similarly a hypothetical monopolist of one type of PPI would not profit169
ably be able to impose a SSNIP, since this would induce distributors to award their
tenders to another underwriter in another category. We therefore find that the
relevant product market is at least as wide as the underwriting of all types of PPI.
7.13 We did not find that there were separate markets for underwriting life and non-life
elements of PPI policies, nor that there were separate markets for PPI underwriting
depending on whether or not the underwriter was vertically integrated with a
distributor. Details of our analysis on these points can be found in Appendices 7.1
and 8.1.
7.14 We then considered whether the relevant product market might be broader,
encompassing the underwriting of some non-PPI insurance policies. Underwriters
told us that PPI was not a complex insurance policy to underwrite, relative to
investment insurance policies, 5 and that general insurance companies should have
little trouble entering the creditor insurance market. 6 There are two large independent
general insurance companies already active in the PPI underwriting market with
significant market share (AXA, Aviva).
7.15 The evidence we saw suggested that there are no fundamental barriers to
underwriters moving into the PPI market from other general insurance or life
assurance markets, though we are not aware of any significant new entry into the
PPI underwriting market in recent years.
7.16 We conclude, therefore, that the relevant product market for underwriting is at least
as wide as the underwriting of all PPI policies, and may be wider to encompass other
general insurance and life assurance policies.
Geographic market
7.17 We considered whether the relevant geographic market for underwriting is UK-wide,
narrower or broader.
7.18 We received no evidence to suggest that the geographic market is local or regional;
we saw no evidence of regional restrictions on underwriting contracts or tender
documents.
7.19 We received one piece of evidence to suggest a wider market—there have been
some tenders to underwrite risk for a pool of creditors in both the UK and the
Republic of Ireland.
7.20 There are different, albeit similar, regulatory regimes in place in the two countries,
and developments in these regimes, even if the developments are similar, will not
necessarily occur at the same time. Insurers underwriting risk in the UK must comply
with FSA rules and regulations. Potential entry into the UK market is constrained to
the extent that companies need to comply with FSA regulations. Although the
regulatory environment in the Republic of Ireland is broadly similar, it will nevertheless be the case that conditions of supply and demand will be to some extent
determined by the regulatory environment. 7 We concluded that this could militate
against underwriting activities in the two countries constraining each other.
5
Genworth, summary of post-issues statement hearing, paragraph 13. Summaries of post-issues statement hearings—Genworth, paragraph 15; HSBC, paragraph 2. OFT, ’Payment Protection Insurance, The OFT’s reasons for making a market investigation reference to the Competition Commission’, February 2007. 6
7
170
7.21 We conclude that the relevant geographic market for underwriting is at least as wide
as the UK, and may also include the Republic of Ireland.
Conclusion on the market definition for underwriting of PPI
7.22 We conclude that the market definition for the underwriting of PPI is at least as wide
as the underwriting of all PPI policies in the UK. It may be wider, encompassing other
general insurance and life assurance policies, and the geographic scope may be
wider to include the Republic of Ireland.
Is there evidence of market power being exerted by underwriters?
7.23 We looked at the structure and operation of the underwriting market in order to see if
there is any evidence of market power being exerted by underwriters. Our analysis is
set out in detail in Appendix 7.1.
Concentration of the market
7.24 We calculated the size of the underwriting market and the share of PPI underwriting
for each of the largest underwriters in 2007. Those calculations are set out in
Appendix 7.1. In 2006, five of the underwriters were vertically integrated (Barclays,
HBOS, HSBC, Lloyds TSB and RBSG); HSBC was partially vertically integrated but
sold its Hamilton businesses to Aviva during the course of the inquiry.
7.25 Based upon responses from underwriters to our market questionnaire, we calculated
that the total size of the PPI underwriting market was approximately £3.7 billion in
2007. The largest underwriters in 2007 were Aviva and Lloyds TSB, with market
shares of [] per cent.
7.26 We looked at the degree of concentration in the underwriting market in 2007. We
considered two measures of concentration, referred to in our guidance: 8 the
concentration ratio, and the Herfindahl-Hirschman Index (HHI). The concentration
ratio of the four largest firms in the market is [60–80] per cent. A concentration ratio
in excess of 60 per cent can be considered to raise concern over competition that, if
unchecked, may be to the detriment of the market. The HHI for the entire market is
approximately 1,500, also indicating a concentrated market. 9 The evidence therefore
suggested to us that the market is relatively concentrated.
Capacity
7.27 It was suggested that some underwriters may not be large enough to be considered
credible bidders for the business of the largest distributors, which might suggest that,
for large distributors, the market is more concentrated than indicated in paragraph
7.26. For example, a large distributor ([]) told us that there were seven or eight
possible companies that it would consider to underwrite its PPI business. 10 This view
was supported to an extent by the evidence we saw from the tenders of large
distributors. Bids for underwriting contracts with large distributors appear to be
predominantly from larger underwriters (see Table 7.1). However, we saw nothing
8
CC3, paragraph 3.9. In its merger guidelines, the OFT states that it is likely to regard any market with an HHI in excess of 1,000 as concentrated,
and any market with an HHI in excess of 1,800 as highly concentrated (OFT 516 Mergers Substantive Assessment Guidance, May 2003). 10
The reference to seven or eight firms in the market is likely to cover almost all of the large insurers, those with annual GWP in excess of £100 million. 9
171
which suggested to us that underwriters would be unable to expand capacity
relatively quickly and easily.
Entry, exit and expansion
7.28 We considered whether there are any significant barriers to entry into, or expansion
within, the underwriting market. The reputation of the underwriter is important, so
entry from a related insurance field may be easier than entry into the underwriting
market without any prior underwriting experience.
7.29 We looked at whether there are significant switching costs which inhibit switching by
PPI providers, and therefore expansion by underwriters (see Appendix 7.1). We
received evidence that the costs of changing underwriter could be significant, largely
in integrating IT. Further, using evidence from a large underwriter, we found that
there appears to be a pronounced incumbency advantage—underwriters tend to win
the majority of tenders they bid for if they are the incumbent provider, and lose the
majority of tenders they bid for where they are trying to dislodge an incumbent
provider. However, it was clear to us that distributors can and do switch underwriter
as a result of tenders.
7.30 We looked at the duration of contracts, to see if the length of contracts might inhibit
switching. We looked at the length of existing contracts between underwriters and
distributors with a total annual GWP above £5 million.
7.31 From a total of 67 contracts, 54 were found to be rolling contracts with an initial fixed
period and the remainder (13), fixed-duration contracts. The fixed-duration contracts
were found to have an average duration of five years. The average length of the
initial period for rolling contracts was 3.7 years and of the 37 rolling contracts 32
(86 per cent) are running beyond their initial period and on average they have been
running for five years. The notice periods on the contracts tend to be six months long.
7.32 We concluded that neither the length of contracts nor notice periods were
excessively long. We found that the relatively high proportion of contracts that roll
past their indicated termination date might be further evidence of the role of
reputation in the underwriting market, but may also simply reflect the opportunity cost
of issuing a tender to market with some companies reporting the whole process to
implementation being up to a year in duration.
7.33 We concluded that there were not significant barriers to entry or expansion into the
market for the underwriting of PPI.
Countervailing buyer power
7.34 We looked for any evidence of countervailing buyer power being exercised by the
customers of underwriters.
7.35 Evidence provided by underwriters and distributors indicated that distributors have
choices as to which underwriters to use, and underwriters are prepared to compete
hard to win business, as having access to consumers at the point of sale is an
underwriter’s only way of gaining access to significant numbers of customers. It is
typically the distributor who designs the policy, rather than the underwriter who offers
a ready-made policy during a tender (see paragraph 7.5).
7.36 We looked at how the premium paid by consumers is split between underwriters and
distributors (see paragraphs 4.57 and 4.58). Our analysis suggested that distributors
172
receive a relatively large proportion of the premium in comparison with underwriters
and nearly all of any profit sharing after claims have been paid. Typical commission
rates are 50 to 80 per cent on PLPPI and CCPPI and 40 to 65 per cent on MPPI.
Further, distributors typically take 90 per cent of any profit share and we have seen
instances where the distributor receives 100 per cent of any profit share after payout
of claims.
7.37 We concluded that there is a significant degree of countervailing buyer power being
exercised by distributors.
Profitability of underwriting
7.38 We looked at the contractual arrangements between underwriters and distributors,
the profitability of underwriting PPI, the risk borne by underwriters, and the capital
requirements of underwriters, in order to see how large the returns made by
underwriters were, and to see whether these returns might indicate anything about
the extent of competition between underwriters. Further details of our analysis can be
found in Appendix 7.1.
Documentary evidence from the parties
7.39 The documentary evidence on the profitability of underwriting we saw suggested that
reasonable rates of return have been earned on PPI business. The underwriter is
typically entitled to a fixed percentage of GWP to cover expenses and will also
typically retain some investment income. A small proportion of profit share may also
be earned in the event that claims are less than expected, and the underwriter may,
in certain circumstances, benefit from the tax treatment applicable to any life
business written. Documentation, such as business plans, deal summaries and
strategy documents, from a number of parties indicated typical rates of return of
around 10 to 30 per cent. 11
Financial performance of underwriting
7.40 We asked underwriters for estimates of their return on capital in recent years, and
noted that achieved rates specified by underwriters were generally between 10 and
20 per cent.
7.41 We also looked at the second version of the Market Economics Model (see Appendix
4.4 and Appendix 7.1) which analysed the profitability of underwriting. We did not,
however, make the model our own by verifying the assumptions and collecting data
from the main parties, or revise the model in any way. The underwriting part of the
creditor insurance industry estimated RoEs of 29 and 38 per cent in 2004 and 2008
(forecast).
Conclusions on the profitability of PPI underwriting
7.42 In summary, we looked at underwriting profitability, using a range of board papers,
strategy and financial documents provided by the parties; and a model of market
economics provided by one significant provider. This evidence suggested that:
11
[] presentation showed an after-tax ROCE of [] per cent; [] strategy document showed a post-tax ROCE of [] per
cent; [] strategy document showed historic average RoE of [] per cent, quoted [] per cent as being reasonable
considering the risk profile of the business; and [] internal figures showed an IRR of [] per cent for 2006.
173
• a large share of GWP and profit goes to distributors;
• the insurance risk is borne by the underwriter, most notably because it would
suffer any losses resulting from claims exceeding expectations;
• regulatory capital requirements reflect this risk; and
• achieved returns on capital were generally in the range 10 to 20 per cent.
7.43 On balance, although one piece of evidence (the Market Economics Model)
suggested that returns may have been quite high, we concluded that, based on the
bulk of the evidence that we had seen, underwriters had not earned unreasonable
returns on PPI.
Conclusions on the underwriting market
7.44 We concluded that the underwriting market is relatively concentrated, and that some
distributors think that not all underwriters are sufficiently large to be able to
underwrite their business. There do not appear to be significant barriers to entry or
expansion, though switching costs may be significant and incumbent providers of
underwriting services do appear to have an advantage over other underwriters when
retendering for a contract.
7.45 However, the tender process leads to underwriters competing vigorously with each
other in order to win contracts that will give them access to a significant number of
consumers at the point of sale of a credit product. The evidence relating to
profitability that we have seen indicated that underwriters appear to be making
reasonable rates of return on PPI business. This suggested to us that underwriters
were unable to exert a significant degree of market power.
7.46 We found little to suggest that there are features which prevent, restrict or distort
competition in the underwriting market to any significant extent, and that there is a
significant degree of countervailing buyer power which would prevent market power
being exerted.
7.47 We conclude, therefore, that there is a significant level of competition between
underwriters of PPI, and as a result there are not features of the market which
prevent, restrict or distort competition in the underwriting of PPI.
174
8.
V
ertical integration
8.1 Having considered the state of the markets at the underwriting and retail levels, we
completed our analysis of the state of competition between PPI providers by
considering the effects of vertical integration of the PPI underwriting and distribution
functions.
8.2 We considered the issue of foreclosure of downstream customers by vertical
integration to be primarily a matter related to the extent of downstream market power.
8.3 Appendix 8.1 explains the rationale behind a company being vertically integrated,
and considers in detail the possible issues arising from vertical integration in the
underwriting and distribution of PPI. We summarize here the issues and our analysis
of them.
Vertically-integrated businesses
8.4 As noted in paragraph 2.39, there are four vertically-integrated firms (HSBC was
partially vertically integrated but sold its Hamilton businesses to Aviva during the
course of the inquiry). Table 8.1 shows the percentage of in-house provision of
underwriting by product type for the three main types of PPI policy.
TABLE 8.1 Percentage in-house provision of underwriting, as at June 2007
per cent
Distributor
PLPPI
MPPI
CCPPI
Abbey
A+L
Barclays
Capital One
Cattles
HBOS
HSBC†
LTSB
MBNA
Nationwide
Northern Rock
RBS
0
0
100
0
0
99
[<50]
[<50]
0
0
0
99
0
0
100
0
0
99
[<50]
100
0
0
0
99
0
0
100*
0
0
99
[<50]
[<50]
0
0
0
99
Source: CC, based on information provided by the parties.
*This does not include the Goldfish business, acquired by Barclays in March 2008. PPI distributed by Goldfish is underwritten
by AXA.
†0 per cent since 1 November 2007.
Ways in which vertical integration might harm consumers
8.5 We focused on four ways in which vertical integration might restrict competition:
• by reducing the number of bidders for non-integrated underwriting business;
• by limiting the size of the underwriting market that non-integrated underwriters can
bid for, and so preventing them from achieving an efficient scale of operation;
• by removing the competitive pressure on in-house underwriters; and
• by creating a greater incentive for in-house underwriters to reject PPI claims.
175
Reducing the number of bidders for non-integrated underwriting business
8.6 We considered whether vertically-integrated underwriters could be less likely than
independents to bid for contracts with lenders, reducing the number of bidders for
tenders compared with a situation where all underwriters were independent of PPI
providers. If this were the case, the remaining bidders might have lower incentives to
make competitive bids, and prices might be higher than would otherwise be the case.
8.7 However, we found that most vertically-integrated PPI underwriters do bid to supply
insurance services for third parties. Further, we found that in their dealings with thirdparty PPI providers they appeared not to have any significant advantages over, or
disadvantages compared with, independent underwriters.
Limiting the size of the underwriting market
8.8 It was put to us that the presence of several vertically-integrated distributors has had
the effect of foreclosing a significant portion of the PPI market to non-verticallyintegrated underwriters. 12 We considered whether this reduction in the size of the
market made it harder for independent underwriters to achieve economies of scale.
8.9 Evidence on bids led us to the view that either scale in PPI underwriting is relatively
unimportant, or that scale economies are not PPI-specific and that firms active in
other areas of insurance can achieve scale or scope economies that allow them to
compete successfully for PPI underwriting contracts.
8.10 The evidence suggested to us that, even if non-integrated underwriters were unable
to compete for customers of vertically-integrated businesses, businesses with a
relatively small underwriting presence in relation to PPI are able to achieve sufficient
scale to compete effectively with those with significantly more contracts.
Removing the competitive pressure on in-house underwriters
8.11 It was suggested to us that vertically-integrated underwriters faced lower competitive
pressure than independent underwriters, whose business was put out to tender on a
regular basis. The downside is that a lack of competition may lead to verticallyintegrated underwriters charging higher prices, which would be passed on to
consumers, or having less incentive to innovate.
8.12 We looked first at a simple comparison of the price of PLPPI charged by verticallyintegrated and non-vertically-integrated PPI providers. This showed that nonintegrated firms charged on average £13.76 per month and integrated firms charged
£19.64 per month—a statistically significant difference.
8.13 However, this analysis did not control for policy quality. We therefore looked at the
cost of PPI as a proportion of personal loan monthly cost as a function of quality,
using Defaqto score as a measure of quality. This showed no clearly identifiable
difference between the price or quality of policies offered by vertically-integrated firms
and those offered by non-vertically-integrated firms.
8.14 We also looked at the monthly cost of an insured personal loan as a function of
product quality, as the CC GfK 2008 survey showed that most consumers who
compare two or more PPI policies do so by comparing combinations of PPI and
12
Cardif Pinnacle, response to issues statement, p2; Cardif Pinnacle, summary of hearing, paragraph 21.
176
credit. This indicated that prices for insured personal loans from vertically-integrated
PPI distributors are not higher than from non-vertically-integrated PPI providers (and
may in fact on average be lower).
8.15 We considered whether there was a difference in consumer choice offered by the two
different types of firm. We did this by looking at which firms offered one PPI policy
associated with a credit product and which offered multiple PPI policies (such as
unbundled products—see Appendix 8.1). We saw no clear difference; the evidence
suggested that three of the four vertically-integrated underwriters offered multiple PPI
policies with at least some of their brands.
8.16 In terms of the incentive to innovate, we found that vertically-integrated underwriters
have a strong incentive to innovate in order to maximize the take-up of their products,
and we saw no evidence that this incentive was weaker than for independent
underwriters. We noted that innovation in the PPI industry was rare. We received
evidence of some innovation, in terms of age banding of PPI price, carer cover,
services to help customers find new jobs when made unemployed, and tiered
products. We were told that all these innovations were introduced by independent
underwriters, and that, to the extent that they were available from vertically-integrated
underwriters, they had been copied from independent underwriters. However, we
noted that three recent new policy launches for selling PPI separately from credit
(Barclays’ Plan B for multiple credit cards, Churchill’s short-term IP policy and
HBOS’s Total Mortgage Protection Plan) were all introduced by vertically-integrated
firms. We could not conclude therefore that the level of innovation among verticallyintegrated firms is lower than among non-vertically-integrated firms, and certainly not
to an extent that customers of policies underwritten by vertically-integrated firms were
disadvantaged.
8.17 Overall, we found that there was insufficient evidence to conclude that verticallyintegrated underwriters are insulated from competitive pressure.
Creating a greater incentive for (in-house) underwriters to reject PPI claims
8.18 It has been put to us that, where distributors are vertically integrated and therefore
part of the same group as the underwriter of their PPI policies, the underwriter has an
incentive to maximize profits from the claims reserve. The rationale for this is that :
• If more claims are rejected by the underwriter, the amount of money available for
profit sharing between underwriter and distributor is increased.
• As the distributor is usually paid most of the profit share (typically 90 per cent), a
non-vertically-integrated underwriter stands to gain little from maximizing the
money available for profit sharing.
• If the underwriter and distributor are in the same group of companies, the
underwriter has an incentive to maximize money for profit sharing, through
rejecting more insurance claims, because the increased profit will go to a business
in the same group of companies.
8.19 No evidence to support this view was provided by the companies which put forward
this theory. This is not surprising; we understand that companies do not have access
to data on their competitors’ claims payouts.
8.20 We had doubts that vertically-integrated underwriters had a strong incentive to
maximize claims rejections. It was put to us that the primary reason for choosing a
lender for credit (and hence PPI) is an existing relationship with the lender, and we
177
found it reasonable to assume that this would be the case as much, if not more, for
the clearing banks as for other PPI providers. Rejecting claims would be likely to
reduce the satisfaction of customers whose claims were rejected, and increase the
risk of those customers switching to another lender. The loss of such customers’
business may outweigh the gain from rejecting individual claims.
8.21 We considered whether there may be different incentives to check claims in detail
between integrated and non-integrated underwriters. However, we provisionally
concluded that non-integrated PPI providers would seek to ensure that their
underwriters would act in the way they wanted them to through terms written into
contracts. This would lead to integrated and non-integrated underwriters having
similar incentives for checking claims.
8.22 We do not find, therefore, that vertically-integrated businesses have clear incentives
to reject more claims than non-integrated companies. Concerns over maintaining
their reputation, and the risk of losing customers who source multiple services from
them, would reduce any incentive they may have to maximize claims rejections.
Further, in our view, vertically-integrated companies are uniquely placed to influence
the underwriters’ claims policies, as non-integrated PPI providers appear able to
influence their underwriters’ policy through contractual arrangements.
Conclusions on vertical integration
8.23 Our analysis on vertical integration of PPI underwriters and distributors did not
identify issues which would prevent, restrict or distort competition in the underwriting
market to the detriment of consumers. We did not identify any issues with respect to
the distribution of PPI policies which are unique to vertically-integrated distributors;
any issues that may exist appear to be linked to market conditions which nonintegrated PPI providers experience equally.
178
9.
F
indings
The distribution of PPI
9.1 We are charged with determining whether any feature or combination of features of
any relevant market are present which are preventing, restricting or distorting competition in the supply of PPI (except store card PPI) to non-business customers in the
UK. A feature for the purposes of section 134(1) of the Act can take the form of the
structure of the market and/or conduct of any of the participants in the market,
including consumers (see paragraph 1.3). ‘Conduct’ includes any failure to act
(whether or not intentional) and any other unintentional conduct. We consider if any
feature of relevant markets prevents, restricts or distorts competition in the market for
the provision of PPI to non-business customers in the UK such that there is an
adverse effect on competition.
9.2 We conclude that there are features of relevant markets which, alone or in combination, prevent, restrict or distort competition in the supply of PPI to non-business
customers in the UK:
(a) Distributors and intermediaries fail actively to seek to win customers from their
rivals by using the price or quality of their PPI policies as a competitive variable.
(b) Consumers who want to compare PPI policies (including PPI combined with
credit), stand-alone PPI or short-term IP insurance policies are hindered in doing
so. Product complexity (the variations in pricing structures (in particular in relation
to single-premium policies) and in terms and conditions, the way information on
PPI is presented to customers); the perception that taking PPI would increase
their chances of being given credit; the bundling of PPI with credit; and the limited
scale of stand-alone provision act as barriers to search for all types of PPI
policies. The bundling of retail PPI with credit accounts and with merchandise
cover (also known as purchase protection insurance) acts as a barrier to search
for retail PPI. In addition, the time taken to obtain accurate price information is a
barrier in relation to the provision of PLPPI, MPPI and SMPPI. These barriers to
search impede the ability of consumers to make comparisons, and therefore
effective choices, between PPI policies. They also, therefore, act as barriers to
expansion for other PPI providers, in particular providers of stand-alone PPI.
(c) Consumers who want to switch PPI policies to alternative providers or to
alternative types of insurance policies are hindered in doing so. Terms which
make switching expensive (in the case of single-premium policies) as barriers to
switching for PLPPI and SMPPI policies. Terms which risk leaving consumers
uninsured (for a short period of time or in case they suffer a recurrence of a
condition) act as barriers to switching for all types of PPI policies. In addition, the
lack of access to consumers’ balance information acts as a barrier for switching
for CCPPI and retail PPI and the bundling of retail PPI with merchandise cover
acts as a barrier to switching for retail PPI. These barriers to switching limit
consumer choice. They also therefore act as barriers to expansion for other PPI
providers, in particular providers of stand-alone PPI.
(d) The sale of PPI at the point of sale by credit providers and, in relation to retail
PPI, continued exclusive access to customer accounts further restricts the extent
to which other providers can compete effectively.
9.3 We have found, pursuant to section 134(1) of the Act, that there are features of
relevant markets, which alone or in combination, prevent, restrict or distort
competition in the supply of PPI to non-business customers in the UK and
179
accordingly there is an AEC within the meaning of section 134(2) of the Act. The
features are those that we have identified in Section 5 and Section 6 of this report
and which are summarized in paragraph 9.2.
9.4 We concluded that the detrimental effects on consumers of these features in relation
to all types of PPI (except retail PPI) were higher prices for, and less choice in, PPI
policies than would be expected in a well-functioning market. We also concluded that
demand for PPI was distorted. We also concluded that it was possible that there was
less innovation than would be expected in a well-functioning market. For retail PPI
we concluded that the detrimental effects on consumers of these features were
higher prices for retail PPI policies than would be expected in a well-functioning
market. We also concluded that it was possible that there was less innovation than
would be expected in a well-functioning market.
The underwriting of PPI
9.5 We concluded that the underwriting market is relatively concentrated, and that some
PPI providers think that not all underwriters are sufficiently large to be able to
underwrite their business. There do not appear to be significant barriers to entry or
expansion, though switching costs may be significant and incumbent providers of
underwriting services do appear to have an advantage over other underwriters when
retendering for a contract.
9.6 However, the tender process leads to underwriters competing vigorously with each
other in order to win contracts that will give them access to a significant number of
consumers at the point of sale of a credit product. The evidence relating to profitability that we have seen indicated that underwriters appear to be making reasonable
rates of return on PPI business. This suggested to us that underwriters were unable
to exert a significant degree of market power.
9.7 We found little to suggest that there are features which prevent, restrict or distort
competition in the underwriting market to any significant extent. We also found that
there is a significant degree of countervailing buyer power which would prevent
market power being exerted.
9.8 We conclude, therefore, that there is a significant level of competition between
underwriters of PPI, and as a result there are not features of the market which
prevent, restrict or distort competition.
Vertical integration
9.9 Our analysis on vertical integration of PPI underwriters and distributors did not
identify issues which would prevent, restrict or distort competition in the underwriting
market to the detriment of consumers. We did not identify any issues with respect to
the distribution of PPI policies which are unique to vertically-integrated PPI providers.
180
10.
R
emedies
Introduction
10.1 We now consider measures to remedy, mitigate or prevent the AEC, or resulting
detrimental effects on customers, as set out in our findings and conclusions summarized in Section 9 of this report. This section describes our consideration of evidence
relating to each individual element of the remedies package and to remedies in
general, and sets out our decisions on remedies.
10.2 On 5 June 2008 we issued a Notice of Possible Remedies (the Remedies Notice)
and on 10 October we issued a Supplementary Notice of Possible Remedies for
retail PPI (the Supplementary Remedies Notice) which set out various options that
we considered might be effective, alone or in combination, in addressing the AEC
that we provisionally found, or in addressing the customer detriment resulting from
the provisional AEC. We received 51 written responses to the Remedies Notice and
8 responses to the Supplementary Remedies Notice; non-confidential versions of
these can be found on our website. In addition, we held 25 remedies hearings with
main parties to the inquiry, consumer bodies, the OFT and the FSA. We also
received evidence from credit reference agencies (CRAs) and others (see Appendix
2.8, which includes parties that responded to the Remedies Notice). In addition, we
commissioned research into consumer views on price disclosure and asked the
larger parties a number of additional questions, including questions about claims
profiles over the life of their single-premium policies. On 14 October, we consulted on
some further analysis conducted by the CC in support of our provisional decision on
remedies, to which we received 23 responses. In the light of this evidence we published our provisional decision on remedies on 13 November 2008, our provisional
decision on retail PPI remedies on 21 November 2008, and a paper describing the
application to intermediaries of the provisional decision on remedies on 12 December
2008. Our remedies assessment takes account of the responses to the Remedies
Notice, the Supplementary Remedies Notice and of the responses to our provisional
decisions on remedies, and the other evidence on remedies that we received during
our investigation.
10.3 The remainder of this section is structured as follows:
(a) First, we set out our framework for the assessment of remedies (paragraphs 10.4
to 10.10).
(b) Next, we consider general issues regarding PPI markets which, it has been put to
us, we should consider when deciding what remedies to impose (paragraphs
10.11 to 10.30).
(c) We then look at the individual elements that we have decided should form part of
the overall remedies package. For each of these we set out the relevant element
of the remedies package, a summary of the responses received during consultation and our views on those responses (paragraphs 10.31 to 10.339).
(d) Next, we set out those remedies we decided not to implement, summarize the
responses we received during consultation and set out why they are not being
implemented (paragraphs 10.340 to 10.373).
(e) We then consider where there are any relevant customer benefits arising from
the adverse features that we have identified (paragraphs 10.374 to 10.463 ).
181
(f) We assess the overall effectiveness and proportionality of the remedies package,
including whether to modify the remedies package that we would otherwise
impose and recommend in order to retain any relevant customer benefits arising
from the features that we have identified (paragraphs 10.465 to 10.514).
(g) We consider issues relating to the implementation of the remedies package
(paragraphs 10.515 to 10.566).
(h) Finally, we summarize our decisions on remedies (paragraphs 10.567 to 10.572).
Framework for the assessment of remedies and relevant customer benefits
10.4 Under section 134(4) of the Act, if the CC has decided that there is an AEC, it must
decide the following additional questions:
(a) whether action should be taken by the CC for the purpose of remedying, mitigating or preventing the AEC concerned or any detrimental effect on customers so
far as it has resulted from, or may be expected to result from, the AEC;
(b) whether we should recommend the taking of action by others for the purpose
outlined in (a) above; and
(c) in either case, if action should be taken, what action should be taken and what is
to be remedied, mitigated or prevented.
10.5 In choosing appropriate remedial action, the CC has a statutory obligation to have
regard to the need to achieve as comprehensive a solution as is reasonable and
practicable to the AEC and any detrimental effect on customers so far as resulting
from the AEC. 1
10.6 As noted in our guidance (CC3, paragraph 4.9), we must consider the effectiveness
of different remedies and their associated costs and will have regard to the principle
of proportionality when deciding on appropriate remedies. Our guidelines outline
several factors relevant to our consideration of effectiveness (CC3, paragraph 4.13
and following). First, an effective remedy will make clear the persons to whom it is
directed and any other persons who might be interested in it. Second, we should
consider the prospects of a particular remedy being implemented and complied with.
A third relevant consideration is the time period within which the remedy will be
effective. Other factors may also be relevant to our consideration of effectiveness,
depending on the facts of the case.
10.7 In considering whether a remedy is reasonable and practicable, we should consider
its implementation costs (CC3, paragraph 4.10). We should endeavour to minimize
any ongoing compliance costs to the parties, provided that the effectiveness of the
remedy is not reduced (CC3, paragraph 4.12). However, we should balance those
costs against the benefit to the UK economy and to consumers in particular.
10.8 We should also take account of the proportionality of any remedies or package of
remedies in relation to the AEC and any resulting detrimental effect on consumers. If
we are choosing between two remedies or packages of remedies which we consider
would be equally effective, we will choose that which imposes the least cost or that is
the least restrictive (CC3, paragraph 4.10).
1
Section 134(6) and section 138(2) and (3) of the Act.
182
10.9 In deciding on appropriate remedies, we may also have regard to the effects of any
remedial action on any relevant customer benefits arising from the adverse feature or
features of the market concerned. 2 Such benefits comprise lower prices, higher
quality or greater choice of goods or services in any market in the UK (whether or not
the market to which the feature(s) concerned relate) or greater innovation in relation
to such goods and services. To qualify as a relevant customer benefit within the
meaning of section 134(8), we must believe that the benefit has accrued as a result
(whether wholly or partly) of the feature(s) concerned, or may be expected to so
accrue within a reasonable period and that the benefit would be unlikely to accrue
without the relevant feature or features.
10.10 In general, we should seek to implement (or recommend) remedies that address the
AEC, though we may also choose to address the detrimental effect on consumers
(CC3, paragraph 4.6). 3 In our guidelines, we say that it is unlikely that, having
decided that there is an AEC, we will decide that there is no case for remedial action,
at least before we have given attention to any relevant customer benefits that may
accrue from the market features.
General issues
10.11 A number of general issues have been put to us in the course of the remedies
process. These related to:
(a) the likely impact of Insurance Conduct of Business sourcebook (ICOBS) 4
regulation on competition;
(b) the impact of the economic downturn;
(c) the impact of any remedies on the ‘protection gap’ and default rates on credit;
(d) the impact of any remedies on adverse selection and the viability of supplying
PPI;
(e) comparison with remedies put in place following other market investigations; and
(f) recent competitive developments in PPI markets.
Impact of ICOBS regulation
10.12 Many parties told us that the revised ICOBS regulation would make a significant
impact on the sales process of PPI and as a result would change the market. It was
suggested to us by some parties that we should allow the changes introduced by
ICOBS to ‘bed down’ before we considered whether remedies were required. Others
suggested that the changes to how insurance is sold introduced by ICOBS were in
themselves sufficient to remedy some or all of the adverse effects we found. We
noted that we have to publish our final report on this investigation by February 2009,
and are therefore unable to wait and see what changes in the market result from
ICOBS. However, the FSA told us that it was fairly convinced that the kind of information and disclosure remedies that it typically deployed were unlikely significantly to
affect the structural problems seen in this market.
2
Section 134(7) of the Act. The CC has said (CC3, paragraph 4.22) that it ‘will first look for a remedy that would be effective in dealing with the adverse effects on competition of the market features rather than seeking to deal with any detrimental effect on consumers’. 4
See Appendix 2.5, paragraphs 27–35. 3
183
10.13 We do not believe that the measures introduced by the FSA, such as increased
cooling-off periods and a greater provision of non-price information—which are
intended to increase consumer protection rather than to address competition
problems in the market—would be nearly sufficient by themselves to remedy the
AEC that we have identified in PPI markets. For our analysis regarding cooling-off
periods and the provision of non-price information, see paragraphs 10.68, 10.69 and
10.186 to 10.192.
The impact of the economic downturn
10.14 Several parties told us that the prevailing economic conditions had led to a significant
increase in claims on PPI policies, particularly for unemployment. The ABI provided
evidence that the number of new claims because of unemployment was 69 per cent
higher in September 2008 compared with September 2007. [], [] and [] also
told us that unemployment claims had increased. A large distributor ([]) provided
us with evidence that the number of involuntary unemployment claims was
significantly higher in September 2008 than in September 2007, and similarly higher
in October 2008 than in October 2007, and that involuntary unemployment claims
represented a significantly greater proportion of all claims. [] evidence also showed
that, whilst involuntary unemployment claims had increased, the overall number of
claims in September 2008 had decreased by 11 per cent on the levels of September
2007. The overall number of claims in October 2008 was significantly higher than the
number made in October 2007. Aviva told us that its analysis of the impact of the
recession of the early 1990s showed that the average annual cost to a scheme for
unemployment per person covered increased by [300–500] per cent over three years
(from 1989 to 1992). For sickness and disability, the corresponding increase was [0–
200] per cent. This was as a result of increases in both the frequency and the
duration of claims. Further, as set out in paragraph 10.424, we were told that some
changes to policies had occurred recently. More generally, we were told that some
distributors had stopped selling PPI policies (though we were not told that these
decisions were related to the current economic downturn). Sterling Insurance told us
that some providers had stopped selling new PPI policies since September 2006,
citing, among others, Virgin and M&S Money.
10.15 It was put to us, in this context, that we could place little weight either on the results
of our analysis of claims profiles for single-premium policies (see paragraphs 10.409
to 10.417) or our analysis of waterbed effects (see paragraphs 10.443 to 10.457 and
10.480 to 10.492), primarily because the likely effect of the economic downturn would
be to raise claims costs, in particular those associated with unemployment.
10.16 The current downturn has not yet run its course, and neither we nor the parties are
able to predict reliably what will happen. The previous economic downturn started
nearly 20 years ago. Even if the PPI and credit data needed to conduct the analysis
of waterbed effects were available for that period, because it was so long ago, we do
not think we could place much weight on the results as an indicator of what is likely to
happen in the current downturn.
10.17 We were not persuaded that the current economic downturn would clearly result in a
sufficiently different outcome from our analyses on claims profiles for single-premium
policies and waterbed effects to warrant a change in our approach to remedies. In
paragraph 10.410, we note that claims costs were relatively small compared with the
price that consumers pay for PPI. Whilst claims costs may increase in an economic
downturn, many other factors affecting the size of claims costs relative to the price
that consumers pay may also change—for example, the prices that consumers pay
for PPI, or the level of demand for PPI.
184
10.18 Our analysis of potential waterbed effects was based in part on assumptions regarding the profitability of PPI, the intensity of competition for credit consumers and the
responsiveness of credit demand (at a market level) to changes in average credit
prices.
10.19 As noted in paragraph 10.17, whilst claims costs may be affected by an economic
downturn, many other factors affecting PPI prices and the profitability of PPI distribution may also change. Therefore it is not clear to us that PPI profitability would
reduce significantly in an economic downturn given the current market structure. We
note in paragraphs 10.458 to 10.463 that there is some uncertainty regarding the
intensity of competition for credit consumers going forward. Consequently we are not
confident that the scale of the relevant customer benefit of lower credit prices that we
have observed in the period up to December 2006 will persist at this level in the
future. One third party ([]) told us that the sensitivity of credit customers to changes
in average credit prices could increase substantially as the economy entered
recession and customers tightened their expenditures and their budgets. However,
given that there are likely to be many factors affecting credit demand in an economic
downturn, we felt that we could make no clear prediction regarding its effect on the
responsiveness of credit demand to changes in average credit prices.
10.20 Finally, we were told that we should take a cautious approach to remedies in light of
the economic outlook. As set out in paragraph 10.17, we see no clear reasons why
an economic downturn would necessarily diminish the adverse effects on customers
of the monopoly that distributors enjoy over sales of PPI to their own credit
customers. As set out in paragraph 10.463, there is some uncertainty over whether
the relevant customer benefits in the form of low credit prices would persist at the
levels that we have observed up to December 2006, which would increase rather
than reduce the case for intervention. If anything, in an economic downturn the case
for intervention to address a competition problem in PPI could be seen to be more
pressing, since the current high prices discourage PPI uptake and could result in
consumers being uncovered at a time of increased risk. We concluded that we
should not change our approach to remedies, in the light of the current economic
outlook.
Impact on the ‘protection gap’
10.21 We were told that any measure that reduced the uptake of PPI would increase the
protection gap, 5 which, given the economic downturn, we were told would be particularly unfortunate. In addition, we were told that a reduction in the proportion of loans
for which customers had taken out PPI could increase default rates to a significant
degree.
10.22 Our analysis of PPI price elasticities indicated that the current high price of PPI
deters some borrowers, who would take it out if it were competitively priced, from
taking out PPI policies. Measures which increase competition are likely to bring down
the price of PPI policies, which, in turn, is likely to make PPI more attractive to consumers, increasing the likelihood of consumers taking out PPI policies. We also
noted that, in times of economic downturn, consumers are more likely to think about
the possible consequences of taking on, or having taken on, debt. We concluded that
our remedies package, which is designed to increase competition between PPI
5
The ABI defined ‘protection gap’ in its report as existing when households did not have sufficient insurance or other coping
strategies in place to match the loss of income resulting from unemployment, illness or loss of life. We concentrate on ‘debt’
and ‘essential’ expenditure, since ‘lifestyle’ expenditure can be reduced. Source: Coping with crises: household protection
needs, ABI Insurance Market Study No 5, 2008.
185
providers and reduce the cost of PPI policies to consumers, would not be detrimental
to customers during an economic downturn.
Increased risk of adverse selection
10.23 We were also told 6 that some of the proposed remedies (in particular, the prohibition
on selling PPI at the credit point of sale) could result in the PPI market disappearing
altogether. For example, Aviva said that any initial fall in sales completed at the point
of sale of credit could lead to increased adverse selection, which in turn would lead to
reduced customer choice, higher prices and a further fall in sales. This in turn would
lead to PPI providers and underwriters exiting the market, and that in turn would lead
to further reduced consumer choice, higher prices and a further fall in sales. We were
told that it was important for underwriters to get a suitable balance of high- and lowrisk customers in sufficient volumes in order to sustain a commercially viable PPI
product that was available to a broad range of consumers with minimal exclusions. 7
10.24 However, there are likely to be a number of conflicting factors affecting the balance of
low-risk and high-risk customers. In particular, we would expect our remedies to have
a significant impact on PPI prices. It is a generally accepted economic principle that
adverse selection problems are exacerbated by high prices. This is because high
prices discourage low-risk consumers more than they discourage high-risk consumers. Given the scale of distributor margins, we would expect our remedies to
deliver a large reduction in PPI prices and therefore a significant improvement in the
average risk profile of PPI customers.
10.25 In addition, we consider that the package will decrease adverse selection risks in the
stand-alone market. A requirement not to sell PPI at the same time as credit would
lead, in our view, to more consumers searching for PPI, increasing the number of
customers considering stand-alone PPI policies. The increase in customers looking
at stand-alone policies, and the resulting change in the balance between high- and
low-risk customers, when taking out credit would, in our view, reduce overall adverse
selection risks for stand-alone providers.
10.26 As regards adverse selection in credit markets, we note that there may be some
increases in credit prices for unsecured and secured personal loans due to waterbed
effects (see paragraphs 10.442 to 10.464). However, we would expect that price
effects in the credit markets will be smaller in magnitude than the effects on PPI
prices. 8 Distributors can also protect themselves from credit risk to an extent using
their credit-scoring systems. In our view, any increase in adverse selection in the
markets for credit resulting from an increase in PPI and credit prices as a result of
our remedies is likely to be modest.
10.27 As regards distributors, it is possible that adverse selection for some might increase,
if their penetration rates were to fall significantly. We consider this unlikely in the
context of falling prices for PPI. However, we note that current claims ratios would
have to rise, in some cases by more than six times, to reach the levels of claims
ratios seen in other insurance markets (see Table 4.3). Moreover, we were told that
adverse selection and moral hazard risks in the PPI market were different from those
6
HSBC response to Remedies Notice, p12.
Aviva response to provisional findings, paragraph 2.3.
8
This is because distributors offer the same credit price to a credit customer regardless of whether that customer takes out PPI
or opts for credit only. Where there is a waterbed effect, the expected profit on PPI sales is therefore competed away by
offering lower prices to both credit-only and credit and PPI customers. By way of example, if PPI sales result in £1 of ‘excess
profit’ per policy and PPI is purchased on half of all loans, then under a full waterbed the loan price to all customers will be
discounted by 50p per loan.
7
186
in the motor or household insurance market, 9 as the risks were less affected by
customer behaviour and a customer’s risk in terms of their unemployment and health
tended to be of public record, allowing firms to underwrite that customer based on
known medical conditions or exclusions. We consider that these differences make
the potential for adverse selection to be a smaller problem for distributors, and easier
for them to mitigate by using pre-existing condition clauses in the contracts, if this
were necessary. But it was not clear that the risk profile of customers choosing to buy
from a distributor or an intermediary offering PPI at the credit point of sale would
change dramatically to the disadvantage of that provider if there were a point-of-sale
prohibition. Our judgement is that, if adverse selection did increase for some distributors, it would not do so to a level which would lead to a significant reduction in the
number of providers of PPI.
10.28 We therefore disagree with the parties that our package of remedies will lead to a
significant reduction in the number of different PPI policies being offered. We expect
that stand-alone sales should increase significantly and the number of providers of
stand-alone PPI should increase. Sales by individual distributors may decline, but we
would expect lower PPI prices to encourage PPI take-up, and improve the balance of
low-risk and high-risk customers. We therefore do not expect that the balance of
high- and low-risk customers buying their products would deteriorate dramatically, or
that there would be a significant reduction in the number of providers of PPI.
Comparison with extended warranties and store cards
10.29 Some parties referred us to the CC’s remedies 10 in Store Cards 11 and Extended
Warranties 12 and said that the CC had not proposed such extensive remedies in
these cases. However, we are not bound by the remedies that have been introduced
in inquiries into other markets; each reference market needs to be assessed individually on its merits and appropriate remedies put in place where adverse effects are
identified. We did, however, consider the implications of the report by LECG regarding the impact of the remedies imposed in the Extended Warranties case, and our
analysis of this is set out in paragraph 10.469.
Recent competitive developments
10.30 Some parties told us that our findings did not fully reflect competitive developments
that had recently taken place in the PPI market. For example, the FLA 13 told us that
80 per cent of its members that had responded to a recent survey had improved
terms and conditions in some way in the last year, and 36 per cent had reduced PPI
prices. However, we have also been made aware of price rises which have occurred
during the course of our inquiry. 14 We noted this evidence but concluded that these
price changes and other developments that have been brought to our attention do
not represent a decisive shift in nature of competition in PPI markets. We also note
that a number of parties over the period of this investigation have either stopped
selling retail PPI or stopped selling retail PPI via certain distribution channels. Further
details on recent market developments in retail PPI can be found in Section 6.
9
‘Moral hazard’ is the concept that people with insurance may take greater risks than they would without insurance, because all or part of that risk is covered. For example, in the context of PPI, policyholders may take less care to avoid accidents, sickness or unemployment. 10
Genworth referred to Extended Warranties at pp7&8 of its response to the Remedies Notice. 11
Store Cards, TSO, March 2006. 12
Extended warranties on domestic electrical goods, TSO, December 2003. 13
FLA response to Remedies Notice, p5. 14
For example, between June 2006 and June 2007 moneyfacts.co.uk reported that some providers, including Alliance & Leicester, Direct Line, Lloyds TSB and Liverpool Victoria had increased PPI prices (it also noted that some had reduced prices). Moneyfacts.co.uk press release, 18 June 2007. 187
Remedies the CC has decided to implement
10.31 This next section sets out the package of remedies that we have decided to implement. We start by explaining each individual element of our remedies package:
(a) a prohibition on selling PPI at the credit point of sale and within a fixed time
period of the credit sale (‘the point-of-sale prohibition’);
(b) an obligation to provide a personal PPI quote (‘the personal PPI quote’);
(c) an obligation to provide information about the cost of PPI and ‘key messages’ in
PPI marketing material (‘information provision in marketing material);
(d) an obligation to provide information to the OFT and the FSA for monitoring and
publication; and an obligation to provide information about claims ratios to any
party on request (‘provision of information to third parties’);
(e) a recommendation to the FSA that it uses the information provided to it under this
obligation to populate its PPI price comparison tables;
(f) an obligation to offer retail PPI separately from merchandise cover where both
are offered together as a bundled product (‘unbundling retail PPI from merchandise cover’);
(g) a prohibition on the selling of single-premium PPI policies (‘single premium
prohibition’); and
(h) an obligation to provide an annual statement of PPI cost and a reminder of the
consumer’s right to cancel (‘annual statement’).
10.32 For each element of the remedies package, we set out our decision, how it contributes to addressing the AEC we have found, parties’ views on it and specific points
relevant to its design.
10.33 We considered each element of the remedies package separately for each type of
PPI policy for which an AEC has been found (ie PLPPI, CCPPI, MPPI, SMPPI and
retail credit PPI) and specified the individual elements of the package in such a way
as to reflect the different characteristics of the different types of PPI and the underlying credit. We considered that it was necessary to include all elements of the
remedies package to achieve a comprehensive solution to the AEC that we found
and resulting customer detriment.
Prohibition on selling PPI at the credit point of sale and within a fixed time
period of the credit sale (the ‘point-of-sale prohibition ’)
Summary of this element of the remedies package
10.34 This element of the remedy package prohibits distributors and intermediaries that
arrange credit for a consumer from selling PPI or contacting the consumer regarding
PPI during the credit sale period and either for seven days after the conclusion of the
credit sale period or for seven days after the provision of a personal PPI quote (if this
occurs after the credit sale). Distributors and intermediaries arranging the credit for
the consumer, as well as certain other associated persons, will be permitted to
market the PPI products that they offer during the credit sale process, but will not be
able to enter into a contract to provide PPI at that stage or for another seven days
after the conclusion of the credit sale period. There is a limited exemption for
188
consumer-initiated transactions made over the Internet or by telephone at least
24 hours after the credit sale (or provision of a personal PPI quote) if they can
confirm that they have received a personalized quote (see paragraphs 10.157 to
10.181).
10.35 Figure 10.1 summarizes the key aspects of this prohibition. Paragraphs 10.157 to
10.181 set out the contents of the personal quote.
FIGURE 10.1
The point-of-sale prohibition
Before the credit sale
PPI advertising is allowed; sales personnel can discuss PPI in general terms with consumers.
To avoid circumvention of this remedy, the distributor or the intermediary arranging the credit cannot ‘pre-sell’ PPI to insure a credit agreement it has discussed with a consumer and has reasonable grounds to expect it will agree with that consumer within the following seven days. During the credit sale period*
The distributor or the intermediary arranging the credit cannot conclude the sale of a PPI policy at any point during this period. The distributor or the intermediary arranging the credit cannot include PPI in the ‘primary credit agreement’ (the credit agreement which relates to the credit that the PPI is primarily sold to insure). The distributor or the intermediary arranging the credit can provide a personal PPI quote to the consumer during this period and is obliged to do so if they provide any information about PPI during the credit sale period. If a distributor or the intermediary arranging the credit does not provide a personal PPI quote (see Figure 10.2) during the credit sale period, but subsequently contacts the consumer to offer PPI, a personal PPI quote must be provided at that time and the prohibition period starts from the date on which the personal PPI quote is provided to the consumer. If the salesperson mentions PPI orally, then they must also orally disclose the key messages (see paragraphs 1 to 3 of Figure 10.3). After the start of the credit sale period
The following businesses are prohibited from selling PPI to the consumer or from contacting the consumer in relation to PPI until the end of the prohibition period: (a) the distributor or the intermediary offering or arranging the credit; (b) any business mentioned during the credit sale period with whom the distributor or the intermediary arranging the credit has a commercial referral arrangement for PPI; or (c) any business to which information regarding the consumer which was obtained in the credit sale period has been passed for the purpose of selling PPI to the customer. The consumer can buy PPI from any other provider at this stage and during the prohibition period. 24 hours after the conclusion of the credit sale period†
The distributor, the intermediary arranging the credit or any business covered by the prohibition during the credit sale period may sell PPI to the consumer over the Internet or telephone provided that the consumer has initiated the transaction. Any business covered by the prohibition, which is selling the policy covered by the personal PPI quote, must obtain confirmation that the consumer has seen the personal PPI quote, before any PPI sale can be made. Any business covered by the prohibition must be able to verify to monitoring agencies that any sales within this exception period were genuine consumer-initiated sales. Seven days or more after the conclusion of the credit sale period†
The distributor (or any of the other businesses covered by the prohibition) can contact the consumer
with regard to PPI. Any party covered by the prohibition who is selling the policy covered by the quote must obtain confirmation that the customer has seen the personal PPI quote, before any PPI sale can be made. This contact can be made in the activation process of a credit card, provided this is seven days after the conclusion of the credit sale period.
*Defined as starting with a consumer’s application for credit and lasting until the consumer is provided with confirmation in a durable medium that the credit provider is bound to provide the credit. See paragraphs 10.106 to
10.108 for further details in relation to different forms of PPI.
†Or after the provision of a personal PPI quote if one were not provided during the credit sale period.
189
How the point-of-sale prohibition contributes to addressing the AEC
10.36 This element of the remedies package will directly address the AEC arising from the
sale of PPI at the point of sale by distributors or intermediaries. It will also address
each of the following barriers to search, that we found contributed to the AEC (see
paragraph 9.2):
(a) the perception that taking PPI would increase consumers’ chances of being given
credit;
(b) the bundling of PPI with credit;
(c) the limited scale of stand-alone provision; and
(d) the time taken to obtain accurate price information. 15
10.37 We have decided that a point-of-sale prohibition is the only option that will effectively
address the point-of-sale advantage, and as such is essential to achieving a comprehensive solution to the AEC and resulting consumer detriment. It complements and
enhances other measures to address barriers to search—in particular, the personal
PPI quote, the provision of information in marketing material, unbundling retail PPI
from merchandise cover and the provision of information to third parties.
10.38 In our assessment of this element of the remedies package, we first considered a
number of potential risks highlighted to us by parties associated with a point-of-sale
prohibition and possible alternatives to a point-of-sale prohibition. We then considered the detailed design of the point-of-sale prohibition.
Risks of a point-of-sale prohibition
10.39 Our proposals for a point-of-sale prohibition were supported by the OFT, the FSA
and the Financial Services Consumer Panel, as well as some stand-alone providers. 16 However, many parties, including all distributors who commented on the
proposals and many of their underwriters, were not in favour of a point-of-sale prohibition. These parties suggested that a point-of-sale prohibition would have the
following risks (which we review in paragraphs 10.40 to 10.66; in doing so we also
consider the possible benefits of the remedy proposal):
(a) aspects of the point-of-sale advantage would not be addressed;
(b) there would ultimately be a reduced take-up of PPI;
(c) a reduction in consumer choice would result;
(d) product quality would decline;
(e) higher distribution costs would be incurred;
(f) differential effects on distributors would be experienced; and
(g) the freedom of establishment for PPI distributors would be restricted.
15
The point-of-sale prohibition will work in conjunction with the provision of the personal PPI quote to reduce the time taken to
obtain accurate price information. Post Office response to Remedies Notice p4. Claim2Gain response to Remedies Notice p2. Paymentcare response to the provisional decision on remedies. 16
190
Inability of the remedy to address all aspects of the point-of-sale advantage
10.40 Some parties said that the prohibition would not comprehensively address the pointof-sale advantage. 17 These parties said that the incumbency advantage enjoyed by a
distributor was not limited to the consumer interaction of the point of sale but was an
issue throughout the term of the credit as the distributor was the only firm that would
be aware of the credit. In addition, one party (Lloyds TSB) told us that distributors
and intermediaries arranging the credit would still have a superior opportunity to
market their PPI to their new credit consumers because they would be aware of the
consumer’s existing level of demand, likely need for PPI and the suitability of the
particular PPI product or products that they offered. We were also told that the pointof-sale prohibition would not address a ‘goodwill’ advantage which would remain
because the consumer has already exhibited a preference for the lender’s credit
product and discussed their financial requirements with them. Citizen’s Advice told us
that it hoped that the prohibition would open sufficient space for stand-alone providers to compete prices down and quality up, but that it was concerned that the
growth of stand-alone policies might be fairly small and the net effect could be to
shrink the market, which would not be wholly beneficial for consumers.
10.41 We agree that this remedy will not entirely remove all aspects of the incumbency
advantage enjoyed by distributors. However, we do not think that we need to remove
all incumbency advantages of distributors in order effectively to remedy this aspect of
the AEC (see paragraphs 10.96 to 10.98). We concluded that this remedy will make
a very substantial contribution to the overall effectiveness of the remedies package
and it would be neither realistic (without a prohibition on distributors ever selling PPI
to their credit customers) nor necessary to seek to remove all aspects of distributors’
incumbency advantages.
10.42 Others argued that we could not be certain that the remedy would be sufficiently
effective in stimulating consumer search or the development of competition on a
stand-alone basis. Some parties (Aviva, RBSG, Egg, Abbey) said that some consumers will fail to take out PPI when faced with a more complex sales process and
some distributors are likely to stop selling PPI policies, in some cases providing
alternative products, instead of competing in a more regulated stand-alone PPI
market. 18 In paragraphs 10.380 to 10.403, we also discuss a number of points that
were cited by parties as relevant customer benefits of point-of-sale PPI products.
10.43 We acknowledge that—as with any intervention aimed at enhancing competition—
there is a risk that this element of the remedies package will not generate the
changes in behaviour necessary fully to address the AEC. However, we found that
the sale of PPI at the credit point of sale was a significant barrier to consumer
search, and that currently only between 11 per cent (SMPPI) and 21 per cent (MPPI)
of consumers compared two or more PPI policies or two or more loans with PPI
policies in making their choice of provider. By creating a clear break between the sale
of credit and the sale of PPI, and providing consumers with the tools that they require
to compare PPI policies (personal PPI quote, information provision in marketing
materials and provision of information to the FSA for use in its comparative tables),
we expect this aspect of the remedies package significantly to increase incentives
for, and ability of, consumers to search for the best-value PPI policy that meets their
needs.
17
ABI response to Remedies Notice p7, Axa response to Remedies Notice p6, Capital One response to Remedies Notice p3.
RBSG also said in its response to the provisional decision on remedies (pp5&6) that consumers might be discouraged from
taking out PPI because it was suggested that the majority of consumers preferred purchasing credit and PPI together and it
would be more complex for the consumer if these purchases were separate.
18
191
10.44 In our judgement, this element of the remedies package is also likely to increase
marketing spend on PPI, as distributors seek to increase the profile of their PPI
products to increase the likelihood of achieving a subsequent sale, and from standalone providers seeking to take advantage of the prohibition. In the current market
structure, few consumers search the market, which reduces the effectiveness of
marketing activity. 19 We consider PPI marketing spend to be low given the size and
profitability of the market. For example, the UK travel insurance market is estimated
to be worth around £709 million 20 in terms of net written premium and has an advertising spend of £5.5 million 21 (not including Internet advertising). We also noted that
the household insurance market grew to £6,392 million in GWP in 2006 and had a
marketing spend of over £100 million 22 and that [] predicted that it would spend
£7.5 million on direct marketing business in 2008. PPI sales generated a GWP of
£4.4 billion in 2006 and only a few examples of distributors actively promoting their
PPI or advertising campaigns specifically featuring PPI policies (see paragraph 4.20).
An increase in marketing is also likely to generate additional consumer search,
intensifying competitive pressure.
10.45 By encouraging consumers to shop around after the credit sale, we considered that a
point-of-sale prohibition will open up the possibility for substantially greater sales to
be made on a stand-alone basis and would provide a stronger incentive than
currently exists for distributors to offer PPI at keener prices to achieve PPI sales at
the end of the prohibition period. We would expect adverse selection risks for standalone providers to reduce, as they would have access to a larger pool of potential
consumers, more similar to the consumer base of distributors (see paragraph 10.25).
We note that the Post Office and [] said that they thought additional providers
would enter the stand-alone market. Although Cattles 23 opposed a point-of-sale
prohibition, it did acknowledge that consumers, having purchased credit, and being
unable to purchase PPI at the point of sale to cover their loan from inception, may
have an additional incentive to shop around. We disagreed with the submissions
from RBSG and SDGFS that the requirement to provide information about PPI at the
credit point of sale is inconsistent with a prohibition on concluding the sale at that
stage. We concluded that the point-of-sale prohibition would stimulate consumer
search and further develop the stand-alone market and would increase incentives for
stand-alone sales by distributors to consumers who have their credit with other
providers.
Reduced take-up of PPI
10.46 Many parties said that a point-of-sale prohibition would reduce the take-up of PPI in
the short term and, in the views of many, in the longer term as well. Three main
arguments were put forward.
19
RBSG said that the effectiveness of our information provision in marketing material remedy was endangered by the point-ofsale prohibition, as the prohibition would lead to fewer companies advertising and greater withdrawal from the market. For the reasons set out above, we thought that the point-of-sale prohibition would increase competition between suppliers. 20
Defaqto—Travel Insurance 2008; Adapting to a changing world, 2008. 21
Datamonitor—UK Travel Insurance 2007, 2007.
22
Datamonitor—UK Household Insurance 2007, 2007. 23
In responding to a working paper on intermediaries. 192
(a) Unless distributors or the intermediaries arranging credit promoted PPI at the
point of sale, certain consumers would not identify their need, particularly as
consumer awareness of PPI before the point of sale was low. 24
(b) A high proportion of consumers would not bother to purchase PPI following the
credit sale as it would be significantly less convenient for them to do so. 25 Providers would be likely to find it difficult to arrange another meeting or conversation
with the consumer following the conclusion of the credit arrangement (indeed
some customers specifically ask not to receive any marketing or other calls from
their credit provider).
(c) Distributors and the intermediaries who arranged credit would be unable to contact credit consumers regarding PPI after the credit sale as many consumers do
not want to be contacted at a later date for marketing purposes. 26
10.47 Some examples were put forward to indicate the potential scale of reduction in sales.
Nationwide said that only 3 per cent of its unsecured personal loan and credit card
consumers had bought PPI or altered an existing PPI policy to cover a new credit
agreement after it temporarily withdrew its PPI product in August 2007. 27 Nationwide
told us that this indicated that only a small number of people would buy PPI if it was
not sold at the point of sale. Parties also said that consumer complaints might or
would rise; different reasons were put forward by different parties, including that consumers would not understand why they could not purchase at the point of sale and
that some consumers might become suspicious of the product. 28
10.48 Parties put forward further evidence in relation to the convenience associated with
purchasing PPI at the credit point of sale:
(a) Convenience featured prominently in reasons given for purchasing PPI at the
credit point of sale. For example: in GfK research [] 91 per cent of respondents
cited convenience as a reason for simultaneous purchase of PPI and credit.
Barclays submitted that 75 per cent of its PLPPI consumers considered that the
best time to purchase PPI was before or at the same time as applying for the
loan. 29 An ABI-commissioned YouGov survey found that 58 per cent of
respondents said they would prefer to buy PPI at the point of sale. 30
24
ABI response to Remedies Notice p6, Axa response to the Remedies Notice, p6, Capital One response to Remedies Notice
p3, Legal and General response to Remedies Notice pp3&4, Lloyds TSB response to the Remedies Notice p15, RBSG
response to the Remedies Notice p11. London General Insurance response to the provisional decision on remedies p1. Cattles
response to the provisional decision on remedies pp2&9.
25
Abbey response to Remedies Notice p10, BBA response to Remedies Notice p4, Citizens’ Advice response to Remedies
Notice pp10&11, Genworth response to Remedies Notice pp6&7, Lloyds TSB response to Remedies Notice p16, RBSG
response to Remedies Notice pp11&12. HBOS response to the provisional decision on remedies p8, HSBC response to the
provisional decision on remedies pp4&18, RBSG response to the provisional decision on remedies pp1&6, Nationwide
response to the provisional decision on remedies p3, and BBA response to the provisional decision on remedies pp1&2.
26
Barclays response to the Remedies Notice, p15, BBA response to Remedies Notice p4, Banque PSA response to Remedies
Notice p6, Council of Mortgage Lenders response to Remedies Notice p5, FLA response to Remedies Notice p21, Lloyds TSB
response to the Remedies Notice, p16. The FLA said that a large proportion of credit consumers did not want to be followed up
with additional marketing materials. The BBA said that a significant proportion of bank consumers used the option of ruling out
any further contact from their bank at the credit point of sale. According to data it had received from its members, this applied to
up to 76 per cent of all bank consumers in some cases. HBOS response to the provisional decision on remedies p8.
27
On 16 August 2007 Nationwide temporarily stopped selling CCPPI and PLPPI. The revised sales process included a discussion on PPI and a suggestion of useful sources for information regarding PPI. In January to April 2008 Nationwide contacted
946 of its consumers and found that 5 per cent of them had gone on to get a PPI quote and 1 per cent had gone on to purchase
a new PPI policy, while 2 per cent altered an existing policy to get the additional cover to protect the new credit.
28
Axa response to Remedies Notice p7, RBSG response to Remedies Notice p11.
29
Barclays also referred to a Money Marketing article (in its response to the provisional decision on remedies, p5) which stated
that 75 per cent of the people questioned thought that the sale of PPI at the point of sale should not be banned. However, there
was no indication of the survey specifications and it appeared to be a simple news commentator poll which we were unable to
place any reliance on.
30
We were told that the survey included around 2,000 participants.
193
(b) Other ‘natural experiments’ have been submitted as showing the potential impact
on take-up rates of breaking the credit point of sale. For example, HBOS carried
out a pilot where consumers at the loan point of sale were handed a telephone to
discuss PPI with a specialized insurance team. This resulted in a halving of takeup rates. Capital One told us of evidence of a month in October 2007 when it did
not offer PPI through the Internet channel at the application stage. It saw only a
small uplift in take-up at the activation stage, ‘demonstrating that even just a few
days following the point of credit application, consumer engagement is significantly reduced’.
10.49 AXA (and others) referred us to behavioural research it had commissioned (‘Turning
Good Intentions into Actions’, by Decision Technology) 31 which showed that, in
experimental situations, even small delays could significantly reduce the likelihood of
taking action. In this context, we also looked at the change in the number of
LifeChoices policies and PPI policies that HSBC sold through its branch network. In
December 2007, HSBC withdrew advised sales in branches of PLPPI and instead
offered personal loan consumers the opportunity to discuss their wider protection
needs with a Financial Planning Manager, with LifeChoices one of the products
offered to those consumers. 32 Figure 2 of Appendix 2.3 showed that there was a
decline in PLPPI sales for which the increased sales of LifeChoices did not fully compensate, and we considered that this could be explained by the delay in the time from
when the consumer buys the credit product to the time they spoke to the Financial
Planning Manager to discuss their protection needs.
10.50 While we acknowledge that this element of the remedies package reduces the convenience of purchasing PPI at the credit point of sale, we consider that the potential
reduction in PPI sales has been overestimated by some parties. By increasing competition and thereby reducing price, we expect our remedies package to lead to an
increase in PPI sales that would partially or fully offset a decline from a reduction in
convenience. Some of the estimates of the potential drop in PPI sales submitted to
us by the parties have been based on analysis of PPI take-up rates after the credit
point of sale, at which point the majority of consumers who were interested in buying
PPI have been able to do so. 33 In Nationwide’s case, the estimates of the potential
fall in penetration were based on a situation where stand-alone PPI products are not
widely promoted, at a time before the FSA comparison website was available and
where Nationwide was not selling PPI at all. To the extent that this measure reduces
the convenience for some customers of purchasing PPI compared with buying it at
the credit point of sale, we are satisfied that this is both necessary to stimulate
competition so as to contribute towards remedying the AEC identified and justified in
light of the scale of the detriment identified.
10.51 We have considered the arguments put to us concerning the risk of reduced take-up
of PPI and we have designed our remedies package to reduce the risks of a substantial fall in take-up. Permitting distributors and intermediaries to market PPI at the
credit point of sale (see paragraph 10.98) and requiring distributors and intermediaries to provide a personal PPI quote (see Figure 10.1) would significantly reduce
the risk that consumers would not consider the product due to lack of awareness of
PPI when taking out credit. Our remedies package also allows for consumers—on
their own initiative—to buy PPI directly from the distributor or intermediary over the
31
www.dectech.org/Links/Brief-AxaPart2.pdf. Advised sales in branches of CCPPI were also withdrawn at this time. HSBC’s MPPI product had been replaced by
LifeChoices earlier in 2007. See paragraphs 3 to 7 of Appendix 2.3.
33
For this reason the experiences of underwriters in their previous attempts to compete using a stand-alone product were only
of limited usefulness when assessing the likely effects of our remedies package on PPI take-up. This applies, for example, to the experience of Norwich Union's initiative to establish a stand-alone PPI product through Norwich Union Direct, in May 2002. 32
194
Internet or over the telephone 24 hours after the credit sale. This would reduce the
inconvenience associated with a delayed sale and the follow-up costs for lenders in
marketing to these consumers. We also note that advertising is not restricted to direct
marketing to individual consumers: television and press advertising can be used
effectively to obtain sales. In this regard, if distributors were prevented from following
up with their consumers individually—for example, because customers had
requested not to be contacted after the credit sale—they would still have the
opportunity to attract sales through other forms of marketing. 34
10.52 We were also told that the point-of-sale prohibition would mean that fewer credit
sales would be made. [] said that a consequence [] was that it had significantly
increased its credit cut-offs, resulting in a large reduction in credit sales. It said that a
point-of-sale prohibition would result in a combination of losses to consumers through
higher credit prices and reduced credit availability, the cost to consumers of adverse
selection, and the cost to consumers of additional search. We concluded in
paragraph 10.483 that, with increased competition in PPI markets, credit prices and
cut-off scores would reach a new equilibrium, and we concluded in paragraph 10.492
that we should not amend our remedies to take account of a relevant customer
benefit of lower credit cut-off scores.
Reduced consumer choice
10.53 Many parties said that any point-of-sale prohibition would lead to a reduction in consumer choice. The main reason put forward for this was that consumers would be
unable to purchase PPI from the distributor or intermediary who sold or arranged the
credit during the period of the prohibition. Several parties also considered that the
reduction in choice would be particularly detrimental to consumers as the PPI provided by the credit provider was generally of a better quality, or might be better value,
than that provided by stand-alone PPI providers. 35 Conversely, one party ([]) said
that the prohibition would increase consumer choice as they would have a choice of
provider and policy whereas now they were only presented with one option.
10.54 Most parties also suggested that the point-of-sale prohibition would lead to market
exit by some providers, and hence a reduced choice for consumers; 36 suggested
reasons for market exit included increased adverse selection, increased costs or
negative publicity on PPI.
10.55 [] told us that, on the information available (in our provisional decision on
remedies), the most viable option at this time was that [] would not provide PPI to
new consumers in the future. 37 [] told us that, as set out in our provisional decision
34
SDGFS told us that because it had not marketed PPI as a stand-alone product, it did not consider it realistic to advertise it
other than by direct marketing.
35
Abbey response to Remedies Notice p10, Aviva response to Remedies Notice p14, Barclays response to Remedies Notice
p16, MBNA response to Remedies Notice p7.
36
Abbey response to the Remedies Notice, pp10&11, Aviva response to the Remedies Notice, pp1&15, Banque PSA response,
Barclays response to the Remedies Notice, p16, HSBC response to the Remedies Notice, p12, Lloyds TSB response to the
Remedies Notice, p16. Lloyds TSB response to provisional decision on remedies p16, RBSG response to provisional decision
on remedies pp2&11, Egg response to provisional decision on remedies p2, AXA response to provisional decision on remedies
pp3&4, Abbey response to provisional decision on remedies p1, Nationwide response to provisional decision on remedies
pp2&3, Aviva response to provisional decision on remedies p2, Paymentshield response to provisional decision on remedies
pp1&3–4, HBOS response to provisional decision on remedies pp1&2, Everyday Loans response to provisional decision on
remedies pp1&2, GMAC response to provisional decision on remedies pp3&4, and BBA response to provisional decision on
remedies pp1&2. Express Gifts response to provisional decision on retail PPI remedies pp3–5, SDGFS response to provisional
decision on retail PPI remedies pp33&44 and BRC response to the provisional decision on retail PPI remedies, p3.
37
[] submitted internal documents underpinning this view, including an assessment of the financial implications of a point-ofsales ban, based on a scenario in which the remedy gave rise to a 95 per cent drop in penetration for CCPPI and an 86 per
cent drop in penetration rate for PLPPI. This was somewhat higher than [] initial estimate, provided in its response to the
Remedies Notice, that sales would reduce to one-third of the current level based on lower levels of sales coming from postpoint-of-sale marketing.
195
on remedies, the impact of this remedy would result in [] ceasing to offer PPI to
any new consumers. [] told us that it had already started to consider carefully the
future options for PPI and that withdrawal from the market was an option that it was
considering. [] said that it would need to consider the economics of selling PPI in
such an environment, including what the success rate would be in contacting the
consumer and getting them to listen to advice and to take the product. [] said that if
a prohibition were imposed, a stand-alone PPI proposition might be necessary to
continue to operate in the market.
10.56 For distributors and intermediaries, the decision to stay in the market would depend
on how many consumers would still buy PPI and how many they could access—in
other words, it would be a straight business decision based on expected returns. We
noted the views of distributors who said that they or others would stop selling PPI.
However, we considered that incentives would remain for distributors and intermediaries to continue to provide PPI to their credit consumers. PPI is a significant source
of income, for which consumers have demonstrated a considerable willingness to
pay. Margins on PPI are currently high (see Appendix 4.4) though we would expect
them to decrease with increased competition. 38 PPI sales also benefit distributors,
whose repayments become insured as a result. We expect that even if marketing
costs increase and/or penetration rates fall as a result of this remedy, many
distributors and intermediaries would still find it beneficial to continue to sell PPI. To
the extent that particular distributors or intermediaries choose to exit the market, this
would create an opportunity for others to offer PPI on a stand-alone basis to their
consumers.
10.57 We concluded that this remedy would not reduce consumer choice. If consumers
wish to buy the distributor’s product or products offered to them by the intermediary
arranging the credit, they can do so on their own initiative from 24 hours after the
credit sale. Alternatively, consumers can use the seven-day prohibition period to
shop around for alternative policies. While some distributors and intermediaries may
leave the market, the number of alternatives for individual consumers is likely to rise
as distributors and specialist stand-alone providers are more likely to offer more
stand-alone PPI policies, because of the increased pool of consumers searching for
such policies. We concluded that the remedy will increase the choice of PPI provider
for most consumers from one to many.
Lower-quality products
10.58 We were told that a prohibition on selling PPI at the credit point of sale, or any option
that increased stand-alone provision, would lead to a decline in quality and a lesstailored or comprehensive PPI product. 39 Parties highlighted our provisional findings
which, they said, indicated that in general the stand-alone providers had less featurerich policies and longer exclusion periods. 40 They also said that as stand-alone providers had greater adverse selection issues, it was unlikely that the policies could
ever be of the same quality. Furthermore, distributors highlighted their unique
position, not shared by stand-alone providers, of being able to pay the insurance
against the loan—thus not affecting state benefits—and being able to assess the
38
The aggregate income as a percentage of GWP was 69 per cent for all PPI, 66 per cent for PLPPI, 77 per cent for CCPPI and 52 per cent for MPPI. 39
See, for example, Abbey response to the Remedies Notice, paragraph 4.1.5, ABI response to the Remedies Notice, p7, Aviva response to the Remedies Notice, p16, RBSG response to the provisional decision on remedies pp1,2&14, ABI response to the provisional decision on remedies pp1&2 and London General Insurance response to the provisional decision on remedies pp2–
4&6. 40
See, for example, Abbey response to the Remedies Notice, paragraph 4.1.5, Aviva response to the Remedies Notice, p16. 196
combined affordability of the credit and the PPI so that consumers did not become
over-indebted.
10.59 Whilst there are stand-alone products which offer a lower level or quality of cover,
these products are also generally sold at a significant discount to those sold at the
point of sale. However, using Defaqto score as a comparative indicator of quality, 41
we noted that some stand-alone MPPI products achieve similar scores to MPPI products sold at the point of sale (see paragraph 10.396 and Figure 10.8). We did not
carry out a similar analysis for stand-alone and point-of-sale PLPPI policies, as we
could not find a suitable metric which allowed a simple and fair comparison of standalone PLPPI, which is sold on a regular-premium basis, with point-of-sale PLPPI,
which was predominantly sold as on a single-premium basis, during the course of our
investigation (see paragraphs 2,92 to 2.90 and Appendix 2.1). Further, we consider
that the remedies package, and the point-of-sale prohibition in particular, is likely to
increase the size of the market for stand-alone PPI and that, as a result, many of the
adverse selection issues faced by stand-alone providers are likely to decrease (see
paragraph 10.25).
Higher costs for distributors
10.60 It was also put to us that the point-of-sale prohibition would lead to additional costs,
as distributors would have to contact consumers at the end of the prohibition period
and in many cases the information from the loan would have to be collected again. 42
Capital One told us that this increase in costs would mean that it would be uneconomical proactively to offer PPI to consumers with lower credit limits (who were, in
general, higher risk) and as a result higher-risk consumers might have reduced
access to PPI (although if these consumers wanted PPI they would be able to buy it).
10.61 We note that the design of the remedy is intended to mitigate some of these costs
compared, for example, with a longer prohibition period or one in which consumers
were not able to return to the credit provider on their own initiative soon after the
credit sale. We did not see any reason, in particular, why a distributor would need to
re-collect significant amounts of information in relation to a loan and a consumer only
a small number of days after issuing that loan, in order to offer PPI to the same consumer.
10.62 We acknowledge that this remedy will impose set-up and ongoing costs on distributors. Our analysis of evidence submitted to us on the costs of implementing our
remedies package can be found in paragraphs 10.497 to 10.508 and in Appendix
10.13. Our assessment of the effectiveness and proportionality of the remedies
package, including the point-of-sale prohibition, can be found in paragraphs 10.465
to 10.514.
Different impact on different providers
10.63 We were told that a point-of-sale prohibition would give an unfair advantage to standalone PPI providers. 43 A number of parties talked about the removal of a level
playing field. Barclays said that the option was ‘pro-competitor’ rather than ‘proconsumer’. However, two firms, [] and [], told us that it would give a competitive
41
We considered the merits of using the Defaqto rating system as a measure of product quality in paragraph 4.27 of our provisional findings. 42
See, for example, Capital One response to the Remedies Notice, p3, and Lloyds TSB response to the Remedies Notice, p15. 43
See, for example, Abbey response to the Remedies Notice, pp3&12, Barclays response to the Remedies Notice, p14, Lloyds TSB response to the Remedies Notice, p17, RBSG response to the Remedies Notice, pp1&13. 197
advantage to high street providers because they would be able to cross-sell PPI in
different circumstances (such as when consumers came into a branch) and they
could leverage their knowledge of consumers’ current accounts to cross-sell PPI
policies on new loans.
10.64 We consider that a point-of-sale prohibition will provide additional commercial opportunities to firms that sell stand-alone PPI, but we see no reason why distributors
could not also benefit from these opportunities by selling PPI on a stand-alone basis.
We acknowledge that distributors’ and intermediaries’ point-of-sale advantage is
diminished by this element of the remedies package, and that this may result in some
disadvantage to them compared with stand-alone providers, who could sell a PPI
policy to a customer the same day as that customer takes out credit. However, we
consider (see paragraphs 10.67 to 10.71—effectiveness of POS prohibition and
alternatives) that it is necessary to address the point-of-sale advantage in order to
remedy the AEC that we have found. By allowing a limited exemption for consumerinitiated transactions 24 hours after the credit sale, we are ensuring that consumers
can buy PPI from those parties covered by the prohibition within a similar timescale
as they could buy it from a stand-alone provider.
Restriction on the freedom of establishment
10.65 One party ([]) also raised concerns that the prohibition could amount to a restriction of the right of freedom of establishment under EU law, which could only be
justified where overriding requirements of public interest were at stake (such as protection of consumers), and provided that the measures proposed were suitable for
securing the attainment of this objective and did not go beyond what was necessary
in order to attain this objective.
10.66 Whilst we do not agree that any part of the remedies package would infringe the right
of freedom of establishment, we note that: the aim of our remedies package, in
accordance with the aims of the market investigation regime, is to ensure that where
market features lead to consumer detriment, these are addressed thereby safeguarding the overriding public interest in markets working well for consumers through the
promotion of vigorous competition; 44 and proportionality is in any event a relevant
consideration under the Act. The test that we are obliged to apply in considering
remedies under the Act requires us to have regard to the need to achieve as comprehensive a solution as is reasonable and practicable to the AEC and any detrimental
effects on consumers resulting from the AEC. 45 In addition, our guidance makes
clear that in considering remedies we will have regard to the principle of
proportionality. 46 Our assessment of the effectiveness and proportionality of the
remedies package can be found in paragraphs 10.465 to 10.514.
44
The White Paper that led to the introduction of the market investigation regime, ‘A World Class Competition Regime’, 30 July
2001, states in paragraph 1.1:
The importance of competition in an increasingly innovative and globalised economy is clear. Vigorous competition
between firms is the lifeblood of strong and effective markets. Competition helps consumers get a good deal. It
encourages firms to innovate by reducing slack, putting downward pressure on costs and providing incentives for
the efficient organisation of production. As such, competition is a central driver for productivity growth in the
economy, and hence the UK's international competitiveness.
It goes on to state in paragraph 1.2 ‘Competition between firms protects consumers. In markets which lack effective competition, firms know that customers have little or no choice but to buy from them and so they can treat their consumers unfairly’. We consider that it is clear from this that, by ensuring effective competition, our remedies package is justified as an overriding requirement in the public interest. 45
The Act, section 134(6). 46
CC3, paragraph 4.9. 198
Effectiveness of the remedy and alternatives to a point-of-sale prohibition
10.67 During consultation on our proposed remedies, several parties suggested alternatives to the point-of-sale prohibition which they considered would address the AEC in
a more proportionate manner than the point-of-sale prohibition that we have decided
to implement. These alternatives are listed below:
(a) Some parties suggested continuing to sell PPI at the credit point of sale, but
emphasizing or increasing the cooling-off period in which PPI could be cancelled
without cost to the consumer (for example, increasing it to 90 days). 47 In addition,
the consumer could be given information about the existence of other PPI
policies at the point of sale and use the cooling-off period to search for alternatives.
(b) Lloyds TSB suggested that the point-of-sale prohibition was not necessary if,
instead, PPI providers were obliged to provide ‘opt-out’ renewal terms to PPI
consumers every 12 months as part of the annual statement provided to consumers. It said that this proposal, together with the remaining remedies package,
would stimulate consumers to shop around for PPI on renewal and avoid the
detriment of consumers suffering a lapse in PPI cover. We discuss this as a
remedy option in paragraphs 10.345 to 10.347. We consider here whether this
remedy option would be effective as part of the remedies package instead of the
point-of-sale prohibition.
(c) Two parties ([]) suggested an access remedy as an alternative to this remedy.
This would involve a number (possibly three or four) of PPI providers having
access to the consumer at the point of sale. The credit provider would have PPI
policies from these PPI providers available at the point of sale of credit and would
advise on the most appropriate one for the consumer’s needs. One party ([])
said that this option would have to be introduced with [] an IT system that took
the decision out of the hands of the salesperson. The remedy envisaged that
sales would be made on an advised basis.
(d) Which? suggested that all PPI sales should be advised and that sales personnel
who sold PPI would have to explain to the consumer why PPI was better than an
income protection product. It said that this requirement would result in a similar
system to how pensions were sold under the RU64 FSA rules for pension
advisers. 48
10.68 We consider that the alternative suggestions from the parties would either be more
complicated to monitor and likely to be ineffective or would not address the AEC that
we found. In particular:
(a) Increasing the cooling-off period would not effectively address the AEC as consumers seem less inclined to change policies once they have purchased them,
even when they consider that they could get a better value-for-money product
elsewhere. 49 For example, Cardif Pinnacle said that it was difficult to convince
consumers to cancel an existing policy once they had made the emotional commitment of making a purchase, and the selling process in these situations was
therefore more difficult. We note that, on average, 5 per cent of consumers
47
See, for example, Barclays response to the Remedies Notice, p18, HSBC response to the Remedies Notice, pp11&12, Lloyds TSB response to the Remedies Notice, p17. 48
RU64 requires an adviser to explain to a consumer in writing why the personal pension they are recommending is at least as suitable as a stakeholder pension—source www.fsa.gov.uk/pages/Library/Communication/Statements/2007/ru64.shtml. 49
This is sometimes referred to as ‘status quo bias’—see FSA, Financial Capability: A Behavioural Economics Perspective, 2008. 199
between January 2006 and December 2007 cancelled their PPI policy in the first
30 days, and at least some of those did not switch to another policy. Further, we
found that few consumers who did compare policies revisited their purchase
decision during the cooling-off period (see paragraph 5.109), and we found low
levels of switching during the cooling-off period noted, which indicated to us that
few consumers are switching within this initial period (see paragraph 4.41).
(b) We consider that a requirement for PPI providers to offer consumers the opportunity to ‘opt-out’ of a PPI policy on an annual basis, together with the other
elements of the remedies package (excluding the point-of-sale prohibition) would
not effectively address the AEC. This proposal would not address the point-ofsale advantage at all, but would instead put in place an additional remedy to
promote switching. We concluded that a remedies package with this focus would
have insufficient impact and emphasis on the point-of-sale advantage and
barriers to search effectively to address the competition problems that we found.
Whilst we decided that the single-premium prohibition and the annual statement
would promote competition through switching, and that it would therefore not be
necessary also to require an annual renewal on an opt-out basis in order to
address that aspect of the AEC, we concluded that a stronger up-front intervention was required to address the other aspects of the AEC that we identified. As a
result, the proposal put forward by Lloyds TSB would not be a comprehensive
solution to the problems that we had identified in this market, even in conjunction
with all of the other elements of the remedies package. We concluded that this
proposal would not be sufficient to stimulate competition to such an extent that
the point-of-sale prohibition would no longer be required.
(c) Controlling the sales process to require the sale or promotion of alternative PPI at
the point of sale would make any access remedy very challenging and difficult to
monitor and enforce, particularly given the benefits to distributors of selling PPI
and the benefit of selling their own PPI in a vertically-integrated firm.
(d) A requirement to sell PPI on an advised basis would not, in our view, address the
competition issues we have identified, as an advised sale relates only to the suitability of the product or products offered by that distributor or intermediary.
(e) In addition, ensuring that consumers are told about income protection products
would not necessarily address the lack of shopping around or significantly
decrease the point-of-sale advantage.
10.69 We also considered whether the provision of a personal PPI quote at the point of sale
would be sufficient, alongside the other measures we are implementing and an existing or extended cooling-off period, to remedy the AEC without the need for a point-ofsale prohibition. 50 As noted in paragraph 10.68(a), we found that few consumers who
did compare policies revisited their purchase decision during the cooling-off period. In
contrast, we noted that a recent study on a remedies package implemented for
extended warranties (which included the incorporation of a written quotation but not a
point-of-sale prohibition) indicated that, taken as a package of remedies, there was
some impact on consumer behaviour as a whole. However, in that study there was
little evidence to suggest that the observed change in consumer behaviour could be
attributed to the informational remedy of the written quotation, which appeared to
have had little impact in the absence of a clean break in time between the sale of the
50
Both HSBC and RBSG said that with a ban on single-premium PPI and information remedies in place the rationale for the
point-of-sale prohibition had been removed and it was therefore not required.
200
primary and secondary product. 51 We concluded that, without a point-of-sale
prohibition, providing a personal PPI quote at the point of sale, after the consumer
has taken out the loan and during the sales process for a linked PPI product, would
not have a sufficient effect on a consumer’s searching behaviour for such a remedies
package to be effective in remedying the AEC identified.
10.70 We also found in paragraph 5.87 that there are significant barriers to entry for standalone PPI providers seeking to sell PPI products without access to consumers at the
credit point of sale, due to adverse selection, poor consumer awareness and high
marketing costs. These factors would not be addressed by the provision of a personal PPI quote, as part of a package of information remedies. We concluded that a
package of information remedies alone would not be sufficient and an additional
structural measure would be needed to achieve a comprehensive solution to the AEC
that we found.
10.71 We decided that a prohibition on selling PPI at the credit point of sale was a necessary part of the remedies package that we have identified as a comprehensive,
reasonable and practicable solution to the AEC that we found. It complements and
enhances other measures to address barriers to search, in particular the personal
PPI quote (paragraphs 10.157 and 10.158), provision of information in marketing
material (paragraphs 10.182 and 10.183), provision of information to third parties
(paragraphs 10.222 and 10.223) and unbundling retail PPI from merchandise cover
(paragraphs 10.278 and 10.279).
Conclusion on the need for a point-of-sale prohibition
10.72 We concluded that the point-of-sale advantage contributed significantly to the AEC
that we had identified. Given the severity of the competition problems and the scale
of the resultant consumer detriment, we concluded that it was necessary to introduce
a remedies package that would lead to a new, more competitive, market structure.
10.73 We acknowledge that the point-of-sale prohibition will represent a very significant
change to current PPI sales practices and noted the strenuous opposition by the
parties to this element of our remedies package. We considered carefully the extent
to which any of the arguments put to us regarding the alleged risks of the point-ofsale prohibition were well founded and accordingly whether we should revisit our provisional decision on this element of the remedies package. However, we concluded
that this prohibition, taken together with other elements of our remedies package,
was the only effective way to address key aspects of the AEC that we found.
10.74 It was put to us that the point-of-sale prohibition would not eliminate the incumbency
advantage of distributors and intermediaries (paragraph 10.40). We agreed that this
was the case but concluded that it was not necessary to eliminate all aspects of the
incumbency advantage of distributors and intermediaries in order effectively to
51
Evaluating the impact of the Supply of Extended Warranties on Domestic Electrical Goods Order 2005, prepared for the OFT
by LECG, October 2005. LECG found that consumers did not always receive the quotation and where they did there was little
evidence that they used it for shopping around. Following publication of our provisional decision on remedies, LECG wrote to
us noting that some elements of the evaluation showed a clear improvement in consumer behaviour, in particular the increase
in the proportion of consumers shopping around from 4 to 15 per cent and the reduction in the proportion of consumers who
received their extended warranties from the point-of-sale supplier from 82 to 68 per cent (although roughly two-thirds of this
change was due to the offer of free extended warranties). LECG also noted that the estimate of the reduction in consumer
benefit was a rough but conservative estimate and that even if it were agreed that the effect of the extended warrantees
remedies on consumer behaviour was small, the preferences of consumers to buy at the point of sale meant that there was no
reason to assume that stronger remedies would make consumers better off. We agreed with LECG that the extended
warranties remedies had had some impact on consumer behaviour as a whole, but did not consider that this invalidated our
views about the overall effectiveness of the package or, specifically, about the effectiveness of requiring the provision of a
written quotation without also addressing the point-of-sale advantage.
201
address that element of the AEC (paragraph 10.41). We were also told that we could
not be certain that our remedy would be sufficiently effective in stimulating consumer
search and developing competition. However, we concluded that by creating a clear
break between the sale of credit and the sale of PPI and providing consumers with
the tools they require to compare PPI policies, this would create significant incentives
for consumers to search for the best value-for-money PPI policy which we expect will
stimulate competition (paragraph 10.43). This would also open up the possibility for
substantially greater sales to be made on a stand-alone basis and judged that the
remedy was also likely to increase the incentives to increase spend on marketing
products (paragraph 10.45).
10.75 We were told that a point-of-sale prohibition would reduce take-up of PPI. Whilst we
acknowledge that a point-of-sale prohibition reduces the convenience of purchasing
PPI, we concluded that the risks of reduced take-up had been overstated by the
parties (paragraph 10.50). We expect that our remedies package, by increasing
competition and reducing prices, would lead to an increase in sales and partially or
fully offset any decline from a reduction in convenience. We also designed aspects of
the point-of-sale prohibition specifically to minimize this risk. In particular, we will
permit, as a limited exception to the point-of-sale prohibition, consumer-initiated sales
after a short interval, provided that the consumer has the tools and the opportunity to
search the market for the best-value PPI policy that meets their needs (paragraph
10.51).
10.76 We also concluded that the arguments put to us that this prohibition would lead to
reduced consumer choice were unfounded. Consumers who wish to buy PPI from
the distributor or intermediary arranging the credit will continue to be able to do so on
their own initiative after a short interval (as specified in the limited exception to the
prohibition) or as a result of further marketing by the distributor or intermediary following the end of the prohibition period. In the meantime they will have had the opportunity to search the market and the tools to facilitate this. By stimulating competition
between PPI providers, we expect there to be greater choice for consumers (paragraph 10.57). It was also put to us that stand-alone policies are of lower quality than
those sold at the point of sale and that our package would therefore lead to a decline
in product quality, but this was not borne out by the evidence (paragraph 10.59).
10.77 We were told that this prohibition would lead to increased costs for distributors and
intermediaries and we recognize that, compared with current sales practices, this is
the case. To address this concern, in designing the point-of-sale prohibition we have
sought to mitigate some of these costs (paragraphs 10.60 to 10.62). It was put to us
that this prohibition gives an unfair advantage to stand-alone providers. However, we
concluded that some disadvantage to distributors and intermediaries as they
currently operate would be necessary in order to create a more competitive market
structure, and we could see no reason in principle why distributors could not also
seek to supply PPI on a stand-alone basis to consumers who had obtained credit
elsewhere (paragraphs 10.63 and 10.64). We also concluded that the arguments put
to us that the prohibition would be unlawful under EU law were unfounded (paragraph 10.66).
10.78 In view of the scale of the competition issues and resultant consumer detriment that
we have identified, we considered that imposition of some cost on distributors and
intermediaries was justified in order to achieve an effective remedy to the AEC.
Moreover, we were confident that our remedies package would be an effective and
proportionate solution to the very significant competition problems and resultant
consumer detriment that we had identified.
202
10.79 We considered alternatives to the point-of-sale prohibition put to us and whether
these would effectively address the AEC and resultant consumer detriment that we
had found, and we concluded that these alternatives would not be effective in doing
so (paragraphs 10.67 to 10.70). We decided therefore that a prohibition on selling
PPI at the credit point of sale was a necessary part of the remedies package
(paragraph 10.71) in order to achieve as comprehensive a solution to the AEC and
resultant consumer detriment as was reasonable and practicable.
The design of the point-of-sale prohibition
10.80 In paragraphs 10.40 to 10.66 we considered a number of risks that have been raised
by the parties about imposing a point-of-sale prohibition. In paragraphs 10.67 to
10.71 we reviewed a number of alternatives to a point-of-sale prohibition that we
ultimately concluded were not sufficient to remedy the AEC that we found. In paragraphs 10.81 to 10.156 we consider the specific design of the point-of-sale prohibition remedy and how it might best maximize the effectiveness of the remedy, while
avoiding unnecessary burdens on PPI providers and consumers and how it is likely
to interact with other regulations in this area. The following issues arise in relation to
the design of the remedy:
(a) whether there should be a prohibition on providing information about PPI at the
credit point of sale;
(b) the information that must be provided to the consumer about PPI at the credit
point of sale;
(c) the interaction with ICOBS and the extent to which the distributor or intermediary
can carry out elements of the PPI sales process at the credit point of sale;
(d) the application of the point-of-sale prohibition; and
(e) whether there should be any exemptions to the point-of-sale prohibition for
particular products, distribution channels or types of retailer.
Whether there should be a prohibition on providing information about PPI at the
credit point of sale
10.81 We looked first at whether distributors and intermediaries should be prohibited from
providing any information about their PPI products at the credit point of sale.
However, we decided that providing information about PPI at the credit point of sale
would not undermine the effectiveness of the remedies package and that this could
enhance some aspects of the package of remedies. In this context, we noted the
submission of one party that a discussion of PPI options enabled lenders to meet
their duty of care to their consumers (although views on this point were mixed) 52 and
would enable consumers and lenders to consider the overall affordability of credit
with PPI. 53 We also noted some parties’ (Egg, Abbey, ABI) submissions that
52
One party (RBSG) raised concerns that a point-of-sale prohibition would impact on its duty of care to its consumers. It said
that regulatory requirements meant that it had to have appropriate discussions with credit consumers about the consequences
of entering into credit arrangements and the steps that they could take to protect their interest before selling credit to them. In
RBSG’s view, these requirements would require it to inform consumers about the existence of protection products such as PPI.
However, HBOS did not agree and said that it had found nothing in the Consumer Credit Act 2004 which would require that the
credit provider gave advice on or offered protection for the credit product. Another party (Legal & General) said that for MPPI
the point-of-sale prohibition would be in direct conflict with the Government’s sustainable house ownership initiative where consumers were encouraged to make provision for times of hardship.
53
Some distributors and the FLA said that lenders needed to assess affordability of the package. Lloyds TSB said that it needed
to inform consumers about the overall affordability of the loan and PPI so that they would have a clear view of the cost.
203
consumers’ engagement with PPI was likely to be particularly high at the credit point
of sale. We concluded that there should not be a prohibition of providing information
about PPI at the credit point of sale.
Information provided about PPI at the credit point of sale—the personal PPI quote
10.82 Next we considered whether distributors and intermediaries should be required to
provide particular information about PPI at the credit point of sale, in the context of a
point-of-sale prohibition. In particular, we considered whether, to help stimulate
search and assist comparisons of PPI policies after the credit sale, distributors
should be required to provide consumers with a personal quote setting out the price
of taking PPI and key messages (such as where to find information about other forms
of PPI and a reminder that PPI is optional).
10.83 The Post Office said that its preferred outcome was that credit providers would
encourage consumers to consider their protection needs through a number of agreed
statements and be able to offer a firm quote and product information to assist the
consumer in comparing alternatives. However, SDGFS told us that this requirement
would be inconsistent with other parts of the remedies package as it was counterintuitive to provide consumers with significant amounts of information about PPI but
then tell them that they could not purchase it straight away.
10.84 The FSA told us that, to be effective, the information provided would have to be given
to all consumers who requested it rather than just to consumers who went through
the sales process, as was currently the case, to enable consumers to use them for
comparative purposes.
10.85 In our view, identifying the appropriate information that should be required to be provided about PPI at the credit point of sale—or after the credit point of sale, but before
the purchase of PPI—depends on a number of factors:
(a) the amount of information required by consumers to make an informed choice
when shopping around (after the point of sale);
(b) the cost to the distributor of providing the information (both in terms of interview
time and marketing materials);
(c) the need to ensure that consumers who would benefit from PPI are informed
about it; and
(d) the need for credit providers to exercise their duty of care to consumers as well
as the need for consumers to be able to assess the affordability of credit with or
without PPI.
10.86 The most appropriate point in time at which a personal quote should be provided to
customers is likely to vary from case to case. In the case of credit cards, Capital One
suggested that a simplified personal quote and policy summary could be provided to
consumers at the point of credit application or the point of credit acceptance. We
noted that in other situations it could make sense for a personal quote to be provided
at the conclusion of the credit sale or, if this were not convenient, at another, later,
date.
10.87 We decided that requiring the provision of a PPI personal quote, as a precondition of
selling PPI, following a credit sale, would enable consumers to compare the distributor’s PPI policies against others in the market and make an informed choice as to
which would most suit their needs and their income. It would also help consumers to
204
search for the best combination of credit and PPI and make use of the cooling-off
period on the credit sale to switch credit provider if they find a better combination of
credit with PPI. In addition, the provision of a personal PPI quote is consistent with
any duty of care that RBSG considered it owed to its consumers (see paragraph
10.81). Our consideration of the content of the personal PPI quote may be found in
paragraphs 10.160 to 10.181. We also decided that if PPI is mentioned at the credit
point of sale, the key messages (see Figure 10.3) should be orally disclosed.
10.88 We decided that if a distributor or intermediary provides information about PPI during
the credit sales process, they should also be required to provide a personal PPI
quote in a durable medium. 54 We noted that some distributors or intermediaries may
not offer PPI at the credit point of sale because it is an optional financial product. If
no information is given about PPI during the sales process, then there is no need to
provide a personal PPI quote at that stage (although a personal quote would have to
be provided to the customer before the distributor or intermediary could make a PPI
sale—see paragraph10.87). No quote would be required if the consumer did not
meet the relevant eligibility criteria for a PPI policy.
The interaction with ICOBS and the extent to which the distributor or intermediary
can carry out elements of the PPI sales process at the credit point of sale
10.89 A prohibition on selling PPI during the credit sale raises the possibility that a sale of
PPI by a distributor or intermediary after the prohibition period had expired could be
carried out over the course of two or more distinct sales events (ie provision of information about PPI at the credit sale and the subsequent transaction at which the
customer takes out a PPI policy). We explored with the FSA the potential implications
of this change to the current sales process for the way it regulates PPI sales.
10.90 We thought that, in principle, there were two options relating to the extent that distributors subject to a point-of-sale prohibition were able to market PPI at the credit point
of sale:
(a) Distributors and intermediaries could be limited to giving a general description of
the terms and conditions of the PPI products that they offer, find out basic
information about eligibility of the consumer (eg age, country of residence and
employment status) and provide the consumer with a personal PPI quote. In
order to finalize the sale, the distributor would need to take the consumer through
a full sales process at a subsequent PPI sale.
(b) Alternatively distributors and intermediaries could go through a fuller PPI sales
process at the credit point of sale, for example conducting a demand and needs
analysis, making all necessary ICOBS disclosures and potentially making a personal recommendation as to suitability or otherwise of a particular PPI product,
but not concluding the transaction. In this scenario, the consumer would leave
with a credit agreement and a personal PPI quote for a product that had been
recommended as being suitable for the consumer, but the PPI sale would be
concluded at a later date.
10.91 The FSA considered this issue in relation to its regulatory duties and also gave us its
views about the potential effectiveness of the remedy in addressing the AEC we
found. From a regulatory perspective, the FSA was concerned about the implications
for the integrity of the ICOBS sales disclosures of permitting companies to conduct a
54
The FSA defines ‘durable medium’ at http://fsahandbook.info/FSA/html/handbook/Glossary/D. In summary, the definition
includes paper and a number of electronic storage devices.
205
fuller PPI sales process at the credit point of sale. The FSA was concerned, in particular, that ‘there is a potential for some of the disclosures required under ICOBS to
be omitted or not appropriately undertaken if the point-of-sale prohibition results in
the ICOBS sales dialogue, being split over several interactions between the credit
provider and the consumer’. The FSA concluded that firms could not rely on oral
information about PPI given during the credit transaction as meeting the FSA’s oral
disclosure requirements and informed decision standard for PPI sales, and that ‘the
full ICOBS sales discussion and disclosures should take place not at the credit point
of sale or PPI quotation point [where those are different] but at the subsequent
contact between the consumer and the credit provider when the prohibition period
has come to an end’.
10.92 The FSA also had concerns about the impact of permitting a fuller sales process on
the effectiveness of the remedy. It was concerned that the point-of-sale prohibition
would be undermined if firms could undertake the substance of the PPI sale at the
time of the credit sale. The FSA said that there was a risk that firms could seek to
secure the consumer’s agreement, as at present leaving only the formal ticking of a
box to complete the deal on hold until the end of a waiting period prescribed by the
CC. If this were to happen, the FSA was concerned that consumers would then be
much less likely to shop around, in which case the point-of-sale advantage of credit
providers may be largely untouched. We were referred to research showing that, in
the context of insurance sales, consumers are highly susceptible to seller influence
and persuasive salesmanship, 55 and that consumers tend to stick with their prior
choices. 56
10.93 The FSA told us that it will continue to look closely at the full range of regulatory tools
and actions available to ensure that firms selling PPI meet its requirements to enable
customers to make informed purchasing decisions. This may include consulting on
new rules if necessary.
10.94 Lloyds TSB commented on the views of the FSA. It said that, unless information and
advice about PPI could be provided at the credit point of sale, advertising of PPI and
a general terms discussion is unlikely to be effective to enable consumers to identify
the potential benefits of buying PPI and to shop around accordingly. It said that many
firms were likely to wish to adopt an advised sales process and, if they were not permitted to complete the advised sales process at the credit point of sale, they would
find it much more costly to complete that process later. We did not agree with Lloyds
TSB that a general terms discussion would be insufficient to enable consumers to
shop around, but noted its comments about firms’ likely preferences and costs.
10.95 We noted the FSA’s concerns about allowing a fuller sales process at the credit point
of sale. It was clear that, if firms were permitted to offer more than a general terms
discussion at the credit point of sale, the FSA would need to ensure that this did not
undermine the integrity of the ICOBS disclosures. This might require these disclosures to be repeated at the subsequent PPI sale, in particular if this took place some
time after the credit sale.
10.96 We considered whether, in order to ensure the effectiveness of our remedies in
addressing the AEC that we found, it would be necessary to prohibit firms from
having any more than a general terms discussion of PPI at the credit point of sale.
We found this to be finely balanced. We noted the possibility that a fuller sales discussion could create a sense of obligation to return to the credit provider. In addition,
55
www.fsa.gov.uk/pubs/other/DeMeza_Report.pdf.
www.fsa.gov.uk/pubs/consumer-research/crpr69.pdf.
56
206
if completion of the PPI sale only required a minimal procedure to finalize, then there
was a risk that significant incumbency advantage would remain with the distributor or
intermediary. This was particularly considered to be likely where the consumer
believes that consent has been provided at the point of the first discussion. However,
we did not think that permitting a more flexible approach would undermine the
remedy to the extent feared by the FSA. We thought it likely that some firms—and
some consumers—would prefer a shorter discussion of PPI at the credit point of sale
and that we would see a combination of approaches if both options were permitted.
For example, a distributor offering a suite of ‘tailored’ PPI policies could have a
general discussion at the credit point of sale and provide separate quotes covering
each of the relevant combinations. Other distributors offering a similar product range
might provide advice as to which policy would meet the consumer’s demands and
needs and provide a quote only in relation to the recommended variant.
10.97 We also believed that there was a significant difference between a situation in which
a consumer had purchased a policy and one in which a salesman had marketed a
policy to a consumer, but no sale had been concluded. With a more flexible approach
to this element of the remedies package, it would still be open for consumers to shop
around after the credit point of sale for alternative cover and the break in the sales
process would remove the perception that PPI is necessary for the credit. This would
allow more flexibility for distributors and intermediaries as to how they arrange their
sales processes. We noted that, in some situations (for example, where a firm
offered more than one PPI product), a fuller assessment of a consumer’s demands
and needs could enable the distributor or intermediary to provide a personal PPI
quote that was more useful to the consumer.
10.98 We decided, on balance, not to impose additional restrictions on the extent to which
distributors and intermediaries could market PPI at the credit point of sale. However,
we recognized the FSA’s concerns that a consumer’s decision to purchase from a
distributor or intermediary after the credit point of sale should be more than a passive
commitment and that consumers should not feel a sense of obligation subsequently
to take PPI from their credit provider. We consider that other aspects of the design of
this element of the remedies package help to address these concerns. In particular,
the constraints on how consumers can return to purchase PPI within the prohibition
period ensure that the limited exemption for consumer-initiated transactions after
24 hours cannot be used to circumvent this element of the remedies package (see
paragraph 10.124). In addition, the prohibition on contacting the customer in relation
to PPI for seven days and the specification of the point from which the point-of-sale
prohibition commences reduces the scope for consumers to perceive an element of
obligation or conditionality (see paragraphs 10.100 to 10.109). We will continue to
have regard to the issues raised by the FSA as we finalize the detailed implementation of the remedies package. Once the remedies are in place, the selling practices of
distributors and intermediaries, and in particular the risks of passive selling, are an
area that we expect to be kept under particularly close review by both the OFT and
the FSA.
The application of the point-of-sale prohibition
10.99 This element of the remedies package creates a break in time between the credit
sale and the sale of PPI, during which period the distributor or intermediary arranging
the credit (and certain other businesses—see paragraphs 10.148 to 10.150) is
prohibited from selling PPI to the consumer or from contacting the consumer in
relation to PPI. We have considered four aspects of the specification of this measure
that are important to define to give effect to the prohibition:
(a) the point in the credit sales process from which the prohibition period begins;
207
(b) the duration of the prohibition period after the credit sale;
(c) the ability of consumers to purchase PPI on their own initiative before the end of
the prohibition period; and
(d) businesses to which the prohibition applies.
• Starting point for the prohibition period
10.100 To put a point-of-sale prohibition into place, it is necessary to specify when the prohibition should start and end. The prohibition on selling PPI needs to cover both the
credit sale period and a defined time period afterwards. To avoid circumvention, the
prohibition also applies to ‘pre-selling’ of PPI to insure credit agreements discussed
with a consumer by a distributor or intermediary and in which the distributor or intermediary has reasonable grounds to expect will be agreed within the following seven
days.
10.101 We noted that a credit sale is a process involving a number of stages (eg application,
formal offer, acceptance, execution, confirmation, transfer of funds). The end of the
credit sale period, and hence the starting point of the prohibition period after the
credit sale, is easiest to identify and define for those credit sales which are completed
in a single interaction between consumer and provider (eg some branch-based
personal loans or Internet loans where the customer uses an electronic signature).
However, definition of the end of the credit sale period becomes more complex for
credit sales which take place over a number of events at different points in time (eg
some distance sales might have separate interactions for the application, formal
offer, acceptance, execution, confirmation, transfer of funds).
10.102 We received a variety of views on the appropriate definition of the point at which any
prohibition period should start. The OFT suggested that this should generally be
defined as the point at which the credit agreement is executed. The OFT advised us
further that, in the case of Consumer Credit Act 1974 (CCA)-regulated credit
agreements, the date on which the consumer receives a copy of the unexecuted
agreement would generally not be sufficient, as the creditor would not at that stage
be committed to the agreement and could withdraw. The OFT noted that the CCA
(section 57) contains special provisions dealing with withdrawal (by either party) from
a prospective agreement. Furthermore, the risk of withdrawal may increase in the
future under the new CCD obligation to assess creditworthiness as the creditor may
become aware of relevant information after the provision of an unexecuted
agreement and be obliged to modify or withdraw the offer. This implied that, for CCAregulated agreements, the prohibition period should run from when an executed
agreement is received by the consumer, apart from those cases in which an
unexecuted agreement is sent to the consumer and is executed by the consumer
signing it. The OFT considered that it may be reasonable, in those cases, for the
seven-day period to run from the date on which the consumer receives the
unexecuted agreement, because the creditor is bound at that point. The FSA proposed that this be defined as the date on which consumers ‘receive a letter from their
credit provider confirming that their application for credit has been agreed’. The FSA
was concerned that any earlier date could enable firms to reorganize their credit processes to circumvent the prohibition. [] suggested that we consider a later date in
those markets (such as car loans and mortgages) where a significant time gap exists
between the sale of the loan facility and its ‘draw-down’. Others suggested that an
earlier point in the credit sale would be more appropriate—for example, HSBC
suggested an earlier stage in the process, namely the earliest point at which a
consumer asked about the credit product. RBSG suggested that the completion of
208
the formal application for credit and provision of the personal PPI quote would be the
appropriate starting point.
10.103 We took the view that three conditions should be met in order for the prohibition
period to commence:
(a) The consumer should be in absolutely no doubt that they are able to access the
credit irrespective of whether or not they choose to take PPI. This reinforces the
key message that PPI is optional and addresses the barrier to search associated
with the perception that taking PPI would increase consumers’ chances of being
given credit (see Figure 10.3). For this reason, we consider that the completion of
an application or the granting of an ‘offer in principle’ would be too early a point
for the start of the prohibition period.
(b) The start of the prohibition period should correspond to a clearly identified step in
the credit sales process. This would facilitate effective monitoring of whether
businesses were complying with the prohibition.
(c) The consumer should have received the personal PPI quote. This will ensure that
the consumer is able to use the prohibition period to shop around.
10.104 We also consider that the start of the prohibition period should be no later than
necessary. An earlier start to the prohibition period would increase the possibility that
consumers who wanted to could switch both credit and PPI, either during the credit
cooling-off period or potentially before they have accepted a loan offer. An earlier
start would also reduce the possibility that a consumer was left without cover in the
period after the credit has been provided but before they were permitted to obtain
PPI.
10.105 In light of these considerations, we decided that the prohibition period after the credit
sale should start no earlier than the point at which the consumer is provided with
confirmation in a durable medium that the credit provider is bound to provide the
credit or, if this occurs later, when the distributor or intermediary has given the consumer a personal PPI quote.
10.106 In the case of MPPI, this means that the prohibition period would commence when
the consumer receives confirmation of a firm mortgage offer, provided the consumer
has received the personal PPI quote. This starting point precedes the drawing down
of funds, reducing the possibility that consumers who wanted to take out PPI but did
not purchase PPI during the prohibition period would inadvertently be uninsured
when the mortgage commenced.
10.107 For PLPPI, CCPPI and SMPPI, provided the consumer has received a personal
quote, the point at which the consumer is informed that the credit provider is bound to
provide the credit (and hence the starting point for the prohibition) will be either when
the consumer receives a credit agreement which has been signed by the creditor and
only needs the consumer’s signature to become binding, or when the consumer is
notified by the creditor in a durable medium that it has signed the credit agreement.
Where the consumer has not received a personal PPI quote, at the time that he
receives notice that the credit provider is bound to provide the credit, then the prohibition period will start when the consumer receives a personal PPI quote. In these
situations and in the case of agreements concluded by means of distance communication, the distributor may not know the precise timing of receipt of the credit agreement, in which case we would specify in any Order implementing the remedies rules
for deeming when this had arrived.
209
10.108 For retail PPI, we found no reason to think that consumers believed their application
for a credit account was enhanced by taking out retail PPI (see paragraph 6.182). We
decided that for retail PPI, the prohibition period should commence either when the
consumer opens a retail credit account or, if later, the provision of a personal PPI
quote.
10.109 If a distributor or intermediary does not provide a personal PPI quote during the credit
sale (see paragraph 10.88), then the prohibition period starts with the provision of a
personal quote to the customer in a durable medium. This is so that consumers have
the information that they require in order to search the market during the prohibition
period. If this safeguard were not in place, there would be a risk that distributors
and/or intermediaries would not mention PPI at the credit point of sale at all, but
would contact the consumer after the end of the prohibition period to sell PPI, without
having given the consumer the personal PPI quote or any other information about
PPI. This could undermine the effectiveness of the remedies package in facilitating
consumer search. This prohibition would apply only on the first occasion that the
consumer is offered PPI for that credit agreement. Subsequent offers of PPI for that
credit agreement would be treated as stand-alone sales and a personal PPI quote
would have to be provided on request (see paragraph 10.337).
• Duration of the prohibition period
10.110 Three main issues have influenced the length of the prohibition being imposed: the
possibility of a consumer not having PPI cover from the point of the primary credit
purchase; how long it takes consumers to search the market; and the interaction with
other regulatory time periods.
10.111 Nearly all parties highlighted that a point-of-sale prohibition could lead to consumers
being exposed to more risk during the initial period of the credit agreement. 57 Several
pointed to a significant number of claims in the first period of the policy; for example,
7 to 26 per cent of policies that are claimed on are claimed on in the first 90 days. 58
One party (Capital One) said that a shorter ban (eg three days) would be sufficient
time for consumers to search the PPI market and would mitigate some of the
reduced take-up of PPI that they consider is likely (see paragraphs 10.46 to 10.52).
10.112 We sought evidence from the parties on the number of PPI consumers who have
claimable events early in the life of a policy. 59 The evidence, set out in Table 10.1,
suggested that only very few consumers suffer a claimable event in the first 14 days
of a PPI sale, and very few receive the first payment on a claim in the first 14 days.
For example, for most PPI policies sold in 2007, less than one in 1,000 consumers
received the first payment on an insured event within the first 14 days. Given the
small proportion of claimable events that occur within the first 14 days, we did not
collect evidence on shorter periods (eg seven days), though as a point of logic, we
noted that fewer customers still would suffer a claimable event within any shorter
period. We noted that some parties (ABI, Barclays, HBOS) suggested that the negative change in the economic environment that has occurred in 2008 has led to a
higher number of consumers submitting claims. However, we received no evidence
57
Abbey response to Remedies Notice p10, ABI response to Remedies Notice p6, AFB response to Remedies Notice p5,
AIFA/AMI response to Remedies Notice pp3&4, Axa, Barclays response to Remedies Notice p14, Cattles response to
Remedies Notice p16, FLA response to Remedies Notice p21, FISA response to Remedies Notice pp1&2, FSA response to
Remedies Notice p14, Lloyds TSB response to Remedies Notice p15, MBNA response to Remedies Notice p7, Paymentshield
response to Remedies Notice p2, RBSG response to Remedies Notice p11.
58
AFB response to Remedies Notice p5, Barclays response to Remedies Notice p15, Cattles response to Remedies Notice
p21, FLA response to Remedies Notice p21, Genworth response to Remedies Notice p7, Lloyds response to Remedies Notice
p15, RBSG response to Remedies Notice p10.
59
[]
210
to suggest that the number of claims made in the first few days of a policy taken out
at the credit point of sale had risen to a significant number.
TABLE 10.1 Percentage of PPI customers experiencing insured events in either the first 14 days or from 15 to 30 days
of the policy*
per cent
PLPPI
0–14
15–29
days
days
SMPPI
0–14
15–29
days
days
CCPPI
0–14
15–29
days
days
Customers who suffer a claimable event within these time periods
[]
0.06
0.09
0.01
[]
0.05
0.09
0.11
0.12
[]
0.11
0.06
0.16
0.06
0.02
0.01
0.02
MPPI
0–14
15–29
days
days
0.13
0.02
Customers to whom the first payment on a claimable event was made within these time periods
[]
0.00
0.01
[]
0.00
0.00
0.00
0.00
0.00
0.00
0.02
[]
0.18
0.09
0.00
[]
0.04
0.04
0.57
0.00
0.00
0.01
0.17
[]
0.00
0.01
0.00
0.00
0.00
0.00
0.29
[]
0.00
0.00
0.00
[]
[]
0.00
0.00
-
0.09
0.05
0.00
0.00
0.01
0.12
0.00
-
Source: CC, based on information provided by the parties.
*As we are concerned with assessing detriment to the total customer base, we have based the table on the percentage of
claiming customers/total customers who could claim rather than the percentage who have claimed. In most cases, the percentage figures relate to the first payment date. However, for [], [] and [], the figures relate to when the loss event actually
occurred. Numbers are rounded to two decimal places.
10.113 In addition, we consider it is possible that PPI policies sold by the credit provider
could be designed to offer cover backdated to the date of the credit sale. We noted
that in a similar context, when a consumer misses a premium payment some providers ([]) continue providing cover for between 30 and 90 days after the missed
payment and will not prevent a consumer from making a claim. We were also told
that, in some instances, a ‘debt waiver’ or free cover could be provided to a
customer, which indicates that PPI policies could be designed to reduce the risk of
not having PPI during the period of the point-of-sale prohibition. However, two parties
(Barclays and MBNA) told us that policies with backdated cover would be difficult to
design, are not required in other types of insurance and one said it would be difficult
to obtain underwriting for.
10.114 We did not have particularly clear evidence on the amount of time that it would take
consumers to search for PPI in the event of a point-of-sale prohibition, not least
because this situation does not currently exist. We were told that consumers could
search the market in a matter of hours and that consumer interest in PPI tended to
diminish fairly rapidly after the credit sale. Lloyds TSB told us that a consumer who
wished to shop around for PPI would be likely to do so within, at most, seven days of
the credit purchase. Conversely, we were told that stand-alone providers may need
up to 90 days to find and sell to individual consumers. Through other aspects of the
remedies package we are seeking to make it easier for consumers to search the
market and expect increased competition to lead to a higher degree of marketing.
These factors suggested to us that a relatively short prohibition period, of 14 days or
fewer, was likely to be sufficient to enable consumers to search the market and to
have the opportunity to see other marketing from competitors.
10.115 We noted that the cooling-off period for consumer credit products was 14 days. A
prohibition period of fewer than 14 days would enable consumers to switch both
credit and PPI within this period if they found a bundle of credit with PPI that they
preferred. Whilst, as discussed in paragraph 10.67, we did not believe that making
211
use of an extended PPI cooling-off period would be sufficient to address the competition concerns that we have identified, we did not wish to prevent consumers from
making use of the cooling-off period for credit in order to switch provider if, having
had the opportunity to search for PPI policies, the consumer considered that a better
offer on credit and PPI could be obtained from an alternative provider. This possibility
would provide an additional, albeit we expected relatively weak, competitive constraint. We also noted that a shorter period than 14 days could increase the possibility that credit card companies would be able to offer PPI to their consumers at the
point of activation.
10.116 Taking all of the above factors into consideration, we concluded that a prohibition
period of seven days would give consumers adequate time to search the market
making use of the information in the personal quote, to have the opportunity to see
advertisements from other providers and to search for better-value PPI or credit with
PPI (and in such cases make use of the credit cooling-off period to switch to an
alternative provider of credit and PPI) before any further contact from the distributor
or intermediary who arranged the credit. Given the limited number of claims in the
early stages of PPI policies, we do not consider that a prohibition period of this length
will materially increase the risk faced by consumers of being without cover before
they are contacted again by the distributor or intermediary.
• The ability of consumers to return on their own initiative after 24 hours to purchase
PPI before the end of the prohibition period
10.117 We note that a point-of-sale prohibition in effect imposes a restriction on consumers
as well as on the businesses covered by the prohibition. While we consider that such
a restriction is necessary given the nature and severity of the competition problems
we have found, we wish only to impose the minimum restrictions on consumers
necessary to achieve an effective remedy. We concluded that consumers should be
able to return on their own initiative to purchase PPI before the end of the seven-day
prohibition period and consulted in our provisional decisions on remedies on what
constraints should be placed on such sales.
10.118 In general, parties suggested that the restriction on the time before consumers could
return on their own initiative be kept to a minimum (for example, Capital One proposed 2 hours for sales made in branches) to minimize inconvenience to consumers
and reduce the potential negative impact on take-up rates. The FSA also noted that
an unduly long gap could have a negative effect on consumer take-up rates, and
supported the 24-hour prohibition as not hindering consumers who wanted to purchase PPI from their credit provider by creating an unduly long gap. Some parties
were concerned about the practicalities of specifying this period in hours, and proposed ‘the next day’ instead.
10.119 We considered these submissions but concluded that a clean break after the point of
sale was necessary to ensure the effectiveness of this remedy—without a break of
some description it would not be possible to enforce the remedy. We concluded that
any period of shorter than 24 hours would be difficult to monitor, would not constitute
a clean break and was potentially open to circumvention. We noted that some providers could have difficulty in incorporating a 24-hour waiting period into their
systems. Where this was the case, distributors could meet their obligations by ensuring that there was at least 24 hours between the credit sale and the purchase.
10.120 One party (Nationwide) also queried why, in the provisional decision on remedies, we
proposed that consumers would only be able to revert via telephone or Internet, and
several distributors expressed the view that it should be made as easy as possible
for consumers to take up PPI in the manner that they most preferred. For example,
212
RBSG told us that ‘a consumer should be able to respond after 24 hours from the
quote by (i) sending a text message; (ii) sending an email; or (iii) posting an acceptance’ and that consumers should be provided at the point of sale with a PPI offer
which simply required a ‘yes’ answer to conclude the contract. We note that some
parties (Capital One, HSBC) suggested that restricting the mode of communication to
purchase PPI products made it too difficult for the consumer to purchase PPI. In
these circumstances, it was suggested to us that this might have an unnecessary
negative effect on consumer purchasing behaviour. One party (HSBC) requested the
CC to clarify that once the customer has demonstrated their desire to purchase PPI,
in the event that the provider does not wish to sell, or the customer does not wish to
purchase, the product through the telephone or Internet channels, the customer
could then be invited to an appointment in the branch even within the prohibition
period after the sale of the credit product.
10.121 We recognized that there was a risk that the businesses covered by the prohibition
could seek to use this limited exception to the seven-day prohibition to circumvent
the remedy. We also noted the concerns that the FSA had expressed that the
effectiveness of the prohibition could be harmed by ‘inertia selling’. For example, a
salesperson could present the discussion of PPI at the credit point of sale as an
agreement in principle by the customer to take out PPI, merely requiring confirmation
from the customer, for example by signing a form or ticking a box (see paragraphs
10.92 and 10.93). We therefore decided that restrictions were needed on the process
through which consumers could take out PPI from businesses covered by the sevenday prohibition during the prohibition period, to ensure that the following two conditions were met:
(a) First, businesses covered by the prohibition must be able to demonstrate to monitoring agencies that the customer had initiated the contact, that they had done so
at least 24 hours after the conclusion of the credit sale and that the business had
reasonably satisfied itself that the consumer had seen the personal quote.
(b) Second, businesses covered by the prohibition must be able to establish and to
demonstrate to monitoring agencies that the consumer’s decision to purchase
PPI from the provider was more than a passive commitment or confirmation of a
pre-agreed sale.
10.122 We did not think that these conditions could be met through face-to-face and postal
channels. We noted that the FSA has undertaken several thematic pieces of work
focused on PPI selling standards and has published three reports between
November 2005 and September 2007 which identified persistent and serious failures
in the sales processes of a number of firms that related to face-to-face branch
sales. 60 In our view, it would be more difficult to establish and monitor, in the context
of a branch sale, whether a purchase had been initiated by the consumer. For
example, a distributor or intermediary could set up a face-to-face appointment to
discuss a consumer’s general financial requirements during the prohibition period,
during which a salesperson might encourage the consumer to raise the subject of
PPI, in order to complete a sale. We therefore had serious concerns regarding
compliance with this element of our remedies package, if sales could be concluded
60
See the results of the FSA’s third thematic review and mystery shopping exercises published in September 2007 at
www.fsa.gov.uk/pages/Library/Communication/PR/2007/102.shtml. The FSA visited 150 firms which sell PPI. It also commissioned a market research company to mystery shop 114 firms which sold PLPPI on a branch-based face-to-face basis. The
FSA said in relation to the firms it visited that over a third of them had inadequate systems and controls in place around the
sales process, which indicated that the firm and senior management had not given appropriate priority to compliance with regulatory obligations that were designed to provide positive outcomes for consumers. The FSA also said that the mystery shopping
results highlighted standards of behaviour which fell below its expectations and raised questions about firms’ systems and
controls. In addition, it said that this might be indicative of failures in the quality of staff training and competence.
213
through the face-to-face channel after 24 hours. We considered that the postal
channel could too easily be manipulated, for example by completing the paperwork
with the customer in branch and simply advising the consumer to delay posting. This
would be very difficult to detect through an ex-post review. In addition, the only way
of verifying the date of posting would be if providers were required to retain postmarked envelopes.
10.123 The responses that we received from parties to our provisional decision on remedies
(see paragraph 10.120) indicated that they were already giving thought to ways in
which this limited exemption could be used to significantly reduce the impact of the
seven-day prohibition. This reinforced our concerns that there was a significant risk
of circumvention unless this remedy was clearly controlled by way of limiting the
channels by which customers could contact firms to purchase PPI after 24 hours. We
recognize that this imposes an additional constraint on consumers and businesses,
but consider that it is necessary to ensure the effectiveness of the prohibition. We
considered that the alternatives to this limitation would be either: not to allow this
limited exemption at all; or significantly to restrict the ability of distributors and intermediaries to market PPI at the credit point of sale, in order to ensure that firms were
not able effectively to make pre-agreed sales of PPI at the point of sale of credit. We
considered that either of these alternatives would be more intrusive. Not providing
this limited exclusion at all would impose a greater restriction on consumer choice.
We did not want to restrict distributors’ and intermediaries’ marketing activities any
more than necessary, as we consider that marketing of PPI at the credit point of sale
is one way in which search and competition could be stimulated (see paragraphs
10.89 to 10.98).
10.124 We concluded that there were substantial risks of circumvention or frustration of the
remedy, should this exemption apply to face-to-face and postal channels. We therefore decided that the exemption for customer-initiated transactions should not apply
to face-to-face sales or to situations in which the customer reverted by post. We took
the view that the conditions set out in paragraph 10.121 were capable of being met,
and that circumvention risks could be more effectively managed and controlled, in a
telephone or Internet environment. For these reasons, we concluded that the 24-hour
exemption for consumer-initiated transactions should be restricted to telephone and
online channels and that businesses covered by the prohibition should be able to
verify to monitoring agencies that any sales within the prohibition period were
genuine consumer-initiated sales.
10.125 Some distributors (Capital One, MBNA 61 ) were concerned about the practicalities of
confirming that the consumer had seen the specific personal PPI quote. In particular,
we were told that consumers were unlikely to remember the quote and having to
resend a quote would cause complexity and turn consumers away from the process
of purchasing. We consider that requiring distributors to make this check (for
example, by asking as part of the sales script) was a reasonable safeguard in a
situation where consumers were contacting a distributor or intermediary at their own
initiative. We consider that this requirement, which also applies when parties covered
by the prohibition contact the customer at the end of the prohibition period, provides
incentives on distributors and intermediaries to ensure that the personal PPI price
quote was given sufficient prominence in the material provided to consumers, and
was provided quickly.
10.126 We decided that there should be a limited exemption to the seven-day prohibition
period for consumer-initiated transactions from 24 hours after the conclusion of the
61
MBNA response to the provisional decision on remedies, p11 paragraph 40.
214
credit sale (or, if later, after they have received the personal PPI quote) to buy PPI.
We decided that this exemption to the point-of-sale prohibition should only apply to
contacts initiated by consumers over the telephone or the Internet, but not through
other channels, such as visiting a branch or by post. We further decided that any
party covered by the prohibition which is selling the policy covered by the quote must
obtain confirmation that the customer has seen the personal PPI quote, before any
PPI sale can be made. We decided that any party covered by the prohibition must
also be able to demonstrate that the customer had initiated the contact, that they had
done so 24 hours after the conclusion of the credit sale and that any sales made
within this exception period were genuine consumer-initiated sales.
• Businesses to which the prohibition applies
10.127 We decided that the prohibition should apply to the distributor or the intermediary
who offered or arranged credit for the consumer; any business mentioned during the
credit sale with which the distributor (or intermediary) has a commercial referral
arrangement in relation to PPI; and any business to which information regarding the
consumer which was obtained in the credit sale has been passed for the purpose of
allowing the recipient to sell PPI to the customer.
10.128 We consider that the wide ambit of the prohibition is important to prevent circumvention of the remedy. For example, it might otherwise be possible for a distributor or
intermediary to pass on details of a recent credit consumer to an affiliated company
or to a company providing them with recompense of some type for referrals, which
could then market its PPI product during the prohibition period.
10.129 Cardif Pinnacle told us that, in its view, the point-of-sale prohibition should not apply
to companies recommended by the credit provider or to which information has been
passed. It was concerned about the enforceability of the restriction on third parties
who might not know that they had been recommended in this way and the possibility
that PPI competitors could be restricted from marketing to, or contracting with, consumers by a distributor mentioning them at the point of sale and invoking the point-ofsale prohibition. We considered that businesses would normally be aware of when
they had a commercial referral arrangement in relation to PPI with a distributor or
intermediary and the circumstances in which they were likely to be mentioned in
sales scripts. We decided that this safeguard was necessary for the effectiveness of
this element of the remedies package.
10.130 RBSG said that it would be very difficult to put in place fail-safe methods of checking
whether a consumer who asked for PPI had been sold a loan by one of the companies in its trading group in the last seven days. [] It is important that this remedy
is not unreasonably burdensome on the industry but it must also be effective. We
consider that it is not unreasonable for providers of PPI to demonstrate that they
have discharged an onus of checking the eligibility of a consumer in relation to these
requirements at the same time as requesting information on other aspects of
consumer eligibility.
Exemptions or amendments to the prohibition for different types of PPI, sales
channel or transaction
10.131 We received a number of submissions that suggested that the point-of-sale prohibition was not applicable to, or should be adapted for, different circumstances within
the PPI market. In paragraphs 10.132 to 10.135, we review the arguments that have
been put to us for exemption or amendment of the point-of-sale prohibition for MPPI
and CCPPI. In paragraphs 10.136 to 10.155, we consider the point-of-sale prohibition
215
in the context of Internet sales, retail credit providers, credit unions and PPI sales by
intermediaries.
• Mortgages
10.132 Several parties (HBOS, AMI, Abbey, Paymentshield, CML, Legal & General) told us
that the prohibition should not apply to MPPI. HBOS told us that this sector was ‘an
effectively functioning market, with broad distribution, including stand-alone intermediaries, and little cross-subsidisation between MPPI and the core mortgage’. Abbey
noted that in December 2008 the Government proposed a scheme whereby families
could, in certain circumstances, defer all or part of their mortgage interest payments
for up to two years. It said that the perception of customers might be that it is not
worth having MPPI because of this, which could have a significant detrimental effect
on the MPPI market. Legal & General suggested that the prohibition should not apply
to MPPI because the MPPI sales process was more involved than other types of PPI,
the sale occurred over a number of meetings with the consumer and it was already
subject to two separate regulatory regimes.
10.133 We do not agree that the MPPI sector is functioning effectively for the reasons set
out in Section 5. Nor did we consider that the Government’s recent proposal would
address the AEC that we found in relation to MPPI. We therefore did not agree that
this sector should be exempted from the point-of-sale prohibition.
• Credit cards
10.134 [] told us that the impact of this element of the remedies package would result in
[] ceasing to offer PPI. However, it suggested that with modification the effect of
the remedy could still be fully achieved without CCPPI providers exiting the market.
Capital One proposed reducing the prohibition period to three days and simplifying
the PPI quote to combine it with the policy summary and distribute it at the point of
application with PPI marketing materials. It was suggested that these amendments
would enable CCPPI providers to continue to sell PPI at the point of credit activation,
while ensuring that consumers had the materials they needed to search the market. 62
10.135 In paragraphs 5.56, 5.57, 5.76 and 5.77, we found that barriers to switching and
searching exist in relation to CCPPI. In addition, as we discussed in paragraph 10.57,
we consider that incentives continue to exist for distributors to continue to provide
PPI to their credit consumers. We decided that the point-of-sale prohibition should
not be modified as Capital One suggested, though the duration of the point-of-sale
prohibition was reduced compared with the length of prohibition on which Capital
One commented, and we have accepted some of Capital One’s suggestions in
relation to the personal PPI quote (see paragraphs 10.178 to 10.181).
• Internet sales
10.136 Several parties (HBOS, Lloyds TSB, Egg, Express Gifts) argued that, even if a pointof-sale prohibition was appropriate for other distribution channels, it would not be
relevant to credit and PPI sales conducted over the Internet. It was put to us that
there are a number of key differences for an Internet sale:
62
We note that we received mixed evidence on the when credit card activations usually occur. [] told us that most credit card
providers sold PPI during the card activation process and many credit card activations occurred between five and eight days of
the credit sale, although [] told us that most credit card applications occurred after 14 days.
216
(a) Consumers are more in control of the online shopping process because a typical
Internet consumer completes their purchase at home, in their own time and with
no external pressure. 63
(b) There is no risk of consumers perceiving that PPI is necessary in order to obtain
a loan as there is no human contact and the Internet process is fully ‘scripted’. 64
(c) Consumers are in a better position to search for credit and/or PPI online. We
were told that Internet purchases of credit and/or PPI were most likely to be a
consequence of ‘shopping around’ for the best product, using price-comparison
websites (see Appendix 3.8), live comparison, and/or other recommendations
and reviews. 65
(d) A large proportion of consumers opt out of all marketing communication, which
prevents distributors from offering PPI after the prohibition period. 66
(e) If a consumer who has shopped via the Internet opts to take up a credit agreement and PPI, both contract documents are sent out to the consumer by mail for
signature. This process provides time for the consumer to decide whether or not
to take out that particular offer of credit and/or PPI. 67
(f) Consumers who are able to use electronic signatures for loan purposes value the
convenience associated with being able to conclude a transaction in a single
online application. We were told that ‘a significant element of this benefit would
be lost if they were unable to buy PPI instantly (and were, instead, obliged to wait
for a personalized quote to be sent by email or post, and then a further 24 hours
before they could take up the offer it contained’. 68
10.137 In addition, Lloyds TSB noted that for Internet and other remote transactions which
required a signature to conclude the credit agreement, distributors might wish to contact the consumer within the PPI prohibition period to encourage them to sign and
return the credit agreement. Lloyds TSB told us that it would be difficult to follow up
these applications within 14 days without also discussing PPI (particularly if the consumer wanted to discuss PPI). If the credit sale period is ongoing then, as discussed
in paragraphs 10.48 to 10.53, we do not consider that further discussions without
concluding a PPI sale would be prevented by this prohibition. If the contact is during
the prohibition period after the credit sale, then we believe that parties are able to
instruct their representatives not to discuss PPI with consumers.
10.138 We noted all of these points and also noted that there are facilities for searching for
PPI prices online (see paragraph 5.31), although these were currently limited. One
effect of our package of remedies is that the barriers to searching are likely to reduce
as advertising and the coverage of the FSA comparative tables increases.
10.139 In light of these arguments, we gave consideration to whether the point-of-sale prohibition might continue to be effective should it be adjusted or exempted from applying in relation to Internet sales. We first considered whether Internet sales could be
exempted from this remedy. We note that Internet consumers have the opportunity to
63
Egg Banking response to provisional decision on remedies p2. Lloyds TSB response to provisional decision on remedies p20, paragraph 1.25. Egg Banking response to provisional decision on remedies p2(b). 66
Egg Banking response to provisional decision on remedies p2(d). 67
Lloyds TSB told us that this also applied to telephone sales. It told us that ‘These channels simply do not create the risk of ‘impulse buying’. For example, during 2008, [50–75] per cent of consumers who had successful telephone applications for personal loans from Lloyds TSB completed their loan documentation and drew down the loan. 68
Lloyds TSB response to provisional decision on remedies p31, paragraph 2.18. 64
65
217
search for PPI providers prior to entering the sales process with a particular PPI
provider. However, this is also true, in principle, of any consumer who enters into an
agreement for credit or PPI. The more relevant issues are the extent of search and
the barriers to search. Our analysis of data obtained from Internet comparison websites led us to conclude that some consumers do use Internet comparison websites
to search for combinations of PLPPI with credit; however, the number of people
doing this is small in absolute terms and only a small fraction of all those who search
for credit (see paragraph 3.117). Overall, we found little evidence that competition on
PPI was more intense for Internet sales than through other channels. The point-ofsale prohibition gives the Internet consumer the opportunity to continue searching
after the credit sale but prior to committing to a particular PPI provider. In this context, we did not see a clear distinction, either in terms of the need for the prohibition
or its effectiveness in addressing the AEC, between PPI sold via the Internet and PPI
sold via other distribution channels. There is also a risk that an exemption for this
particular element of the remedies package may make it possible for distributors to
circumvent the entire remedies package, for example by concluding a PPI sale in a
branch with the Internet available or simply requesting that a consumer conclude a
telephone sale by visiting the PPI provider’s website. To exempt from the point-ofsale prohibition PPI sold via the Internet would reduce the effectiveness of the package of remedies for Internet consumers and would create circumvention risks.
10.140 We also considered whether it might be possible to exempt Internet PPI sales from
the point-of-sale prohibition and maintain the effectiveness of the remedies package
overall by requiring certain other procedures for these types of PPI sales. For
example, we considered whether it would be effective to require that a consumer
reviews an Internet page showing key PPI messages, only permitting a consumer to
make a PPI purchase at a point after the purchase of credit has occurred and
specifically prohibiting the use of Internet sales of PPI within branches. We decided
that these additional procedures would not be as effective as the point-of-sale prohibition at reducing the barrier to consumer searching and, in the case of the exclusion
of Internet-based sales of PPI in branches, would be very difficult to monitor.
10.141 We also considered whether the point-of-sale prohibition for Internet sales might be
reduced by beginning at a point in time that is earlier than that required for other PPI
sales channels. However, as discussed in paragraphs 10.105 to 10.108, we consider
that the general principle of requiring the consumer to have confirmation in a durable
medium that the credit provider is bound to provide the credit, or, if this occurs later,
a personal PPI quote, should continue to apply in an Internet sales context. We note
that it could take a shorter time for the consumer to receive this document than
through some other sales channels because receipt of the relevant document is more
likely to be via, for example, an emailed PDF than a hard copy document that has
been posted.
10.142 We concluded that it was important that the point-of-sale prohibition should apply on
a consistent basis across all distribution channels, both to ensure the effectiveness of
the remedies package in relation to Internet sales and to avoid possible circumvention risk to the remedies package via sales channels other than the Internet. We
therefore decided that Internet sales should not be exempted nor subject to a different remedy than that required of other sales channels.
• Retail credit
10.143 We noted a number of characteristics of the retail PPI market that some parties
suggested meant that an alternative specification of this element of the remedies
could be appropriate for this sector. We found that the sale of retail PPI at the initial
point of sale and continued exclusive access to customer accounts restricts the
218
extent to which other PPI providers can compete effectively (see Section 6). However, we found the point-of-sale advantage in retail credit to be weaker than for other
forms of PPI, partly because consumers were unlikely to believe that their ability to
purchase goods was dependent on taking retail PPI.
10.144 The average amount outstanding and consequently the premiums on retail PPI were
often quite small compared with other forms of PPI. A number of parties (the BRC,
Express Gifts, JD Williams) told us that average balances insured by retail PPI
ranged between £100 to £400 and average premiums would generally be less than a
few pounds per month. We were told that these typical balances meant that the
profitability in this PPI sector was much lower than that of other PPI sectors. In
addition, the average time taken to settle outstanding balances was typically between
six to nine months. The BRC told us that the AEC for retail PPI has been identified as
less pronounced than that found elsewhere and expressed concern that the proposed remedy was equal to that proposed for mainstream PPI providers. For these
reasons, it was suggested by these parties that this remedy would not be proportionate to the AEC in the retail PPI sector.
10.145 We were also told by a number of parties (the BRC, Express Gifts, JD Williams) that
retail PPI catered for a low socio-economic group of consumers that would find it
difficult to obtain PPI from other PPI providers. One party (Express Gifts) told us that
it was not uncommon for a consumer to maintain an account for over 20 years and
simultaneous accounts with other retail companies. Two parties (SDGFS, Express
Gifts) told us that PPI specifically and finance generally was only a secondary reason
for consumers to choose to shop at a particular retailer. We were told that these
reasons supported a view that this remedy should be modified for retail PPI providers.
10.146 We also noted that JD Williams did not sell PPI at the initial point of credit sale. It
instead preferred to focus its marketing of PPI on those consumers who it considered
would run a credit balance. It was only able to identify these consumers at a point
three months after the initial point of credit sale. In this situation, if a personal PPI
quote were not provided at the initial credit point of sale, then the prohibition period
would start from the subsequent provision to the consumer of a personal PPI quote
ahead of any later marketing activity.
10.147 We decided that the point-of-sale prohibition should apply to retail PPI. Our finding
was that a point-of-sale advantage did exist, albeit to a lesser extent than for PPI sold
in other sectors of the market, and we consider that it is necessary to address the
AEC found in retail PPI with this element of the remedies package unmodified. However, we decided that the prohibition period should start when the consumer first
opens a credit account or the later provision of a personal PPI quote (see paragraph
10.108), and that the specification of the personal PPI quote should be simplified for
retail credit (see paragraph 10.181). We consider that the application of the prohibition in this way acknowledges the differences raised by the parties in respect of retail
PPI and will reduce the burden, but not the effectiveness, of this element of the
remedies package in this sector of the PPI market. In this way, our remedies package addresses the point-of-sale advantage that we identified as arising at the initial
point of sale, as well as the advantage arising subsequently through continued
exclusive access to customer accounts. The point-of-sale prohibition applies only on
the first occasion that the consumer is offered PPI for that credit agreement (we did
not consider that it was necessary to apply the prohibition to every subsequent
communication concerning PPI from the credit provider to the consumer in order to
address the AEC that we have found).
219
• Credit unions and other smaller providers
10.148 CUNA Mutual (CUNA), a supplier of PPI to credit unions, asked for an exemption
from the point-of-sale prohibition for credit unions possibly by treating PPI sales to
their members as having been sold on a stand-alone basis by their insurance
partner. CUNA noted the different role and incentives of credit unions and other thirdsector lenders. It argued that the proposed remedy prohibiting the sale of PPI at the
same time as the credit sale would have significant negative implications for the
credit union sector.
10.149 CUNA noted, in particular, that many credit unions would be at a disadvantage in
offering borrowers the option to apply for PPI after 24 hours via the Internet or telephone as many did not have these channels available to them, and that the limited
resources available to many credit unions meant that they would not be in a position
proactively to resolicit consumers after the waiting period. In addition, it said that its
members may not have ready access to other payment transmission methods
required by stand-alone providers.
10.150 We considered whether a de minimis threshold could be applied to the point-of-sale
prohibition to mitigate the burden that might be experienced by some smaller PPI
providers, such as credit unions. We consider that the same point-of-sale advantage
exists notwithstanding the size of the distributor that is offering the PPI product. We
found no obvious break in the market between ‘small’ PPI providers and ‘larger’ PPI
providers. We also noted that a large part of the MPPI market is serviced by smaller
PPI providers, such as IFAs, and noted that an exemption for small distributors
and/or intermediaries could significantly undermine the effectiveness of our remedy
package. We noted, but were not persuaded by, the arguments put to us by one
provider that a specific exemption was needed for credit unions. On this basis, we
decided that there was no justified reason to preclude certain consumers from the
opportunity to shop around for the best-value PPI after the credit point of sale. We
decided that an exemption should not be developed to exclude smaller providers,
including credit unions, from the point-of-sale prohibition.
• Intermediaries that arrange credit
10.151 We received a range of views from the parties on whether the point-of-sale prohibition should apply to intermediaries. Some parties (Aviva, Capital One, RBSG),
including the FSA and the OFT, suggested that this remedy should apply equally to
all organizations that offered credit and PPI together. Other parties (Genworth,
HBOS, Lloyds TSB, FLA, AFB, AMI, Assurant, GMAC, Renrod, SBPP) suggested
that this remedy should not apply to intermediaries. Reasons cited included that
these companies did not have continuing relationships with consumers, that they
were assisting the consumer to search the market at the point of sale and in fact did
have continuing relationships with their clients, and that many providers would be
likely to leave the market. Three parties (HBOS, Lloyds TSB, FLA) told us that the
application of the point-of-sale prohibition to intermediaries would cause intermediaries to leave the market. GMAC told us that a point-of-sale prohibition would prevent third parties, such as car dealers, from offering PPI because it would require two
separate contacts with a consumer.
10.152 Cardif Pinnacle emphasized to us the practical difficulties of a point-of-sale prohibition in the context of intermediaries. In particular, it highlighted issues relating to the
initiation of the prohibition period where there was a delay between the sale of credit
and completion of the loan, the complexity of selling arrangements between distributors and recommended PPI providers, and the way in which the point-of-sale
220
prohibition would apply to existing consumers who wished to extend their credit
facilities (and PPI cover).
10.153 SBPP suggested that advised sales should be excluded from this remedy, because
in this situation an intermediary was searching the PPI market on behalf of a consumer and making a recommendation on an appropriate policy for their circumstances. It told us that this searching of the market should be encouraged from a
competition perspective because those PPI providers providing the best price and
quality of cover would be selected, resulting in consumer benefits. However, our
analysis in paragraphs 3.64 to 3.74 and Appendix 3.5 indicates that only a very small
percentage of PPI intermediaries search the market in this way. In addition, the FSA
told us that an advised sale did not indicate necessarily that a ‘meaningful whole-ofmarket’ search had occurred because there was only limited stand-alone provision of
PPI in the market. We considered that it would not be appropriate to exempt those
intermediaries that provide advised sales.
10.154 We noted the various views of the parties and the significant role of intermediaries in
the number of PPI policies distributed in the market, particularly in relation to MPPI.
The point-of-sale prohibition is linked to the provision of credit, and we found that the
point-of-sale advantage applied to those intermediaries involved in arranging the
credit sale, but not to other intermediaries. On this basis, we decided that the
prohibition should apply only to those intermediaries involved in arranging the credit
sale and who fall into one of the categories in paragraph 10.127. We also decided
that, for the reasons set out in paragraph 2.44, the point-of-sale prohibition should
not apply to independent price-comparison websites such as www.moneysuper
market.com which consumers use to search the market.
10.155 We decided that an exemption from the point-of-sale prohibition for intermediaries
involved in arranging the credit sale would not be appropriate for similar reasons to
those provided in paragraph 10.150. We acknowledge that this may cause some
intermediaries to exit the market or to enter into different arrangements with PPI
providers. However, we concluded that it is important to apply this element of the
remedies package in an unmodified manner to address the AEC that we have found.
• Particular types of transaction
10.156 Barclays told us that the point-of-sale prohibition should not apply to agreements that
increase a lending amount and associated PPI. We considered that if this constituted
a new credit agreement, there was no valid reason why such agreements should be
exempt from the prohibition. As a general principle, we decided that no particular
type of credit transaction should be exempted from the point-of-sale prohibition, as
we expect consumers to benefit from increased opportunity to search the market. In
addition, we considered that allowing such exemptions could create a circumvention
risk. However, we acknowledge that there may be some exceptional cases in which it
would be impracticable to apply the point-of-sale prohibition, which may become
apparent during the development of the Order.
Obligation to provide a personal PPI quote
Summary of this element of the remedies package
10.157 We found that a lack of comparable information hinders consumers from searching
for alternative PPI policies around the time of purchasing PPI. We consider that this
element of the remedies package will make it easier for consumers readily to compare PPI policies prior to making their decision to purchase PPI.
221
10.158 In order to contribute to effectively addressing the AEC, we decided to require that a
personal PPI quote is provided to consumers in a durable medium if a consumer
receives information about PPI from a distributor or intermediary during the credit
sale period. If information about PPI is not offered during the credit sale period, a
personal PPI quote must be given if the consumer is subsequently contracted and
offered PPI. The personal PPI quote will contain information about the consumer, the
credit product that is being insured (where relevant) and the PPI policy, and will be
substantially in the format set out in Appendix 10.1 subject to consumer testing to be
carried out by the CC prior to the adoption of a final format into any order.
FIGURE 10.2
Personal PPI quote
All distributors and intermediaries who arrange or sell credit for consumers must provide a personal
PPI quote to the consumer in a durable medium (if the distributor or intermediary provides information
about PPI to the consumer during the credit sale period). If the distributor or intermediary arranging
the credit does not provide a personal PPI quote during the credit sale period, but subsequently contacts the consumer to offer PPI, a personal PPI quote must be provided at that time.
The price and terms offered in personal PPI quotes given at or after the point of sale must remain
valid for at least 14 days after the start of the point-of-sale prohibition period.
Stand-alone providers and providers of short-term IP are required to provide a personal PPI quote to
the consumer in a durable medium if the consumer asks the provider about the cost and/or features of
a stand-alone PPI and/or short-term IP policy sold by that provider.
How the obligation to provide a personal PPI quote addresses the AEC
10.159 The requirement to provide a personal PPI quote will contribute towards addressing
the barrier to search arising from the time taken to obtain accurate price information
that we found contributed to the AEC. It complements and supports the point-of-sale
prohibition as consumers can make use of the quote during the period of the prohibition to shop around for the best-value PPI policy (and it is for this reason that the
point-of-sale prohibition does not commence until the personal PPI quote has been
given if this is after the conclusion of the credit sale period). It complements and
supports other measures in our remedies package designed to address barriers to
search, in particular, the provision of information in marketing material, the provision
of information to third parties and unbundling retail PPI from merchandise cover. It
also complements and supports the measures to facilitate switching—the singlepremium prohibition and the requirement to provide an annual statement—as those
consumers who are prompted by these measures to switch will more easily be able
to compare their current product with those on offer from alternative providers.
The content of the personal PPI quote
10.160 We received comments on the following issues relating to the detail of the personal
PPI quote:
(a) the degree to which the format and content of the personal PPI quote was
standardized;
(b) the period over which the price on the personal PPI quote remains valid;
(c) the inclusion of personal information in the personal PPI quote;
(d) the inclusion of information about the credit product in the personal PPI quote;
222
(e) the requirement to calculate a ‘combined APR’ where the personal PPI quote
relates to a specific credit product; and
(f) its use with PPI for revolving credit (CCPPI and retail PPI).
Degree of standardization
10.161 We considered the degree to which the personal PPI quote should be laid out in a
standard format determined by the CC. The purpose of requiring a personal PPI
quote is to give consumers a tool with which they can shop around for PPI after the
credit point of sale and, if considering switching, for comparison with the pricing
details contained in the annual statement. We noted that the Mortgage Conduct of
Business review by BMRB for the FSA said that the different formats for the disclosure documents, such as the key facts illustration, can make it more difficult for consumers to undertake comparisons. 69 We therefore took the view that there was a
strong case in principle for requiring consumers to be presented with the same information in materially the same format, as this would increase the ability of consumers
to compare products.
10.162 However, we were told that there could be costs and risks associated with an
excessively prescriptive format. Barclays told us that allowing flexibility in terms of
formatting would enable providers to develop personal PPI quotes that were ‘user
friendly, cost-effective and in line with their brand image’. We were also told that an
overly prescriptive format may not be able to reflect the characteristics of all the
products on the market and could restrict innovation into the future. For example,
Barclays said that the personal PPI quote on which we consulted in our provisional
decision on remedies did not provide sufficient information for situations in which the
PPI covered more than one product, the credit product was adjusted later but PPI
cover was kept the same, a joint life policy was taken out or where a lump sum was
paid for life cover.
10.163 We noted that there was a balance to be struck between ease of comparability and
flexibility. We concluded that the personal PPI quote needed to contain prescribed
information and that rules would be needed on format and prominence, but that some
flexibility needed to be built in to accommodate differences in products and potential
for future innovation. We also concluded that further consumer testing would be
needed to specify the rules on format and prominence. We therefore concluded that
the personal PPI quote should take substantially the forms proposed in Appendix
10.1 but that consumer testing by the CC should be taken into account, prior to the
adoption of a final format into any order.
Period over which quote is valid
10.164 We considered whether there should be a minimum period over which any price
offered in a personal PPI quote should remain valid.
10.165 Lloyds TSB submitted that individual lenders should be free to decide how long
personalized PPI quotes were valid. The OFT suggested that we should specify a
minimum period of validity covering at least to the end of the prohibition period.
Otherwise, this could deter shopping around.
69
Consumers and mortgage disclosure documentation, September 2006, FSA, p9.
223
10.166 Some limitations were also noted on the extent to which a quote could be ‘firm’.
Lloyds TSB said that a quote could only remain valid subject to confirmation from a
consumer that there had not been any change their circumstances between the date
of the quote and the date on which the PPI policy was purchased. RBSG said that a
revised quote requested by a consumer after the prohibition period had elapsed
(perhaps to protect a different amount of credit) should not be subject to the point-ofsale prohibition. The FSA noted that it was important that, where a personal quote
had been given following a PPI discussion where only information was given and no
advice, it should be clear to the consumer that the quote did not represent a personal
recommendation.
10.167 We noted these points and we concluded that a minimum period needed to be specified for which the quote would remain valid. If not, there was a risk that distributors
would seek to use the quote period to pressurize consumers into returning to their
distributor, for example by offering an exclusive price that would only be available if
the customer returned during the prohibition period. To avoid this, we concluded that
the quote should be valid for at least 14 days after the start of the point-of-sale
prohibition. This would make it clear that, for all customers, the quote would remain
valid for at least a week after the end of the prohibition period.
Inclusion of personal information in quote
10.168 Capital One queried whether there was a need to include personal information about
the consumer (eg name and address) on the personal PPI quote for CCPPI because
this would add to the operational cost and complexity of the quote. It also strongly
recommended that the provision of date of birth should be removed as this could
create an identity theft risk in certain distribution channels. We noted this risk and will
evaluate how much personal information is needed to be presented on the personal
PPI quote during consumer testing.
Inclusion of credit information in the quote
10.169 Lloyds TSB told us that for a stand-alone provider to provide a rival quote to a distributor’s PPI quote, a stand-alone provider would need to know the start date of the
loan, the term of the loan, the total loan amount and the amount of the monthly
repayment for which the consumer required cover. However, Capital One suggested
that including this information would duplicate information provided to a consumer
elsewhere and might indicate to consumers that the credit and PPI were linked; the
credit information would not help consumers to make a decision about PPI, inclusion
of credit information would cause a competitive disadvantage for combined providers,
and operational complexity because it would require PPI systems to link with credit
systems.
10.170 We decided that information about the start date and term of the loan, the loan
amount and the monthly repayments are essential for any competing provider to give
a quote for PPI on a stand-alone basis and should be included on the personal PPI
quote.
10.171 The OFT told us that splitting the monthly repayment stated on the personal PPI
quote into capital and interest components could be confusing and could add significantly to the length of the quote. The OFT told us that it should be sufficient to
indicate the amount of credit, the Total Charge for Credit and the Total Amount
Payable, in addition to the amount of the monthly repayment, for personal loans,
mortgages and second-charge mortgages. Lloyds TSB suggested that, for personal
loans, the cost of credit and the cost of PPI could each be presented on a monthly
224
basis. We consider that the Total Charge for Credit could be compared with the
measure of total premium cost for PPI, which businesses are required to disclose
under ICOBS for those fixed-term products, which are not annually renewable. We
noted that our consumer research had indicated that consumers were generally
happier with monthly measures (see paragraph 10.208), but, as noted above, there
are difficulties associated with a monthly capital/interest split. We also noted that the
Total Amount Payable on the loan could be a useful indicator for stand-alone
providers. We decided that the Total Charge for Credit and Total Amount Payable
should be included in the personal PPI quote, which should not contain a monthly
split between capital and interest.
10.172 Some parties (Lloyds TSB, Capital One, Barclays, HBOS) questioned whether it was
necessary to include information about the price of the insured credit product—in
particular, the APR—in the personal PPI quote, as this was not relevant to comparison of the price of PPI on a stand-alone basis. The comments we received included
that this could add to the cost of preparing a personal PPI quote. We decided that the
credit APR should be included on the personal PPI quote as it would help consumers
put the PPI price into context—particularly when compared with a combined APR for
credit with PPI (see paragraphs 10.173 to 10.177). In addition, we consider that
inclusion of the credit APR would help those consumers who search on the bundle of
credit including PPI.
Requirement to include a combined APR in the personal PPI quote
10.173 Many parties (Capital One, Barclays, HSBC, Lloyds TSB, RBSG, BPF, GMAC,
MBNA) questioned our proposal to require the calculation of a ‘combined APR’ in the
personal PPI quote. These parties argued that requiring the calculation of a combined APR would add to the costs of producing the personal PPI quote, would risk
misleading or confusing consumers and would be contrary to the requirements of the
CCA and forthcoming Consumer Credit Directive (CCD). Parties also cited the
qualitative research that we carried out on price disclosure to suggest that
consumers would not find this metric helpful. Parties also argued that requiring a
‘combined APR’ appeared to be inconsistent with a remedy that had the effect of delinking the credit from the PPI.
10.174 The OFT agreed that a ‘combined APR’ might be useful in addition to other measures, but noted that it might be necessary to specify assumptions (additional to those
in the CCD) to ensure feasibility. The OFT noted that there were residual uncertainties about how the CCD assumptions would apply in the case of credit cards.
10.175 We noted these comments and the general level of opposition by distributors of
including this measure in a personal PPI quote. We also noted that there would be
practical issues in specifying a ‘combined APR’ and cost implications for distributors
of calculating a bespoke price measure for the personal PPI quote.
10.176 Nonetheless, we considered that a ‘combined APR’ provides consumers with useful
information and is an effective tool for comparison. A ‘combined APR’ can be compared with the APR on the credit product to give an indication of the extra cost of
PPI. Seeing the data presented in this familiar way could prompt consumers to think
carefully about the cost of PPI and appreciate the potential gains from search (see
paragraphs 5.10 to 5.15). A ‘combined APR’ is an indicator of the price of a combination of credit with PPI and including this information in the personal PPI quote could
prompt some consumers to shop around for a better combination of credit with PPI
after the credit point of sale. We therefore concluded that including a ‘combined APR’
would significantly enhance the effectiveness of this element of the remedies
package.
225
10.177 We decided that we should include a ‘combined APR’ in the specification of the
personal PPI quote. This specification will be subject to resolution of any practical
issues associated with including this measure and consumer testing of the personal
PPI quote as the remedy is implemented.
Specific issues in relation to revolving credit (ie credit cards and retail credit)
10.178 A number of issues were raised in relation to the credit card quote, reflecting the
different nature of the credit agreement compared with personal loans or mortgages.
Several parties (RBSG, Barclays Capital One, Lloyds TSB, HBOS, MBNA) said that
there was less (or no) scope or need for providing a consumer-specific estimate of
the cash cost of PPI, as this was not known at the time of taking credit and/or PPI.
Some of these parties suggested that annual and monthly costs of PPI would be
better estimated by means of a standard example (eg a consumer monthly balance
outstanding of £1,000 was suggested as being appropriate for credit cards), rather
than attempting to link these estimates to a consumer-specific factor, such as the
credit limit. This would also reduce the cost of producing personal PPI quotes. We
considered that this was an improvement on the original specification and that we
should adopt this proposal in relation to quotes for CCPPI and retail PPI.
10.179 Two parties (Lloyds TSB, Barclays) told us that the credit APR and credit limit were
not always determined at the credit card application stage. We considered that there
was little additional value in requiring the inclusion of the credit limit in the personal
PPI quote. In relation to the credit APR, we decided that, as this did not affect the
calculation of the cash cost of PPI, it would be acceptable to use a representative
APR, based on certain assumptions, which may change by the time of the eventual
agreement, if the actual APR were not known at the time of producing a personal PPI
quote.
10.180 SDGFS told us that it would not be possible to quote the cost of £100 of monthly
benefit from retail PPI without covering a range of possible credit scenarios and
raising a risk of failing to meet its regulatory obligations. Three parties (Capital One,
the BRC, JD Williams) noted that, if consumer-specific credit pricing information were
not required for CCPPI or for retail PPI, the personal PPI quote would be capable of
being provided at the point of credit application rather than at the point of credit
acceptance. This may fit with some providers’ existing processes and could make
PPI pricing information available to consumers at an earlier stage in the process,
helping consumers to shop around for PPI and for credit with PPI. This would not
necessarily bring forward the starting point for the prohibition period for retail credit
and CCPPI products, however, as these would be determined according to the
principles set out in paragraphs 10.107 and 10.108.
10.181 We decided that the specification of the personal PPI quote for CCPPI and retail PPI
need not require the actual APR that would be paid by a consumer, if this were not
known when the personal PPI quote was given. We decided that annual and monthly
costs should be illustrated by reference to a standard example which would be specified by the CC during the detailed implementation of our remedies package.
Obligation to provide information about the cost of PPI and ‘key messages’ in
marketing material
Summary of this element of the remedies package
10.182 We have found that a consumer’s ability to compare products is reduced by an
absence of information provided in a way that would help them compare PPI policies.
226
We also found that few distributors actively seek to win credit and/or PPI business by
using the price (or non-price characteristics) of their PPI policies. In particular, we
found that the time taken to obtain accurate price information is a barrier in relation to
the provision of PLPPI, MPPI and SMPPI. We decided that certain standard information should be provided to consumers in relevant PPI marketing materials to help
them understand the price of PPI and search more effectively for the best-value
stand-alone policy or combination of credit and PPI.
10.183 Figure 10.3 summarizes this element of the remedies package.
FIGURE 10.3
Provision of information in marketing materials
All PPI providers must prominently disclose the following information in any PPI marketing materials,
which include pricing claims or cost information, any indication of the benefits of the PPI product or its
main characteristics:
1. the monthly cost of PPI per £100 of monthly benefit;*†
2. that PPI is optional‡ and available from other providers (without specifying those other providers);
and
3. that information on PPI, alternative providers and other forms of protection products can be found
on the FSA’s moneymadeclear website.
*If the benefit pays out for less than 12 months, notice of this fact must also be clearly disclosed to consumers
alongside the cost of the policy.
†CCPPI and retail PPI providers must also show the cost of PPI per £100 of outstanding balance.
‡If the PPI provider is a stand-alone provider, it does not have to include the information that the PPI is optional
in their marketing material.
How the provision of information in marketing materials will address the AEC
10.184 This element of the remedies package will make it easier for consumers to compare
PPI products offered by different providers and to search for alternative PPI offers,
including stand-alone PPI and short-term IP policies. By increasing the prominence
of PPI prices within the information provided to consumers, the remedy will help to
address the failure of distributors to compete actively on the price of their PPI products. This requirement will also complement and enhance other measures that
address barriers to search, in particular the point-of-sale prohibition (paragraphs
10.34 and 10.35), the personal PPI quote (paragraph 10.158) and the unbundling of
retail PPI from merchandise cover. It also complements and supports the measures
to facilitate switching, in particular the requirement to provide an annual statement
(paragraph 10.302).
Issues raised
10.185 In general terms, all parties were in favour of increased transparency and making it
easier for consumers to compare PPI policies. However, there was a wider range of
views about the benefits of this element of the remedies package as a means of
increasing transparency. We received views on the following issues relating to the
detail of this requirement:
(a) the effectiveness of providing additional PPI information in marketing materials;
(b) whether to include information about PPI in or in close proximity to credit advertisements;
227
(c) interaction of the requirement with ICOBS; and
(d) the application of this requirement to intermediaries.
The effectiveness of providing additional PPI information in marketing materials
10.186 There was a broad consensus that pricing and other information should be included
in marketing materials specifically related to PPI products, though views were mixed
about the impact that this requirement would have in the absence of other measures.
Which? told us that only requiring the key messages to be included in PPI marketing
materials would not be effective. Abbey told us that putting additional information into
PPI marketing materials might confuse consumers. 70 These parties noted that distributors did not generally market PPI, and hence the impact on competition of a remedy
that affected only PPI marketing would be small. The BRC and SDGFS said that
retail PPI consumers already had enough information to enable meaningful comparison and the provision of further information was not necessary and could cause
information overload. SDGFS told us that the cost of PPI for a month was already
shown on a consumer’s monthly statement and it was unnecessary to include it on
marketing materials. JD Williams told us that it would be feasible to add a short statement of key messages but questioned the necessity of including all these statements
in marketing material because it would overload a consumer. One party (HBOS)
suggested that a requirement to include key messages would be likely to result in
more ‘small print’ and bulkier direct mail. However, most other parties, including
stand-alone providers such as the Post Office, considered that mandatory provision
of information about PPI could be helpful on direct PPI marketing materials and on
the Internet. Several parties (Lloyds TSB, Post Office, Paymentshield, RBSG) suggested that there should be a requirement to publish a qualitative ranking of the PPI
cover to enable comparison of what was a diverse product.
10.187 We note that some parties were concerned that any changes in regulations might
make advertisements more difficult to understand and less effective at attracting consumers. 71 In addition, some parties told us that ICOBS already addressed many or
all of the informational issues that we had found and that there would be no benefit
from providing additional information. 72 Cattles told us that the additional information
of price that was included would make all PPI advertising ‘financial promotions’ under
ICOBS regulations. Cattles also told us that additional cost of these requirements
could lead to fewer companies advertising and greater withdrawal from the PPI
market. Other parties told us that this remedy could distort competition by focusing
consumer attention on the price of PPI (rather than looking at both price and quality)
and give distributors an incentive to reduce both the level of PPI cover and innovation. 73 The FSA suggested that further specifications might be necessary to clarify
the relative prominence of the key messages within advertisements. We agreed with
the FSA that it was important for this information to feature prominently in marketing
material for this element of the remedies package to be effective. We consider that
the disclosure of monthly cost/monthly benefit should be sufficiently prominent that
an ordinary PPI consumer would notice it, when looking at/listening to the marketing
material concerned and it should be no less prominent than other descriptions of
price and/or product features. We noted that interpretation of these principles was
likely to vary across media.
70
Nationwide also said this in its response to the provisional decision on remedies p7 paragraph 3.2.2.
AFB response to Remedies Notice p3, AIFA/AMI response to Remedies Notice p2, Aviva response to Remedies Notice p4, Barclays response to Remedies Notice p5 and provisional decision on remedies p7, Cattles response to Remedies Notice p7. 72
Abbey response to Remedies Notice p6, Banque PSA response to Remedies Notice p3, Genworth response to Remedies Notice p4.
73
For example, Lloyds TSB p25 paragraph 2.3 and RBSG p14, in their responses to the provisional decision on remedies. 71
228
10.188 We considered that the form and content of the price information required under
ICOBS is useful when considering whether a particular product is suitable and affordable. However, the information currently provided to PPI consumers does not encourage searching and is not sufficient for easy price comparison of products between
PPI providers. 74 We noted that ICOBS already requires disclosure of the features of
a PPI product including a policy’s significant benefits, exclusions and limitations,
duration, and price information. We concluded that further disclosure of non-price
features of the product in marketing material is not necessary to address the AEC we
found.
10.189 We note that having additional information on the PPI price available in a comparable
form would be effective as a supporting measure to other elements of the remedies
package (notably the provision of personal PPI quotes and annual statements) which
would reduce the point-of-sale advantage and make it easier and cheaper for consumers to search for and switch suppliers. In a more competitive PPI market, we
expect a greater level of PPI marketing than we have observed (see paragraph
10.44). 75 We note that many providers have Internet sites which currently highlight
their PPI offerings and that consumers using these sites would benefit from this
option.
10.190 The OFT suggested that a remedy to require all PPI marketing material to include
key messages might be more than what was required to address the AEC, and
impose unnecessary cost. The OFT and some parties (MBNA, HBOS, RBSG) noted
that, for at least some marketing material, it should be possible simply to advertise
that PPI was available without triggering the requirement to include pricing or key
messages in PPI marketing material. We were told that there might in any event be
practical obstacles to including the key messages or pricing information in some
advertising media such as business cards or directories. The OFT suggested the use
of triggers, similar to the approach in the Consumer Credit Advertisements
Regulations—for example, whether the advertisement refers to pricing claims or cost
information, or any indication of the benefits of the product or its main characteristics.
The OFT suggested that this could form a model for this element of the remedies
package. Two parties (MBNA, HBOS) told us that there should not be a requirement
to advertise the key messages in some advertising media, for example telemarketing,
if a more general trigger for including the key messages was not adopted.
10.191 We consider that the requirement to include specific information in PPI marketing
materials will result in better-informed consumers with enhanced access to more
easily comparable information. We concluded that it was necessary to require pricing
information and key messages in PPI marketing materials. We concluded that the
key messages and pricing information should be as widely disseminated as possible
to maximize the competitive impact of this element of the remedies package. We also
agreed with the OFT that the Consumer Credit Advertisement Regulations were an
effective and proven model for triggering disclosures in such a way that avoided
requiring disclosures to be made when this would be unnecessary or impractical. We
therefore decided that the key messages and disclosure of price should be provided
in any PPI marketing materials which include pricing claims or cost information, any
indication of the benefits of the PPI product (including ‘peace of mind’ which, as
noted in paragraph 2.27, is the most frequently cited reason for taking out PPI) or its
main characteristics. We also decided that this requirement will apply both to direct
74
Under ICOBS, PPI providers have to provide a document that outlines prices in a durable medium—before the conclusion of a contract in a non-distance sale or immediately after the conclusion in a distance sale: http://fsahandbook.info/FSA/handbook/ ICOBS/6/4.pdf. Under ICOBS rules, firms are required to give price information to consumers in a way calculated to enable the consumer to relate it to a regular budget and total premium payable. 75
In this context, marketing materials includes advertising and direct mail. 229
marketing materials (such as statement inserts, emails and direct mail) and nondirect marketing material (such as newspaper and radio advertisements), that the key
messages and price disclosures should be sufficiently prominent that an ordinary PPI
consumer would notice them when looking at/listening to the marketing material concerned, and it should be no less prominent than other descriptions of price and/or
product features.
10.192 We considered whether to require further information—for example, a qualitative
ranking of the PPI product (see paragraph 10.186)—to be included in PPI marketing
materials in addition to the key messages and price disclosures, or whether we
should impose additional restrictions in relation to the presentation of non-price
information in marketing materials. We did not consider that this was necessary to
remedy the AEC that we found. However, we noted that providers may wish to
include a wider range of pricing and other information in their marketing material to
help promote their PPI product or bundle of credit with PPI. Such information may
reflect the information in the annual statement and the personal PPI quote as well as
details of the non-price characteristics of their products. We therefore decided not to
require any further information to be disclosed about non-price characteristics of PPI
products or to impose any further restrictions on other information provided in
marketing material, beyond those which are required by the FSA through the existing
ICOBS rules.
PPI information in or in close proximity to credit advertisements
10.193 We consulted on whether to require key messages and information about the price of
PPI in credit advertisements, or in close proximity to credit advertisements. Parties
were split over the effectiveness of such a requirement. Those parties which were in
favour of requiring more PPI information to be included in credit advertisements told
us that all advertising and marketing material that carried an illustration of the cost of
credit should also show the same illustration in relation to the credit and PPI bundle,
as this would encourage consumers to shop around for cheaper PPI and/or the
cheapest combination of credit with PPI. Abbey and RBSG were concerned that any
combined disclosure could have unintended consequences in the credit market and
on their ability to cross-sell non-PPI products at the credit point of sale. 76 Capital One
said that having a combined disclosure could cause both consumer detriment and a
distortion to competition in the credit market because it risked consumer alienation or
confusion to non-PPI consumers, the additional complexity could mislead the consumer on the cost of the underlying credit and a combined APR would not impact
equally on all PPI providers and would cause a competitive distortion.
10.194 Barclays, Nationwide and the FSA said that, depending on how the cost of credit and
PPI was presented in the advertisement, it might lead to a belief on the consumer’s
part that the PPI was not optional. 77 One party (Cattles) suggested that a
requirement to advertise PPI as part of all credit advertisements would lead to PPI
distributors subsidizing advertising costs for stand-alone PPI providers. HSBC
suggested that the requirement to advertise PPI when advertising loans and in ‘close
proximity’ could distort the market for credit, by substantially raising the cost of
advertising credit, leading to a reduction in advertising, and in any event was
disproportionate and impractical.
10.195 Other parties questioned whether it would be appropriate to require the disclosure of
the cost of PPI on credit advertisements in those situations where a credit provider
76
Abbey response to Remedies Notice p4, RBSG response to Remedies Notice pp3&5.
Barclays response to Remedies Notice, p7.
77
230
offered a number of products that could be bought alongside the loan—such as
household insurance (when selling a mortgage) or fraud insurance. We were told that
it could be misleading to consumers for providers to be required to provide an illustration of combined cost of credit plus PPI, but not to provide a similar illustration of
the cost of other products also offered alongside the credit product. GMAC told us
that it would be disproportionate for the motor industry and confusing for consumers
for the PPI key messages to be included in advertisements of goods where credit
payment options were mentioned, and noted that such advertisements are controlled
by the vehicle manufacturer. Banque PSA Finance also told us that there would be
significant cost implications with this remedy and, in any event, its credit products
and advertisements were different because provision of its retail finance was secured
against an asset (a vehicle). The BRC said that many PPI policies were linked with
other, non-PPI, elements (such as theft cover and legal helplines). It queried whether
PPI elements would be subject to standard disclosure whilst non-PPI insurance
elements would not be.
10.196 The FSA said that it seemed appropriate that a distributor which intended to sell PPI
at the point of sale should be required to provide the key messages when advertising
their credit product and in other marketing material. However, the OFT suggested
that it should be possible for a ‘simple’ PPI advertisement to be made in a credit
advertisement to alert the consumer of PPI being on sale, without requiring the key
messages to be displayed. The OFT and RBSG suggested that it should be permissible for a credit advertisement to mention that PPI was available, provided that any
additional information was in a separate PPI advertisement.
10.197 Some parties highlighted potential issues with credit regulations, 78 and in particular
about the interaction of the proposal to require disclosure of PPI information in, or in
close proximity to, credit advertisements with the CCD. 79 Cattles also said that there
were some general issues in correctly defining advertising and marketing materials.
The CCD will regulate the form and content of credit advertising from 2010. We
consider that the CCD sets out the boundaries of national regulation of the content of
advertisements for unsecured personal credit, including unsecured personal loans
and credit cards, but does not address the content of PPI advertisements.
10.198 Several parties said that requiring PPI information to be included in or triggered by
credit advertising seemed inconsistent with a point-of-sale prohibition. 80 Genworth
said that it would be confusing to consumers, who would be unable to buy the products they had received marketing and advertising information for at the credit point of
sale. RBSG said that a combined APR would send mixed messages to consumers,
who might assume that the PPI had to be bought with the credit product.
10.199 We noted the concerns that the parties have raised in relation to including PPI
content in credit advertisements. On balance, we considered that the effects of
greater competition in PPI markets should encourage PPI advertising generally. This
suggested to us that there is less need to prescribe greater advertising of PPI by
requiring PPI content to be put into or proximate to certain credit advertisements. We
also note the comments raised by the parties and the OFT which indicate that there
are a number of practical considerations concerning how this measure would operate, including exceptions that could be needed, for example for certain advertising
media. We considered that the effect of a number of exceptions to the general rule
78
ABI response to Remedies Notice p3, AFB response to Remedies Notice p2, Sterling Insurance response to Remedies Notice p3.
ABI response to Remedies Notice p3, Banque PSA response to Remedies Notice p3 in relation to the CCD 2008/48/EC. 80
Abbey response to Remedies Notice p10, Genworth response to Remedies Notice p5, RBSG response to Remedies Notice p11, Nationwide response to the provisional decision on remedies p7, paragraph 3.1.1. 79
231
would mean that the remedy would not be able to achieve its aim. We also noted the
additional costs that we were told would be borne by PPI distributors. We decided
that PPI content should not be required to be included in or in close proximity to
credit advertisements.
Interaction of the requirement with ICOBS
10.200 A number of parties referred to the ICOBS rules relating to PPI which came into force
in January 2008. 81 Nationwide told us that the FSA rules and principles for selling
PPI meant that all PPI providers must make it clear that PPI was optional. Lloyds
TSB also suggested that all PPI providers should have to comply with the same
requirements. In contrast, the Post Office suggested that stand-alone providers
should not have to include in their advertising that PPI was optional. We considered
that the link between the sale of credit and the sale of PPI was most relevant to PPI
distributors and decided that this point should be emphasized in the key messages to
be included in their PPI advertising. On this basis, we have not required the same of
stand-alone PPI providers. This remedy will not override any FSA regulation to the
extent that it requires all PPI advertisements to include a statement about PPI being
optional because it only confirms the requirement that is already placed on PPI
distributors by the FSA.
10.201 We consider that all elements of our remedies package are compatible with ICOBS
and we have worked closely with the FSA to date to ensure that this is the case. As
we implement our remedies, we will continue to work with the FSA to ensure that the
detail of the requirements that we put in place is compatible with the financial promotion rules in ICOBS.
Application to intermediaries
10.202 Our expectation is that this remedy will impact primarily on the distributor of the PPI
product and on stand-alone providers, rather than on intermediaries. Where intermediaries supply PPI products which are only available from one or more intermediary or intermediary networks, then the responsibility for ensuring that marketing
materials relating to that product comply with these requirements would lie with the
intermediary or intermediary network. In all other cases, we consider it likely that
intermediaries would rely on marketing materials provided by the supplier of the PPI
product.
The design of a common price metric
10.203 Almost all parties agreed on the need for consumers to have adequate purchasing
information and the usefulness of a common pricing metric that was clear and
succinct. In paragraphs 10.204 to 10.212 we review a number of possible pricing
metrics, and in paragraphs 10.215 to 10.222 we consider the submissions we have
received on the application of a common metric to providers of multiple PPI products,
credit card PPI and retail PPI.
10.204 HBOS said that developing a ‘common currency’ in which to express the price of PPI
would have value prior to the point of sale and would help to drive consumer searching. Lloyds TSB said that it would be necessary to take into account the different
characteristics of PLPPI, SMPPI, MPPI and CCPPI when looking at price but gener-
81
ICOBS 6 sets out product information requirements and 6.4 deals with the pre- and post-contractual requirements for protection policies.
232
ally supported the adoption of a common price metric. Citizens Advice suggested that
distributors should also disclose their commission in cash terms because a cash
amount could be more explicit than a percentage. Capital One suggested that consumers would find it difficult to understand how a price metric related to what they
pay on a monthly basis.
10.205 In addition to these general points, we received comments on a range of ways of presenting PPI prices and the price of combinations of credit and PPI, both in marketing
materials and in material presented to consumers at the point of sale. The price
measures we considered included:
(a) the annual or monthly cost of PPI presented as a cash amount in relation to a
particular scenario or quote;
(b) the monthly cost of PPI per £100 of monthly benefit. This metric compares the
monthly premium paid by consumers with the monthly benefits received if the
consumer makes a valid ASU claim. This metric is already commonly used for
MPPI and short-term IP, and is used in the FSA tables for credit cards;
(c) a ‘combined APR’, incorporating both the cost of the credit and that of the PPI
calculated using the standard approach for measuring APRs; and
(d) a ‘stand-alone APR’ for PPI, calculated as the difference between the combined
APR and the APR on the credit only.
10.206 We undertook some qualitative research with PPI consumers to assess their understanding of the various elements of credit and PPI pricing and to investigate which
ways of presenting price information were most easily understood by consumers. 82
The research showed that consumers were not comfortable working with percentages, and, whilst they could compare two different percentages (such as two APRs)
and understand which one related to the cheaper product, they were unsure what an
APR actually represented or what the actual cost to them would be. We note that
APRs were not supported by a number of parties as they were not considered a
useful indicator of price, quality or might act as a barrier to entry for stand-alone
providers. 83 The OFT said that an APR for the credit and PPI combined might mislead consumers to thinking that PPI is mandatory or a condition of credit. The OFT
accepted, however, that this might be useful information in a personal PPI quote.
10.207 We shared the OFT’s concerns that requiring a combined APR in PPI marketing
materials could mislead customers into thinking that PPI was mandatory or a condition of credit. We decided not to require the price of PPI to be required in PPI
marketing materials. This concern does not apply to the inclusion of a combined APR
in a personal PPI quote in the context of a point-of-sale prohibition (see paragraphs
10.173 to 10.177).
10.208 The consumer research also found that consumers were uncomfortable with large
numbers, generally preferring to think in terms of monthly costs. However, it was
suggested that whilst large numbers such as lifetime costs made consumers uncomfortable, the simple effect of making them uncomfortable might in fact encourage
them to consider the cost more carefully and search around more. Many parties
82
Qualitative research comprising of 24 in-depth interviews. These interviews were undertaken with a range of recent PPI consumer types (MPPI, CCPPI, PLPPI and SMPPI) and socio-economic groups. 83
Citizens Advice response to Remedies Notice p8, Which? response to Remedies Notice p3, Abbey response to Remedies Notice p4, paragraph 3.1.6, Nationwide response to Remedies Notice p6, and the FSA response to the Remedies Notice p9(b). 233
considered that monthly costs were useful for consumers, particularly when
assessing their budgets and overall affordability. 84 Which?, Cardif Pinnacle and
Cattles said that annual costs should also be provided.
10.209 Some parties suggested that monthly cost of PPI per £100 of monthly benefit was a
good measure of the price for PPI, which could be used by stand-alone providers and
credit providers alike and could increase pressure on pricing. 85 It was suggested that
use of this measure would remove the current confusion and increase price transparency across lending products that had led to the current wide variation in prices. 86
Some parties (Capital One, MBNA) suggested that this metric would not capture nonfinancial benefits and would make PPI look expensive and lead to some consumers
not taking PPI at all. Other parties instead supported the use of cost per £100
monthly cover to reflect the variable nature of some PPI products such as CCPPI. 87
One party (Lloyds TSB) suggested that the specific metric should be determined after
additional consultation during the implementation process.
10.210 Given the findings from our research and the responses we have received, we considered that pre-sale marketing materials should include a disclosure of price using a
standardized metric in an easily understandable format. We agreed with HBOS and
others (including the Post Office) that the beneficial effect of price disclosure on competition would be maximized if it enabled the development of a ‘common currency’ for
use in comparisons.
10.211 A pricing metric for use in marketing materials should be widely recognized and allow
any consumer to compare accurately the price of two competing products. We noted
that the monthly cost/£100 monthly benefit was already in use for MPPI and shortterm IP (and is used on the FSA tables for CCPPI) and that this metric had the desirable property for regular-premium policies that it was unaffected by factors that
varied across consumers, such as the size or term of the underlying loan, or the APR
on the credit. Whilst the monthly cost/£100 monthly benefit may differ for consumers
where risk-based pricing is used, we considered that it would still be a useful headline indicator that could be used to compare policies offered both by distributors and
on a stand-alone basis.
10.212 We decided that the monthly cost/£100 monthly benefit should be used as a common
metric for quoting PPI prices in marketing materials for all forms of PPI. In addition,
we decided that the disclosure of price according to this metric should be quoted in
PPI marketing materials (see paragraph 10.191) and should be featured prominently
within them along with the other key messages. This approach would maximize
comparability of the prices of PPI products, including PPI offered on a stand-alone
basis. We noted that one risk of using this metric was that providers could reduce the
number of months for which their products offered cover, in order to appear to offer
better value for money. We therefore propose to require PPI marketing materials to
give a clear notice to consumers if benefits on the product advertised pay out for less
than 12 months.
84
AFB response to Remedies Notice p3, FLA response to Remedies Notice p12, Genworth response to Remedies Notice p5, HSBC response to Remedies Notice p10, Nationwide response to Remedies Notice p6, Sterling Insurance response to
Remedies Notice p2.
85
Defaqto response to Remedies Notice p8, Post Office response to Remedies Notice p1, OFT response to the provisional decision on remedies p5, paragraph 19. 86
See Post Office response to the Remedies Notice, paragraph 14(b).
87
Cost per monthly cover is equivalent to monthly cost per monthly benefit for some products, such as MPPI, though not for variable premium products such CCPPI. Abbey response to Remedies Notice p4, Coventry Building Society response to
Remedies Notice p3, HBOS response to Remedies Notice p3, Council of Mortgage Lenders response to Remedies Notice pp2&3, Nationwide response to Remedies Notice pp7&10. 234
10.213 We do not consider that the provision of data on distributors or intermediaries’ commission, as proposed in paragraph 10.204, would be effective, as many providers are
vertically integrated, making the disclosure of commission meaningless to consumers.
Providers of multiple PPI products
10.214 We noted that some providers will offer more than one type of PPI policy to consumers. In particular, some providers already offer consumers a choice of different
levels of cover (eg LASU, ASU, AS and U). In these circumstances, it may not be
possible to use a single price to communicate the cost of PPI to every consumer who
sees the advertisement. However, we noted that ICOBS includes rules regarding
financial promotions and we considered that these rules adequately address the
issue of how variations in product offers could best be accommodated. 88 We also
noted that providers will have incentives to advertise aspects of their PPI product
characteristics that consumers are likely to find attractive.
Credit card PPI
10.215 We also noted that, in the case of CCPPI, a different metric—the monthly cost/£100
outstanding balance—was already well established.
10.216 Several parties (MBNA, Lloyds TSB, Barclays, Capital One, RBSG) told us that it
was not necessary to require disclosure of monthly cost/£100 of monthly benefit for
CCPPI policies given that it already had a commonly-used metric for comparison and
a dual metric would be confusing for consumers. Capital One told us that this metric
would make PPI ‘look’ more expensive compared with the traditional disclosure.
Some parties (Capital One, MBNA) suggested that for CCPPI it was not clear
whether monthly benefit would be related to the average balance outstanding over a
month, the highest credit balance during the month or the credit limit, as some providers calculated premiums and benefit on different bases. One party (Lloyds TSB)
told us that the price metric of monthly cost/£100 monthly benefit ignored the difference in the level of benefits that might be received after 11 months of claiming
where, for example, in month 12 the CCPPI product paid out the whole of the outstanding balance as at the point of claim. We also noted that the OFT considered
that a dual metric would be useful.
10.217 We decided that, for CCPPI, price should be disclosed both as a monthly cost/£100
monthly benefit and a monthly cost/£100 outstanding balance. We considered that
both forms of price disclosure provided useful information to consumers. The monthly
cost/£100 of monthly benefit provides an indication of value for money and provides
a point of comparison with other providers, including stand-alone PPI. The monthly
cost/£100 outstanding balance provides an indication of the way in which charges to
a consumer are calculated. We noted the comments that were made in paragraph
10.216 about the application of the monthly costs/£100 monthly benefit metric to
88
A financial promotion, with regard to PPI, is an invitation or inducement to engage in investment activity that is communicated
in the course of business. ICOBS 2.2.4 provides:
(1) This guidance applies in relation to a financial promotion that makes pricing claims, including financial promotions that indicate or imply that a firm can reduce the premium, provide the cheapest premium or reduce a
consumer's costs.
(2) Such a financial promotion should:
(a) be consistent with the result reasonably expected to be achieved by the majority of consumers who respond,
unless the proportion of those consumers who are likely to achieve the pricing claims is stated prominently;
(b) state prominently the basis for any claimed benefits and any significant limitations; and
(c) comply with other relevant legislative requirements, including The Control of Misleading Advertisements Regulations 1988. 235
CCPPI, and during the detailed implementation of this remedy we will seek to apply
this metric in a manner consistent with the disclosure of CCPPI prices in the FSA
comparative tables. 89
Retail PPI
10.218 We received submissions from retail PPI providers that emphasized the variable
nature of the credit balance that retail PPI applies to. 90 Similar to CCPPI, overall PPI
costs cannot be calculated in advance either on the PPI or associated retail credit.
The BRC said that any advertising based on the concept of annual cost or whole-life
cost of PPI would have to be based on typical examples which had the scope to be
misleading for non-typical consumers. We considered that illustrative scenarios are
likely to be of limited use to retail PPI consumers when making comparisons between
providers on the most appropriate retail PPI cover for their circumstances.
10.219 SDGFS and JD Williams told us that they already disclosed the equivalent monthly
amount per £100 balance outstanding and a monthly cost based on a typical consumer’s average monthly balance. However, similar to CCPPI, we noted that the cost
per £100 balance outstanding did not give any information about the benefits
received in the event of a claim and did not provide a point of comparison with standalone PPI. We noted that for retail PPI, the level of benefits received in the event of a
claim can vary substantially (see paragraph 6.96). SDGFS, the BRC and JD Williams
told us that stand-alone providers should instead be required to express their PPI
policies on the basis of £100 of protected balance and in this way standardization
could be achieved. Otto told us that it would calculate the benefit of each element of
a policy and average the sum.
10.220 We considered whether the monthly cost per £100 monthly benefit would, for some
reason, be non-applicable to retail credit. We concluded that this metric could be
applied to retail credit in a similar way to that in which CCPPI prices are currently
disclosed on the FSA comparative tables. 91 We noted that, as with other forms of
PPI, some of the benefits of retail PPI (eg life and critical illness cover) are not paid
out on a monthly basis. We also noted that SDGFS told us that this suggested that
this metric was inappropriate and impractical. However, like other forms of PPI, retail
credit providers offer ASU as a core part of their PPI offering and these benefits are
paid out on a monthly basis. We noted that SDGFS estimated that more than half
([] per cent) of its policies in force were held by people who were employed and
aged between 16 and 75 and thus covered for ASU risks under its policy. We also
noted that SDGFS told us that a standard metric was unnecessary because
customers always made a subjective judgement about the value of the benefits being
offered. It also said that, because [] per cent of its consumers had a monthly
repayment of less that £10 and [] per cent less than £50, a metric based on £100
of benefit would be meaningless and misleading.
89
The FSA’s approach to calculating this indicator for its comparative tables can be found in the ‘Credit and Storecard PPI Data
Requirements’ at www.welcomepacksonline.com/.
90
JD Williams response to Retail Remedies Notice p2 paragraph 8, and BRC response to Retail Remedies Notice p2. JD
Williams also told us that the FSA had recognized that there was a difference between premium paid on any non-revolving
credit agreement and on revolving credit, and that, as an authorized provider, it already complied with the FSA’s requirement to
require retail credit providers to disclose a ‘typical cumulative cost’.
91
The FSA’s approach to calculating this indicator for its comparative tables can be found in the ‘Credit and Storecard PPI Data
Requirements’ at www.welcomepacksonline.com/. Where the cost of credit is embedded in the price of the goods, and consumers pay for the goods and credit over a fixed period (eg 20 weeks), this metric can be calculated by assuming that a consumer spends a constant (weekly or monthly) amount and calculating the monthly cost per £100 monthly benefit, once the
outstanding balance reaches a steady state.
236
10.221 We considered that disclosing a monthly cost per £100 monthly benefit would provide
a useful point of comparison and providers could promote other features of their
policy if they desired. Such information may reflect the information in the annual
statement and the personal PPI quote as well as details of the non-price characteristics of their products. We do not propose any restrictions on other information that
may be provided in marketing material, beyond those which are required by the FSA
through the existing ICOBS rules regarding financial products.
10.222 We decided that the metric of price per £100 balance outstanding, already used by
retail PPI providers, was complementary to the disclosure of monthly cost per £100
monthly benefit, and retail PPI providers should prominently disclose both metrics in
their PPI advertising.
An obligation to provide information to OFT and FSA for monitoring and
publication and to provide information about claims ratios to third parties
Summary of this element of the remedies package
10.223 We have found that a consumer’s ability to compare products is reduced by an
absence of information provided in a way that would help them, and that few distributors actively seek to win credit and/or PPI business by using the price (or non-price
characteristics) of their PPI policies as a competitive variable.
10.224 We decided to require the provision of specified information to certain third parties as
set out below.
FIGURE 10.4
Provision of information to third parties
All PPI providers must provide comparative data to the FSA, as specified by, and in the format
requested by, the FSA. We also recommend to the FSA that it uses the information provided to it
under this obligation to populate its PPI price-comparison tables.
In addition to the information that the OFT may request from time to time for the purposes of
monitoring and reviewing the operation of the remedies package, all PPI providers that meet a
specified threshold (see paragraph 10.538) must provide the following information to the OFT on an
annual basis:
1. annual GWP, split by product type;
2. distributor penetration rates, split by product type; and
3. aggregate claims ratios for each provider, split by product type.
All PPI providers should provide to any person on request, aggregate claims ratios, split by product
type, for the previous year. These can be provided in the form of a range to be specified by the CC.
How the requirement to provide information to third parties will address the AEC
10.225 This requirement will make information available which will help consumers to compare the cost of PPI. Requiring the provision of key information on PPI and shortterm IP policies to the FSA, for use in comparative tables, will help consumers to
search for the best-value policy. By facilitating search and switching, this requirement
complements and enhances the point-of-sale prohibition, the provision of a personal
PPI quote, the provision of information in marketing material and the annual statement.
237
Issues raised
10.226 Most parties were in favour of increasing transparency and making it easier for consumers to compare PPI policies. 92 Some parties said that more comparable information would act as an encouragement for consumers to shop around as it would
give them the ability to compare PPI sold alongside credit with stand-alone PPI. 93
However, other parties believed that there was a risk that providing additional information would just increase complexity and could lead to a focus on price rather than
on the appropriateness of the product. 94 Two issues were raised by the parties
during consultation on this remedy:
(a) the effectiveness of using the FSA comparative tables; and
(b) whether claims ratios should be made available if requested.
• The effectiveness of using the FSA comparative tables
10.227 Many parties noted that the FSA tables provided consumers with comparable PPI
information. 95 In general, parties suggested that a remedy to bolster information provided for use in these tables was a positive step forward, a good concept and useful
for consumers. 96 However, some parties were concerned that the tables as presently
published were difficult for consumers to understand as the information was presented in a complex way. 97 For example, Abbey said that its initial impression of the
FSA comparison tables was that product information was not particularly clear and
the interpretation of product quality was left to the consumer. Some parties (Lloyds
TSB, MBNA, the BRC, JD Williams, Capital One, the Post Office, Paymentshield)
suggested that the FSA tables should be improved so that non-price benefits could
be compared. SDGFS and the BRC considered that sufficient information for comparison of retail PPI policies was already provided via catalogues, websites and
policy documentation. These two parties were concerned that incomplete PPI information might be presented or policies might be misrepresented by third-party comparative websites. We were also told by SDGFS that a PPI price-comparison website
did not make sense unless it was linked to the purchase of associated credit. SDGFS
told us that its consumers shopped on a bundle of goods, credit and PPI, although
the costs of PPI were transparent and separate from the costs of credit. JD Williams
told us that, as retail PPI was always sold separately from retail credit, and that
charges for each were entirely transparent, it should not be necessary for retail PPI
providers to provide information to enable comparison of bundles of retail credit.
Capital One suggested that the FSA tables biased stand-alone providers and should
be improved to give equal prominence to linked credit providers.
92
Assurant response to Remedies Notice p2, Axa response to Remedies Notice p4, BBA response to Remedies Notice p3, Barclays response to Remedies Notice p13, Citizens’ Advice response to Remedies Notice p8, Capital One response to
Remedies Notice p4, Cassidy Davis response to Remedies Notice p1, Genworth response to Remedies Notice p2, HBOS response to Remedies Notice p5, IMLA response to Remedies Notice p3, Lloyds response to Remedies Notice p14, Nationwide response to Remedies Notice p9, RBSG response to Remedies Notice p8. 93
Axa response to Remedies Notice p4. 94
Abbey response to Remedies Notice pp8&9, Aviva response to Remedies Notice p12. 95
Abbey response to Remedies Notice p9, ABI response to Remedies Notice p5, Axa response to Remedies Notice p4, Barclays response to Remedies Notice p13, BBA response to Remedies Notice p3, Cassidy Davis response to Remedies Notice p2, Council of Mortgage Lenders response to Remedies Notice p4, Coventry Building Society response to Remedies Notice pp6&7, FLA response to Remedies Notice p18, Genworth response to Remedies Notice p6, HBOS response to Remedies Notice p5, IMLA response to Remedies Notice p3, Nationwide response to Remedies Notice p9, Post Office response to Remedies Notice p4, RBSG response to Remedies Notice p8, Sterling Insurance response to Remedies Notice p6. 96
BBA response to Remedies Notice p3, Barclays response to Remedies Notice p13, Coventry Building Society response to Remedies Notice p6, Nationwide response to Remedies Notice p9. Axa response to the provisional decision on remedies p2, paragraph 7, Lloyds TSB response to the provisional decision on remedies p5, MBNA response to the provisional decision on remedies p5, paragraph 11, and Nationwide response to the provisional decision on remedies p2, paragraph 2.1. 97
Abbey response to Remedies Notice p9, BBA response to Remedies Notice p3, Lloyds TSB response to the provisional decision on remedies p5, and Capital One response to provisional decision on remedies pp23&24. 238
10.228 Nearly all parties, including the OFT, suggested that it would be more proportionate
to make the provision of information to the FSA for use in these simplified tables
compulsory, rather than to set up an additional independent website for consumers to
access. 98 In addition, a few parties (for example, Aviva, Cattles and Lloyds TSB) said
that care should be taken in creating an obligation to supply sensitive data to thirdparty commercial service providers as they might seek to leverage their role as the
operative of comparative tables with commercial objectives. One party (MBNA)
suggested that clear specification of the information to be provided to the FSA was
necessary. JD Williams told us that it thought the FSA comparison website would be
a useful aid to consumer search because some eligibility information was required
before a consumer could view PPI policies.
10.229 Citizens Advice supported the proposal to get better PPI information on to comparative websites, although it suggested that the effectiveness of the option could be
affected by many PPI consumers not having access to broadband Internet at home.
We noted these concerns, although our consumer research gave us confidence that
enough low-income PPI consumers were able to access the Internet for this element
of the remedies package to be effective. 99 We consider that any comparative pricing
table would be most effective if it were in a central location such as the FSA website,
which is easy to monitor and has comprehensive coverage and availability for
consumers.
10.230 Most parties considered that stand-alone providers should be included in any obligation to provide information for use in comparative tables, as the tables should be
used by consumers to compare stand-alone products with products sold alongside
credit. 100
10.231 In relation to retail PPI, JD Williams told us that, due to the nature of the cover
offered, it believed that direct comparison on the basis of monthly price per £100
monthly benefit (which is how prices are specified on the FSA website for credit
cards) might be impractical and/or misleading to consumers. However, we did not
agree that incorporation of retail credit into the FSA tables would be misleading or
confusing for consumers in this way. We acknowledge that there may be differences
in the levels of cover (for example, for different ages/employment status of consumers the benefits can be different within policies offered by the same provider), but
we considered that it is important for consumers to be able to compare the overall
value for money of alternative PPI policies. We also noted that SDGFS suggested
that price per £100 of credit balance is comprehensive. However, we consider that
the monthly price per £100 of monthly benefit provides a useful basis for comparison
with other products, including stand-alone PPI, against which consumers can take
into account other product features that are disclosed on the FSA website.
10.232 We consider that, given that the FSA tables are available, it would be more effective
and proportionate to add to them rather than create an additional table. We also consider that to be effective these tables need to provide a comprehensive view of the
market by including data about all PPI policies, including stand-alone policies and
short-term IP policies.
98
ABI response to Remedies Notice p6, Axa response to Remedies Notice p5, Cattles response to Remedies Notice p24, Coventry Building Society response to Remedies Notice p7, Nationwide response to Remedies Notice p10, Post Office
response to Remedies Notice p4, RBSG response to Remedies Notice p8. 99
BMRB Report—Telephone survey of PPI customers, February 2008, pp8&10. 100
Abbey response to Remedies Notice p9, Aviva response to Remedies Notice p13, Axa response to Remedies Notice p5, BBA response to Remedies Notice p4, Barclays response to Remedies Notice p13, Cassidy Davis response to Remedies Notice p2, Cattles response to Remedies Notice p15, Coventry Building Society response to Remedies Notice p6, FLA response to Remedies Notice p18, FSA response to Remedies Notice p13, Genworth response to Remedies Notice p6, Lloyds TSB response to Remedies Notice p14, Nationwide response to Remedies Notice p10. 239
10.233 We decided that PPI distributors and stand-alone providers should provide details of
their PPI policies to the FSA for use in its comparison tables. We consider that the
FSA comparison tables would help consumers to search for the best-value product
because they contribute to the information that is already available from different PPI
providers, reduce costs of searching the market and will help promote competition
between providers. PPI providers should provide the specific information required by
the FSA to complete its tables, in the format specified by the FSA. This would enable
providers to ensure that the correct information about their policies was used. We
decided to recommend to the FSA that it uses the information that we are requiring to
be provided to it under this element of the remedies package to populate its PPI
price-comparison tables.
10.234 We noted the concerns raised by some parties regarding the complexity of the FSA’s
tables but note that the FSA is in a position to enhance the usability of the website for
consumers should this be necessary. The FSA told us that it intended to discuss
further with the CC and the OFT the practicalities of making the tables more relevant
to consumers and reflective of changes in the market.
10.235 We considered whether distributors should be required to disclose the price of combinations of credit with PPI, as well as the price of PPI, on a comparative website.
We did not think that this would be necessary to remedy the AEC that we found and
we noted that the FSA website did not provide a facility for making such a comparison, which would require the development of a new site. We decided that we should
not require provision of information of combinations of credit and PPI.
• Claims ratios
10.236 During consultation we suggested that claims ratios of PPI providers should be provided to the OFT and third parties on request. Most parties provided a number of
reasons why claims ratios should not be made available to the public. 101 We were
told that consumers would not understand them (and indeed might be put off from
taking out PPI by them); 102 that they would change over time (and relate to the past
rather than the present); 103 that they were not important at an individual level
because they do not relate to the success or value of a claim and could be misleading; 104 that they do not recognize the costs of different distribution models; 105 that
101
Abbey response to Remedies Notice p9, ABI response to Remedies Notice p5, Axa response to Remedies Notice p5, BBA
response to Remedies Notice pp3&4, Banque PSA response to Remedies Notice p5, Barclays response to Remedies Notice
p13, Coventry Building Society response to Remedies Notice p6, Defaqto response to Remedies Notice p10, FLA response to
Remedies Notice p19, FSA response to Remedies Notice pp9&10, Genworth response to Remedies Notice p6, GMAC
response to Remedies Notice pp1&2, HBOS response to Remedies Notice p5, HSBC response to Remedies Notice p11,
Lloyds TSB response to Remedies Notice p14, Nationwide response to Remedies Notice p10, Openwork (noted this for MPPI
only) response to Remedies Notice p1, Post Office response to Remedies Notice p4, RBSG response to Remedies Notice p9
Sterling Insurance response to Remedies Notice p6. Axa response to the provisional decision on remedies p2, paragraph 8,
Banque PSA Finance response to the provisional decision on remedies p2, Capital One response to the provisional decision on
remedies pp23&24, Aviva response to the provisional decision on remedies p15, paragraph 3.2.3, Cardif Pinnacle response to
the provisional decision on remedies p1, Cattles response to the provisional decision on remedies p2, paragraph 5, Lloyds TSB
response to the provisional decision on remedies p30 paragraph 2.15, Barclays response to the provisional decision on
remedies p8, HBOS response to the provisional decision on remedies pp14&15, paragraph 3.1.2, MBNA pp5&6, RBSG
response to the provisional decision on remedies p14 and Aviva response to the provisional decision on remedies p15,
paragraph 3.2.3. Express Gifts response to the provisional decision on retail PPI remedies p3, and SDGFS response to the
provisional decision on retail PPI remedies p23.
102
ABI response to Remedies Notice p5, Banque PSA response to Remedies Notice p5, Barclays response to Remedies Notice
p13, Coventry Building Society response to Remedies Notice p6, Post Office response to Remedies Notice p6. Otto response
to the provisional decision on retail PPI remedies pp2&3.
103
Barclays response to Remedies Notice p13, Defaqto response to Remedies Notice p11, Genworth response to Remedies
Notice p6, Lloyds TSB response to Remedies Notice pp14&15, Post Office response to Remedies Notice p6.
104
ABI response to Remedies Notice p5, FLA response to Remedies Notice p19, Genworth response to Remedies Notice p6,
Nationwide response to Remedies Notice p10, JD Williams response to the retail PPI Remedies Notice p7 paragraph 34 and
SDGFS response to retail PPI Remedies Notice p12, paragraph 4.18. Cardif Pinnacle response to provisional decision on
remedies pp1&2.
240
they do not value non-monetary services that are provided; 106 that they would be
expensive to calculate; 107 that they were commercially sensitive; 108 that an obligation
to provide this information to competitors in the retail PPI industry if requested could
distort competition; 109 and that new entrants into the market would not have claims
ratios to provide. 110 We were told by the BRC and Otto that claims ratios would be
used to distinguish ‘good’ firms from ‘bad’ ones, which was a flawed analysis and
consumers should not be expected to understand what claims ratios were intended
to demonstrate without a detailed and complex explanation. We did not think that this
would be necessary to remedy the AEC that we found.
10.237 Both Which? and Defaqto said that they would find claims ratios helpful when comparing products, as they were an important indicator of how much an insurer would
be able or willing to pay out in claims, and a good indication of value for money.
Capital One said that it did not object to the publication of claims ratios as long as a
consistent method was used to calculate them, although it shared some of the concerns raised in paragraph 10.236. One party (Banque PSA Finance) considered that
provision of claims ratios to the OFT could be useful when monitoring the success of
the remedies package. We noted the views of the FSA and the OFT that claims
ratios could be a useful general guide for consumer groups, although not generally
useful for consumers. 111
10.238 Some parties suggested that as an alternative to the provision of data on claims
ratios, we should require provision of information on the percentage of claims that
were accepted. 112 These parties said that it was more important to consumers to
know the probability of their claim being turned down than for them to know claims
ratios.
10.239 We acknowledge the very strong representations we have received from almost all
the parties on requiring claims ratios to be made widely available. However, we consider that claims ratios are an important measure of both absolute and comparative
value for money and that this is the only readily available quantitative comparison of
PPI providers (all the other measures are qualitative). We anticipate that consumer
groups and others will make use of this information to inform consumers about the
value for money on offer by different providers. Making claims ratio data publicly
available will therefore serve as an additional discipline on providers and will help
address the failure of distributors to compete on price. We concluded that the
availability of the claims ratios was important in order to achieve a comprehensive
solution to the AEC that we found. We note that in a more competitive market for PPI
we would expect to see claims ratios increase, and making this information available
will enable the OFT and others to monitor the effectiveness of our package of
remedies.
10.240 We decided that PPI providers who meet a specified threshold (see paragraph
10.538) should supply claims ratios to the OFT once a year as well as GWP and
penetration rate. We consider that the GWP and penetration rates, as well as claims
---------105
GMAC response to provisional decision on remedies p3. Capital One response to the provisional decision on remedies pp23&26. 107
Cardif Pinnacle response to the provisional decision on remedies p1. 108
SDGFS response to the provisional decision on retail PPI remedies, p23 paragraph 3.74. Barclays response to the provisional decision on remedies p8, RBSG response to the provisional decision on remedies p14, Cardif Pinnacle response to the provisional decision on remedies p2, Paymentshield response to the provisional decision on remedies p3 and HSBC
response to the provisional decision on remedies, p1 paragraph 2.2. 109
SDGFS response to the provisional decision on remedies, paragraph 3.74. 110
Post Office response to Remedies Notice p4. 111
FSA response to the Remedies Notice, p11. 112
Abbey response to Remedies Notice p9, Nationwide response to Remedies Notice p10, RBSG response to Remedies Notice p9. 106
241
ratios, will be essential to the OFT when monitoring the effectiveness and impact of
the remedies and decided that this information should also be provided to the OFT by
large PPI providers once a year. We will also include standard wording in our Order
implementing the remedies package, which requires parties to provide to the OFT
any information and documents that it reasonably requires to enable it to monitor and
review the operation of the remedies package. In addition, we decided that all PPI
providers should make available to any party, on request, their claims ratios. To
address concerns about commercial confidentiality, we decided that claims ratios
provided to parties other than the OFT could be presented in the form of a range (eg
0–10 per cent, 10–20 per cent) with one aggregated banding above a certain figure
(eg 60 per cent). We will determine the ranges to be used, during the detailed implementation of our remedies leading up to the Order. We recognize that the obligation
to provide claims ratios falls on a larger number of companies than are covered by
the compliance reporting arrangements with the OFT. We considered that it was
necessary for this provision to have a wider application in order to ensure its effectiveness in delivering the benefits to competition set out in paragraph 10.239.
10.241 We noted the parties’ concerns regarding the logistics of calculating claims ratios (for
example, in relation to new products, intermediaries and the detail that should be
included to be consistent with other providers) but consider that these issues are not
insurmountable. We note that the largest parties to our investigation have been able
to provide claims ratios to us in the same format and that Datamonitor publishes
claim ratios of the top ten UK insurers across five major lines of business: accident
and health, liability, motor, pecuniary loss and property insurance, which is based on
data from the FSA. 113 We also noted that some PPI distributors, intermediaries and
stand-alone providers do not hold claims ratio information which would need to be
obtained from underwriters, but did not consider this an insurmountable barrier. For
example, intermediaries will be able discharge their obligations by disclosing a claims
ratio from the providers of the PPI policies that they offer to customers. Should it
prove necessary, during the drafting of the Order, we will include consequential obligations on underwriters to provide the information needed to meet their obligations.
We decided that it would not be necessary for new providers to provide information
about expected claims ratios. We consider that this would not add appreciably to the
OFT’s ability to monitor the market and could put new entrants at a competitive disadvantage to established players, who would not have to publish their forecasts. New
providers would only be required to publish actual claims ratios after they have built
up a reporting year’s claims data. We also noted the concerns raised by parties
about the burden of dealing with requests and having continuously to update the
information to be provided. We thought that it need not be unduly burdensome to
make this information available—for example, by publishing banded claims ratios on
a website. To address the latter concern, we decided that claims ratios could be
calculated once a year, rather than requiring claims ratios to be provided for the
12 months prior to a request.
10.242 We considered but decided against imposing a requirement to publish information on
the percentage of claims accepted to give consumers an indication of the quality of a
distributor’s sales process (a lower percentage might give some indication about the
proportion of consumers who have bought policies on which they cannot claim).
However, we decided that it would not help consumers to search for the best-value
policy for them in any material respect.
113
www.datamonitor.com/products/free/Report/DMFS2163/020dmfs2163.htm.
242
Prohibition on the selling of single-premium PPI policies
Summary of this element of the remedies package
10.243 We have found that consumers who want to switch PPI policies to alternative PPI
providers or to alternative insurance products are hindered in doing so. In the case of
single-premium policies, we found that terms which make switching expensive (such
as the level of rebates for early termination) act as barriers to switching for PLPPI
and SMPPI policies. Consumers also experience search barriers because of the
complexity and variety of single-premium pricing structures. We consider that this
element of the remedies package, which prohibits single-premium PPI policies from
being offered, will reduce the financial costs and searching complexities experienced
by consumers when switching PPI providers.
10.244 We have decided to implement the prohibition as summarized in Figure 10.5. These
measures aim to ensure the effectiveness of the remedies package in addressing
barriers to switching and search, which prevent, restrict and distort competition.
FIGURE 10.5
The single-premium prohibition
No PPI provider can charge for PPI on a single-premium basis. The only charge that can be levied on a PPI policy is a regular premium, paid monthly or annually by a consumer. If an annual premium is paid by a consumer, then a rebate must be paid to consumers on a pro-rata basis, if the consumer terminates the policy during the year. No separate charges for administration or for the set-up or early termination of a PPI policy shall be payable by the consumer.
How the prohibition of single-premium PPI policies will address the AEC
10.245 This element of the remedies package will fully address the switching barrier caused
by the terms on which single-premium policies are terminated (see paragraphs
10.256 to 10.263) and is the only option which would do so effectively. Singlepremium pricing also contributes to barriers to searching for consumers because of
the complexity and variety of single-premium PPI policies (see paragraphs 2.84 to
2.89); this element of the remedies package will reduce those barriers. It therefore
complements and enhances the other elements of the remedies package aimed at
facilitating consumer search, addressing the point-of-sale advantage and encouraging switching—in particular, the point-of-sale prohibition, provision of information in
marketing materials and to third parties, the personal PPI quote and the annual
statement.
Issues raised
10.246 Some parties (Paymentshield, [] and the Post Office) supported banning singlepremium policies as they considered that regular-premium policies allowed
consumers to switch more easily between products. Which? and Citizens Advice said
that they could not see how consumers benefited from single-premium policies.
HBOS noted that, although it did not support the remedy, it understood that a ban on
single premiums will help to reduce the perceived barrier to switching and aid
searching.
243
10.247 A number of parties have reviewed whether they will continue to offer single-premium
policies, during the course of the investigation and our consultation on this remedy,
and their current product portfolio will not be directly affected by implementation of
this remedy. HSBC currently offers consumers the opportunity to discuss their protection needs (not just those relating to the credit product acquired) with a Financial
Planning Manager instead of offering single-premium PLPPI. 114 In November 2008,
RBSG took the decision to stop offering single-premium PPI policies, for commercial
reasons (which we understand include, among other matters, changing penetration
rates and loan volumes). This decision took effect from close of business on
12 December 2008. Alliance & Leicester stopped offering single-premium policies
with effect from 1 December 2008. Lloyds TSB decided to stop offering new singlepremium PPI policies from January 2009. On 20 January 2009, the FSA published a
press notice which welcomed the move by Alliance & Leicester, Barclays, the
Co-operative Bank, Lloyds Banking Group (formed by the merger of Lloyds TSB and
HBOS) and RBSG to stop selling single-premium PLPPI by the end of January 2009,
and said that it expected other providers of single-premium PLPPI to review their own
positions. []
10.248 The following issues were raised by parties during consultation on this element of the
remedies package:
(a) whether a prohibition on single-premium policies would increase comparability of
PPI policies;
(b) whether alternative remedies would be effective and proportionate in addressing
barriers to switching; and
(c) whether a prohibition on single-premium PPI would result in consumer detriment.
Increased comparability of PPI policies
10.249 It was put to us that a prohibition on single premiums could help enhance consumers’
ability to search the market for the best value. Defaqto told us that a move to regular
premiums would enable consumers to shop around much more easily.
10.250 In paragraph 9.2(b), we found that those consumers who did want to compare products were hindered by product complexity, and that the variations in pricing structures
were one aspect of this. We therefore analysed whether a prohibition on single
premiums would help address barriers to search, either on its own or in combination
with other measures.
10.251 Single-premium policies have particularly complex pricing because a single-premium
PPI price includes an insurance premium and an APR on the loan that funds the
premium, which is calculated in different ways by different providers. 115 This means
that it is difficult for a consumer to make a comparison between single-premium PPI
policies without obtaining a specific quote from each provider (see paragraphs 2.84
to 2.89).
10.252 Our consideration of how a ‘common currency’ might be developed for use in marketing to compare the price of PPI between competing providers (paragraphs 10.203 to
10.213) indicated that there was no simple way of expressing the price of single-
114
See Appendix 2.3. See paragraphs 3–7 of Appendix 2.3. 115
244
premium products in marketing material that would be relevant to the majority of
consumers who saw it.
10.253 For example, in paragraph 2.89, we noted that, even with a policy offering identical
cover, the price of PLPPI expressed in terms of monthly cost per £100 monthly benefit from a provider can vary substantially for each consumer, depending on the term
of the loan and the APR. All other metrics that we considered exhibit a similar degree
of variation, when applied to single-premium policies or bundles of credit with singlepremium PPI. We also expect that we would observe a similar degree of variation for
other providers of single-premium PPI. We were unable to find any simple metric
which allows consumers easily to compare the cost of single-premium PPI products
in a consistent way.
10.254 These considerations highlight the difficulty that consumers currently face in making
a quick comparison between the cost of a single-premium policy and a regularpremium policy offered on a stand-alone basis (we are not aware of any stand-alone
PPI sold on a single-premium basis).
10.255 We therefore conclude that a prohibition on single premiums would contribute to
addressing the barriers to search arising from product complexity and variations in
pricing structures. We concluded that this was a relevant consideration in assessing
the effectiveness of this element of the remedies package—and its effectiveness
relative to alternatives, such as pro-rata rebates—as part of our remedies package.
Effectiveness and proportionality in removing barriers to switching
10.256 We did not receive any representations to suggest that a prohibition on singlepremium policies would be an ineffective remedy to the switching barrier identified in
relation to single-premium PPI. In paragraph 10.245, we explain why we consider this
prohibition to be effective in addressing the AEC, as part of a package of remedies.
During our consultation on this remedy we put forward two alternative ways of
addressing the switching barrier identified:
(a) requiring distributors of single-premium policies to offer a regular-premium policy
with an identical level of cover as the single-premium policy they offered; and
(b) a requirement mandating that single-premium products offered by distributors at
least meet minimum terms for early settlement rebates and any additional
charges (‘minimum rebates’). In particular, we asked whether early settlement
rebates should be based on a pro-rata calculation to address the switching costs.
10.257 After further analysis and consultation, we did not consider that either of these possible remedies would be an effective way of addressing the switching barrier that we
identified, and we have not included them in our remedies package to address the
AEC (see paragraphs 10.31 to 10.33). In summary, we consider that offering a
choice of products at the point of sale would not be effective at all and pro-rata
rebates would only be partially effective in addressing the switching barrier we identified. We had two concerns about whether a pro-rata rebate would be a sufficiently
effective remedy:
(a) It would remove some, but not all, of the financial costs of switching to a regularpremium policy and the remaining costs could be material, particularly where a
high APR was charged on the loan funding the single premium. We also noted
that those features of the current structure of PPI pricing that may provide an
incentive to switch to shorter-term single-premium policies may not be sustainable in the context of pro-rata rebates.
245
(b) It would require a degree of financial sophistication on the part of consumers to
switch with confidence, and we were not satisfied that a sufficient number of consumers possess the required level of financial sophistication to indicate that this
remedy would be effective.
10.258 Parties suggested that this element of the remedies package was not proportionate
to the AEC that had been identified for various reasons. We were told that, because
almost all PLPPI is single-premium, this remedy would have a significant effect on
the conduct of PPI distributors and providers in this sector of the PPI market
(Cattles). In addition, we were told (Cattles) that no evidence had been provided to
prove that single premiums were themselves ‘bad’ or ‘poor’. Another party
(Nationwide) said that a prohibition was unnecessary and disproportionate because it
would lead to reduced coverage, less consumer choice and higher cancellation
during the term of the PPI policy. One party (Barclays) stated that the CC had not
sufficiently considered the alternatives to a single-premium ban in light of effectiveness and proportionality and relevant customer benefits.
10.259 We accept that a prohibition on selling single-premium policies will have a significant
effect on the way some parties currently conduct business, in particular in relation to
PLPPI and SMPPI where in 2007 [] per cent respectively of active policies were
single-premium (although, as noted in paragraph 10.247, some very significant
suppliers of PLPPI stopped selling single-premium policies during the course of our
inquiry). However, having considered the alternatives carefully, we believe that a prohibition on selling single-premium policies is the only effective remedy to address the
barriers to switching and searching associated with single-premium policies. We consider that the scale of the AEC we have identified justifies robust intervention if that is
the only way the AEC can be addressed effectively. We did not accept that a prohibition on single-premium policies would inevitably lead to reduced coverage (see paragraph 10.419). We also did not accept that the prohibition would lead to increased
cancellations, save to the extent that the remedies package will encourage consumers regularly to consider whether they have the best-value policy for their needs,
at that point in time, and thus increase switching between policies. We consider
whether it will reduce consumer choice in paragraph 10.265.
10.260 One party ([]) told us that [] the decision of [] and other providers to cease
selling single-premium PLPPI (see paragraph 10.247) has a material impact on our
provisional finding of an AEC and makes a prohibition on single premiums redundant.
We did not agree that this development had a material impact on our AEC finding or
that it rendered this element of the remedy package redundant. First, whilst some
very significant distributors have decided to stop offering PLPPI, some distributors
continue to provide PLPPI and SMPPI on a single-premium basis. The FSA’s
confirmation of firms stopping only referred to some large distributors of PLPPI
(paragraph 10.247) []. Our AEC finding in relation to single-premium PPI related to
all forms of single PPI and to all providers. Second, the recent developments are
potentially reversible and we consider that a prohibition is necessary to ensure that
single premiums are not reintroduced at a later date, which would result in the
reintroduction of the barriers to switching and searching that we have identified.
Third, the search and switching barriers associated with single-premium PPI are only
one form of supplier behaviour giving rise to the AEC—this development does
nothing to address the point-of-sale advantage and the other elements of the AEC
and does not require reappraisal of the AEC for the purpose of adopting appropriate
remedies. To address the AEC that we found, we concluded that it remains
necessary to put in place a prohibition to the sale of single- premium PPI that will
apply to all PPI providers and to all forms of PPI.
246
10.261 In relation to a point raised by Cattles, we did not take a view on whether singlepremium policies are a ‘good’ or a ‘bad option’ for consumers; rather we considered
the effectiveness and proportionality of including this measure as part of a package in
addressing the AEC. We have found that single-premium policies result in barriers to
switching and searching (see paragraphs 5.63 to 5.75), in paragraphs 10.404 to
10.426 we conclude that there are no relevant customer benefits associated with
single-premium policies and in paragraph 10.272 we conclude that there would not
be any significant harm to consumers associated with prohibiting single-premium
policies.
10.262 A prohibition on single-premium PPI would have the greatest impact and be most
effective if consumers take account of the greater opportunities to switch PLPPI and
SMPPI and of the reduced barriers to search both before and after the credit point of
sale. This measure therefore complements all the other elements of our remedies
package. We concluded that including this measure in our remedies package was
the only effective means of addressing all of these aspects of the AEC that we found.
10.263 We noted the submission we received from Barclays that the transfer of existing
single-premium customers to regular-premium products would require a significant
implementation timeline and investment across sales channels. 116 We decided that
this remedy will apply to all PPI policies originating on or after the date the Order for
the PPI market investigation has application. We decided not to apply this remedy to
the ‘back book’ of existing single-premium policies, as we did not consider that this
would be necessary to ensure the effectiveness of the remedies package in creating
a competitive marketplace for PPI policies and would require additional investment
from distributors in relation to a type of PPI policy that would not be sold in the future.
Whether a prohibition would result in consumer detriment
10.264 Some PPI providers told us that a prohibition on single-premium policies would result
in significant consumer detriment. We consider the main arguments on this particular
issue in paragraphs 10.404 to 10.426, where we conclude that single-premium PPI
policies do not result in any relevant customer benefits within the meaning of the Act.
10.265 Some PPI providers have consistently told us that a prohibition on single-premium
policies will result in a reduction in choice for consumers. We accepted that a prohibition will remove products with a particular pricing structure from the market and that
this involves some reduction in consumer choice. However, we also considered that
a commensurate increase in choice will be made available via new regular-premium
products entering the market as a result of the prohibition on single-premium policies.
We were not aware of any consumers currently being offered a choice between
single- and regular-premium PPI at the point of sale (despite the FSA’s views that
single-premium policies were only appropriate for a limited number of people—see
paragraph 10.271). 117 We also noted that there is some evidence of consumer confusion about the distinction between single- and regular-premium policies (see pararaph 10.350).
10.266 One party (Abbey) told us that a ban on single-premium policies was likely to lead to
increased risk for underwriters and would make it difficult for companies to sell the
product on a sustainable basis (and thereby ultimately reduce choice). We disagreed
—in our view, parties have significant experience in designing policies and can
116
Barclays response to the provisional decision on remedies, p6, paragraph 4.3. We are aware of one experiment, conducted by [], where consumers were offered a choice of single- or regular-premium policies (see paragraph 10.349). 117
247
create regular-premium PLPPI and SMPPI policies with reasonable risks attaching to
them. Similarly we had no reason to conclude that selling regular-premium policies
would lead to problems of selling on a sustainable basis. We noted that other PPI
policies, such as MPPI, are successfully sold on a regular-premium basis and we
also noted that several large distributors stopped selling PLPPI with single premiums
during the course of this investigation (see paragraph 10.247).
10.267 We concluded that the benefits to consumers arising from the lower switching and
search costs would outweigh any detriment arising from the possibility of a reduction
in choice.
10.268 The ABI told us that the price of regular premiums rose by 40 per cent when singlepremium policies were banned in South Africa. 118 However, the South African
National Credit Regulator told us that the National Credit Act had only come into
force very recently and the early signs were encouraging; banks were still offering
PPI and appeared to be competing on price. 119 It also said that there seemed to be
more advertising of PPI. We therefore do not consider that, based on the limited
evidence available, the South African experience indicates that the price of PPI
policies in the UK will rise as a result of a prohibition on single-premium policies.
10.269 A number of parties have consistently told us that single-premium policies provide
benefits that cannot be replicated through regular-premium policies. For example, we
were told that PPI coverage was guaranteed for the term of the primary credit
arrangement, that more generous benefits could be supported by comparatively
lower-priced single premiums and a single premium was more convenient for some
consumers (see Appendix 2.10 for more detail). In addition, we were told that lower
socio-economic groups would be disadvantaged as they had limited savings, were
more likely to miss regular-premium payments and thereby would not receive the
benefit of PPI cover when they most needed it (Cattles).
10.270 As noted in paragraphs 10.423 to 10.426, regular-premium PPI products can and do
currently include terms that enable consumers to receive continuous cover, even
when they have missed premium payments.
10.271 We sought the views of regulators and consumer organizations on whether consumer detriment is likely as a result of this remedy. The FSA said that it had concerns over whether single-premium PPI products could ever be good value for
consumers and over the potential for such products to act as an impediment to
switching and the development of competition in these markets. In light of this, the
FSA told us that it considered that there were no significant risks to consumers from
a prohibition on single-premium PPI policies. The views of Which? and Citizens
Advice are set out in paragraph 10.246. We also noted the decision of certain parties
to cease selling PLPPI (see paragraph 10.247), which we believe supports the view
that we should not attach significant weight to the arguments put forward by some
parties that a prohibition on single-premium policies will lead to significant harm to
consumers.
10.272 Based on our analysis of relevant customer benefits, and the submissions from the
parties and others, we conclude that a prohibition on single-premium policies will not
118
South Africa introduced the National Credit Act 2005, which requires PPI to be optional (section 106(3)) and for the insurance
premiums to be payable on a monthly basis (or on a monthly or annual basis in the case of large agreements) (section 106(4)).
The National Credit Act also requires the credit provider to give the consumer the opportunity to take insurance from a provider
other than the credit provider, and to disclose the cost of the PPI to the consumer (section 106(5)).
119
www.ncr.org.za.
248
cause significant harm to consumers and should not affect our view of the proportionality of this part of the remedies package.
Design of the prohibition on single-premium PPI
10.273 A number of parties submitted that any remedy prohibiting single-premium policies
should not prescribe how regular premiums were charged to consumers. One
distributor (RBSG) told us that prescribing the form of premiums would be a
disproportionate restriction on providers and would be likely to result in capital repayment benefits, such as life cover, being removed. Two distributors (Abbey and
RBSG) suggested that it was unreasonable to prescribe a constant rate for regularpremium PPI. Abbey told us that this was because it would become more difficult to
absorb tax changes and alter benefits during the term of the cover (assuming
appropriate and reasonable notice to the consumer).
10.274 We considered that the effectiveness of this remedy in addressing barriers to search
would not be affected by permitting regular premiums that were not constant for the
entire term of the primary credit agreement, provided that the consumer can continue
to compare the premium of their policy with those of other PPI policies. For example,
an annually-renewed and annually-paid policy might decrease or increase its price
each year to reflect changes in its costs and perception of risks. We considered that
this form of variable annual premium is consistent with the types of premiums that
might be expected to be observed in a competitive market. For this type of policy, it
would be possible for consumers to use the information that we are requiring to be
provided in annual statements and marketing materials (eg the price expressed per
£100 per monthly benefit) to compare the premium they pay at any point in time, with
PPI policies offered by other providers. Similarly, an annual policy that is paid in
equal monthly instalments allows a consumer to compare the price of their PPI
against that charged by other PPI providers.
10.275 We note that providers might in the future develop a range of other possible pricing
structures. For example, some providers could offer a premium that decreases on a
monthly basis because it is set as a percentage of outstanding balance (an example
of such a policy is given in paragraph 10.408). We note that it could become difficult
to compare products charged on a reducing basis against products charged at a
monthly flat rate, if both charging structures were to become commonly used. This
could reintroduce a degree of complexity that would make it difficult for a consumer
readily to compare the premium they pay on a reducing basis with those premiums of
other regular-premium PPI policies. If PPI pricing were to become increasingly complex in this way, there is a risk that it might reduce the effectiveness of this remedy as
barriers to search could be reintroduced. While this could add to search costs, we
consider that the level of pricing complexity would be substantially less than with the
current mixture of single and regular premiums. We also noted the comments of
other providers (see paragraph 10.408), highlighting aspects of reducing premiums
that could make provision of products with this pricing structure unattractive to providers and customers, suggesting that this risk may not materialize in practice. We
therefore decided not to specify how regular premiums are to be charged, other than
requiring either monthly or annual premiums and the requirement not to charge
additional fees (see paragraph 10.277).
10.276 We also note that an annual PPI premium may, in some instances, be equivalent to a
single premium and would thus be prohibited. For example, an annual premium
being charged to insure a one-year personal loan would be a single payment, made
up front for a PPI policy (see Glossary). In effect, this means that PPI offered for
credit repayable over a term of one year or less is required to be charged on a
monthly premium basis.
249
10.277 A further question arises in regard to other separate charges that might be levied by
a PPI provider when establishing or terminating a PPI policy. One party (Lloyds TSB)
stated that there was no good reason to preclude charging administration fees, setup fees or early termination fees as this would be inconsistent with the OFT approach
to allowing charging for financial services, which favours the levying of cost-reflective
fees for ancillary services. However, we consider that these charges would hinder the
ability of consumers to search and switch PPI providers by adding to the complexity
of PPI pricing and making it more difficult to compare PPI premiums. Permitting these
charges would undermine the effectiveness of the remedies package in addressing
the AEC. We decided that these sorts of charges should be expressly prohibited on
the basis that they would reduce the transparency of premiums and effectively act to
prevent consumers from searching and switching their PPI provider.
Obligation to offer retail PPI separately from merchandise cover when both are
offered as a bundled product
Summary of this element of the remedies package
10.278 We found that the practice of bundling merchandise cover and retail PPI into one
combined policy made it difficult for retail PPI customers to evaluate the value for
money of retail PPI policies and therefore contributed to a lack of transparency. We
decided that requiring retail PPI providers to offer PPI separately would ensure that
consumers were provided with information that would enable them to compare the
price of retail PPI policies both with each other and with other forms of PPI. Improved
price information on their PPI policy will assist consumers to compare and search
effectively for alternative stand-alone PPI policies and short-term IP policies, or
bundled merchandise cover and PPI products.
10.279 This element of the remedies package requires that retail PPI providers offer PPI
separately where PPI and merchandise cover are both offered.
FIGURE 10.6
Unbundling retail PPI from merchandise cover
Where distributors of retail PPI offer an insurance package containing PPI and merchandise cover, they must
also offer, as a separate item, PPI cover alone.
How the requirement to unbundle retail PPI from merchandise cover will address the
AEC
10.280 By separating retail PPI cover for the merchandise cover offered by retail credit providers, this element of the remedies package helps address the barriers to search in
retail PPI, making it easier for customers to compare PPI products offered by different retail credit providers and to search for alternative PPI offers, including standalone PPI and short-term IP. By increasing pricing transparency, it helps address the
failure of distributors to compete on price. We decided that this measure, in combination with the other elements of the remedies package, would address the AEC that
we found in relation to the supply of retail PPI and it was not necessary to consider
whether merchandise cover should also be offered on its own. By increasing comparability of products, this remedy supports and enhances the effectiveness of other
elements of the remedies package in relation to retail PPI, in particular the personal
PPI quote, the provision of information in marketing material and to the FSA for
publication on its price-comparison website.
250
Issues raised
10.281 Three specific issues were raised by the parties during consultation on this requirement and we have considered them in the following categories:
(a) the financial impact of unbundling retail PPI from merchandise cover;
(b) possible conflict with other financial regulation; and
(c) the potential role of stand-alone providers.
• Financial impact of unbundling
10.282 Three financial impacts of unbundling were put to us. First, we were told that, if we
imposed this remedy, the cost to distributors of selling PPI and the bundled merchandise insurance and PPI product would increase, and that, as a consequence, the cost
of PPI would be higher to consumers. For example, SDGFS told us that sales calls
would lengthen considerably, adding significantly to costs. 120 JD Williams told us that
selling two products would lengthen telephone sales and increase costs disproportionately. 121 SDGFS also told us that retailers providing either a bundled or standalone product may withdraw from the PPI market (and only provide merchandise
cover) thereby reducing the range of PPI products available to consumers. However,
JD Williams told us that if we required PPI and merchandise cover to be sold separately, it would probably just sell PPI rather than try and sell three products (retail PPI,
merchandise cover and the bundle) in one telephone call. 122 JD Williams also told us
that, in order to keep costs down, it was considering ceasing to offer merchandise
cover. Express Gifts supported an objective that would enable customers to compare
prices and product features and make an informed choice. Otto told us that, although
it did not currently offer any retail PPI products, it had previously offered a suite of
products which would have complied with the remedy and, should it return to the
market, would be in a position to make available suitable products.
10.283 We considered whether a requirement to offer retail PPI separately from merchandise cover separately would increase selling costs to the extent that providers would
stop selling either or both components. We noted that telephone sales are a major
but not a dominant method of selling retail PPI as SDGFS sold nearly half ([] per
cent) of its PPI via the Internet and all of Express Gifts’ sales were online. SDGFS
told us that Internet sales were growing. It also told us that in relation to Internet
sales, there would be some development costs associated with unbundling the
products but no other incremental costs. 123 It said that call costs would be the most
significant extra cost of this remedy, although there would also be related administrative and compliance costs by making these changes.
120
SDGFS estimated that each extra minute in call duration would result in incremental costs of £[] a year, equivalent to £[]
per PPI policy. It told us that the current average call time it took for the consumer to make their first order, to set up a
consumer credit agreement and to discuss its current combined product was 7 minutes. It estimated that selling the unbundled
products would add, on average, 5 minutes to the call time, resulting in incremental costs of £[] a year, equivalent to £[]
per new PPI policy. Taking into account the remedies package, which SDGFS told us meant that it would need to offer three
separate policies, it expected the costs per PPI policy to increase by £[] per new PPI policy. In addition, in response to the
provisional decision on retail remedies, it said that additional time was also expected to be added to the average-length call to
enable retail PPI providers to confirm the receipt of a personal PPI quote and explain the key messages.
121
JD Williams estimated that selling the unbundled products would incur additional costs of £[] a year due to the longer
telephone calls required; it estimated that the call length would increase by 50 per cent, taking 11 minutes instead of the current
7 minutes. In addition, in response to the provisional decision on retail remedies, it said that additional time was also expected
to be added to the average length call to enable retail PPI providers to confirm the receipt of a personal PPI quote.
122
Response to Retail PPI Remedies Notice, p6.
123
SDGFS also told us that the added complication of attempting to explain the bundled product and the balance protection only
product alongside each other would result in fewer online sales being made and would require follow-up campaigns (email and
telemarketing) effectively to offer the two or three products.
251
10.284 We concluded that offering more than one product to consumers over the telephone
could take longer and could therefore cost the distributor more. However, we noted
that distributors would have incentives and the expertise to adapt their call scripts to
minimize any such extra costs. We were also not persuaded by the parties that it
would need to take significantly longer to explain to customers two alternatives
between which they would be able to choose, than it would to outline the characteristics of a bundled product comprising both PPI and merchandise cover. As such, we
considered that the parties’ estimates of the additional call times and hence the extra
costs associated with this remedy were likely to provide an upper bound on the likely
costs. In light of the evidence on call costs and the role of Internet sales in the distribution of retail PPI—for which call costs were irrelevant—we did not consider that
any increase in costs would be such as to prevent these distributors from selling PPI
on its own or with merchandise cover.
10.285 The second aspect we considered was whether bundled products offered better
value for consumers. JD Williams and SDGFS said that they sold a bundled product
because consumers wanted to purchase merchandise cover and retail PPI together.
We were also told that the short life and smaller balances of retail credit accounts
meant that bundled products would be likely to offer significantly better value than
unbundled products, because of the likely additional cost of selling retail PPI separately. The BRC, Express Gifts and JD Williams said that offering two separate products had the potential to carry two sets of acquisition costs, which meant that the cost
of PPI was unlikely to differ substantially from the cost of the bundle. It also said that
the additional cost of separating out the various elements to build unbundled products, along with extended call handling times for telephone-based sales, would be
bound to make these sales more commercially marginal.
10.286 We noted the comments that suggested the bundled product was likely to offer better
value for money than either of the components sold separately. We considered that
some consumers may agree with those comments and this measure would not prevent suppliers from offering a bundled product.
10.287 Third, we were told that the previous sales experience of JD Williams and Otto when
offering unbundled products suggested that a requirement to unbundle sales would
result in withdrawal from the PPI market. However, we considered that these
experiences were not representative of, or likely to indicate, what might happen if we
required retail PPI to be unbundled from merchandise cover. 124 Express Gifts told us
that in its experience there was little demand for a stand-alone retail PPI policy (ie a
retail PPI policy that does not include merchandise cover), otherwise one would
already be offered. SDGFS told us that the additional regulatory burden might cause
retail PPI providers to leave the market and leave consumers without the option to
purchase a bundled product (which has economies of scale). However, Express Gifts
told us that it was prepared to introduce an unbundled PPI policy for its consumers
which could be presented on a durable medium to allow comparisons to be made
before and after purchase.
10.288 We noted that consumer claims made under bundled PPI and merchandise policy
related mainly to PPI (in paragraph 6.16 we also note that the PPI premium is the
majority of the total premium). 125 We considered that this supported a view that PPI
124
JD Williams told us that it had sold merchandise cover on its own from June 2006 to July 2008 and ceased to offer the product []. The BRC noted that one of its members (Otto) introduced an unbundled product in 2007 but saw fit, for commercial
reasons, to withdraw it after a short time. In paragraph 6.140, we set out the reasons for Otto withdrawing from the retail PPI
market. []
125
For SDGFS, over three-quarters of claims ([] per cent of claims by number and [] per cent of claims by value) were
attributable to PPI. For JD Williams, almost all claims ([] per cent) related to PPI. For Express Gifts, almost all claims ([])
were attributable to PPI.
252
represented a substantial proportion of the value to consumers, and retail PPI providers, from a bundled product and thus parties would continue to provide PPI if we
implemented this remedy. The relatively low proportion of claims accounted for by
merchandise cover suggested that a stand-alone merchandise cover policy may be
of limited value to consumers. SDGFS told us that an insurance policy’s value to a
consumer could not be measured by reference to the number of claims made in
relation to it. For example, buildings or life insurance have a high value to consumers
but a very low number of claims. We noted this comment but did not change our view
about the contribution of PPI to the value to consumers of the bundled product.
10.289 We have concluded that the unbundling of retail PPI from merchandise cover, along
with other remedies, would make a substantial contribution to an increase in competition among PPI suppliers for consumers of retail credit and we have decided that
this measure should therefore be included in our remedies package. Our analysis of
profitability of retail PPI showed that distributors have earned profits that were persistently and substantially in excess of the cost of capital and led us to conclude that
PPI prices are higher than they would be in a well-functioning market (see paragraphs 6.134 to 6.138). 126 In light of our assessment of profitability and the
downward pressure on prices that would be imposed by our remedies package, we
considered that it was unlikely that any additional costs of providing retail PPI that
was not bundled with merchandise cover would be passed on to consumers as
higher prices.
• Regulatory risk
10.290 SDGFS and JD Williams told us that there was a risk that this remedy would bring
them into conflict with their obligations to the FSA concerning advised sales. The
BRC also said that there could be significant regulatory risk and operational difficulties concerning advised sales. In particular, there was a concern that it would be
very difficult to sell the separate products on a non-advised basis, as the higher
prices of the unbundled products would make it difficult to avoid an implicit recommendation of the better-value bundled product. JD Williams told us that this remedy
would increase its compliance and monitoring costs greatly. SDGFS told us that the
risk of not complying with these regulations could result in withdrawal from the
market because a distributor might not be prepared to accept the regulatory risk and
the compliance cost of this remedy.
10.291 We checked with the FSA. It said that distributors would not be obliged to move to an
advised sales process as a result of PPI and merchandise cover being unbundled.
To offer consumers a choice on a non-advised basis, it would be necessary to outline
the features of each of the products on offer (including significant exclusions, limitations, benefits and price) and undertake the relevant disclosures as required by
ICOBS. In a non-advised sale, the salesperson is required to provide information only
and must not make a personal recommendation. In PPI sales, which do not involve a
personal recommendation, the FSA requires firms to take reasonable steps to ensure
that a consumer understands that they are responsible for deciding whether a policy
meets their demands and needs. We concluded that it would be feasible to sell retail
PPI separately from a bundled product on a non-advised basis. For this reason, we
considered that any increase in compliance and monitoring costs as a result of this
remedy would not be substantial and would not affect our conclusion of the proportionality of inclusion of this element of the remedies package (see paragraphs 10.465
126
We note that JD Williams did not agree with this analysis in its response to the provisional decision on retail PPI remedies. It
suggested that because the merchandise part of the policy was so small as to be negligible against the additional requirements
for the sale, premiums would increase or at least cost the same as a bundled product.
253
to 10.514). We noted the view of SDGFS that, notwithstanding the view of the FSA, a
retail PPI provider may consider the regulatory risk and compliance cost too large
and withdraw from selling retail PPI altogether. We considered that a decision to
remain or enter into the retail PPI market will be driven by a number of commercial
decisions, including the profit available and the ability of a provider to comply with the
regulation associated with selling a particular product. We also noted that some other
retail credit providers (see paragraphs 10.282 and 10.287) told us that they considered that they could comply with the obligations associated with this element of
the remedies package, which we consider is necessary to address the AEC that we
have found.
10.292 SDGFS told us that unbundling each element of cover would present a clear conflict
with regard to Treating Customers Fairly. It said that it was not possible to comply
with its obligations whilst charging different rates to consumers for the same product.
10.293 We checked with the FSA, which said that it was unlikely that the pricing of products,
either individually or as a bundle, would create an issue with the Treating Customers
Fairly obligations. These obligations require that the appropriate disclosures are
undertaken and that sufficient timely information is provided about the product
options (including pricing) to enable an informed decision to be made.
10.294 We noted that some of the concerns raised by the parties related to the possibility of
being required to sell both PPI and merchandise cover separately, as well as the
bundle. We have decided not to require the offer of merchandise cover on its own, as
we do not consider that it is necessary to remedy the AEC identified. 127 We have not
needed to come to a view on whether offering all three products could be done on a
non-advised basis.
• Potential role of stand-alone providers
10.295 JD Williams told us that it did not think that a requirement to offer merchandise and
retail PPI cover separately was necessary. It did not accept our finding that there was
a lack of transparency created by the sale of retail PPI and merchandise cover
together and it believed that consumers could and did make an assessment as to the
value of each to them. It believed that stand-alone providers would be in a position to
offer bundled PPI/merchandise cover as an IP product.
10.296 We considered that stand-alone providers would not find it economically attractive to
develop new PPI products bundled with a form of merchandise cover on goods
bought through retail credit, given the relatively small size of this part of the PPI
sector. 128 On this basis, we considered that it was particularly important to increase
the comparability of stand-alone PPI.
•
Other issues
10.297 JD Williams told us that unbundling merchandise and retail PPI cover would be of
little benefit to consumers as retail PPI providers sold the goods, credit and the
insurance covering the goods and the credit. On this basis, the PPI offered by other
PPI providers would not be comparable and unbundling PPI would not increase the
options considered by consumers. SDGFS told us that consumers compared and
searched for alternative retail PPI policies based on the bundle among other retail
127
SDGFS told us that, if it were to be able to keep selling merchandise cover at the initial point of sale, then it would need to
offer three separate policies. 128
The four largest distributors of retail PPI received £73 million of GWP from consumers in 2007. 254
PPI providers. SDGFS also told us that its customers were primarily retail customers
and had a number of information sources available to compare goods, credit and PPI
when making a purchasing decision. We considered that unbundling would allow
consumers to compare the PPI offering with alternatives, including CCPPI and standalone PPI, while still allowing consumers to purchase a bundled product if they
wanted merchandise cover.
10.298 SDGFS told us that it did not believe that the requirement to offer the option of PPI
cover and merchandise cover separately was within the CC’s terms of reference. It
said that the CC should not put in place remedies that might affect the sale of other
products such as merchandise cover. We considered that offering PPI cover is within
our terms of reference. We have found that the bundling of retail PPI with merchandise cover acts as a barrier to search, contributing to the AEC. Taking remedial
action to address this barrier by requiring retail PPI to be offered without being
bundled with merchandise cover is within our statutory powers. 129
10.299 SDGFS also told us that the CC should take into account practical considerations
and proportionality of the proposed remedy against the small sums involved in this
sector. We considered that the detriment of the AEC is not small in absolute terms or
in relation to GWP. We review the practical considerations and proportionality of
implementation of our remedies in paragraphs 10.464 to 10.510.
10.300 JD Williams and Otto told us that we should consider the potential length of call and
the implications for consumers of the amount of information that would be required to
be presented to potential consumers when selling retail PPI over the telephone.
These parties noted that the FSA asked firms to consider the amount of information
given to consumers at the point of sale to avoid obscuring or confusing the message.
SDGFS told us that this remedy would be confusing because of the number and
different types of products (bundled and unbundled) that a provider would be offering
at the point of sale. We considered whether the offer of separate policies would
become more complex and confusing for consumers. We considered that there was
no reason why offering choice to consumers needed to be inherently confusing or
difficult. 130 We considered that this remedy is necessary to enable consumers to be
able to compare retail PPI more readily with alternatives, and together with our other
remedies requiring providers to provide certain information to consumers to facilitate
such comparison, the offer of two separate policies would not be confusing to consumers.
10.301 The BRC told us that instead of a requirement for PPI providers to offer an unbundled PPI policy alongside other products, its members would prefer to offer
consumers information about stand-alone policies (ie to split out the PPI and
merchandise cover costs) prior to any sale via a durable medium. 131 It suggested that
information about, and an offer of, an unbundled PPI policy at the point of sale could
create information overload for the customer, a lengthy telephone conversation and
higher premium costs due to the additional costs of sale. We were also told that,
because retail PPI was generally a distance sale over a telephone, consumers would
be unable effectively to compare products because they were in a ‘live sale’ situation.
We considered that other elements of the remedies package, particularly the point-ofsale prohibition and the provision of the personal PPI quote, would enhance customers’ ability to compare PPI products before committing to a particular provider.
However, these would not address the barrier to search associated with the bundling
129
We note that a similar remedy was imposed in the Store Cards Order. As noted in paragraph 2.3, some PPI distributors offer more than one type of PPI policy to their consumers. We received no evidence that consumers were confused by these offers. 131
This was also suggested by JD Williams in its response to the provisional decision on retail PPI remedies, p5 paragraph 24. 130
255
of merchandise cover with PPI. Providing information about the relative costs of providing PPI and merchandise cover would not fully address this barrier to search, as
consumers would still not be offered a retail PPI product that they could purchase,
whose price was directly comparable to PPI cover available from other providers. We
therefore did not think that the proposal simply to make information available about
the cost to distributors of providing retail PPI on an unbundled basis, without actually
unbundling PPI from merchandise cover, would be sufficient to address the AEC we
have found, either on its own or in conjunction with other measures.
An obligation to provide annual statement of PPI cost and a reminder of the
consumer’s right to cancel
Summary of this element of the remedies package
10.302 We decided that an annual statement would raise consumer awareness of their
ability to switch PPI provider. The annual statement should include information
similar to that provided in a personal PPI quote to assist consumers to compare the
cost of their PPI policy with alternative policies as well as information about consumers’ rights to cancel the policy. The summary of this element of the remedies
package is set out below.
FIGURE 10.7
Requirement to provide an annual statement
All PPI providers will provide an annual statement to all their PPI consumers* (see Appendix 10.2 for
details).
Provision of this statement will be the responsibility of the company which sold the PPI policy to the
consumer (ie the distributor or the stand-alone provider), other than for sales made by intermediaries
where provision of this statement will be the responsibility of the underwriter (or distributor or standalone provider) with which the consumer has an ongoing relationship.
The statement must be provided separately to any information on a credit product held by the consumer but might be included with other information relating to the PPI policy.
*PPI consumers in this context do not include those PPI consumers who have not paid any PPI premium on that
policy in the previous year.
How the requirement to provide an annual statement will address the AEC
10.303 We found that there are barriers to switching (paragraph 5.87). The prohibition on
single premiums (see paragraph 10.34) removes the financial costs of switching from
single-premium PPI policies. By raising consumer awareness of their ability to switch
PPI provider, this element of the remedies package will enhance the effectiveness of
that measure and will encourage consumers periodically to consider whether their
PPI policy still represents the best-value option for them. The specification of the
annual statement has been designed to complement the requirements to provide a
personal PPI quote, to provide pricing information in marketing materials and to
provide information to the FSA for use on its website. This will help those consumers
who are prompted by receiving the statement to consider alternatives to their current
PPI policy to make comparisons with other products.
Issues raised
10.304 Many parties, including the OFT and the FSA, were in favour of an annual statement.
In general, parties said that annual statements would increase transparency and
256
clarity and would be likely to increase consumer switching. 132 Parties told us that an
annual statement would at least ensure that consumers were regularly considering
whether the policy continued to meet their needs (the Post Office) and would remind
consumers of their policy (PaymentShield). Openwork said that annual statements
would bring PPI more in line with consumers’ understanding of other general insurance products, such as motor and home insurance. [] and [] suggested that
annual statements were an opportunity to gain consumers through increased
marketing opportunities.
10.305 Some parties said that this remedy would encourage consumers to cancel their policies, rather than switch, which they said would be detrimental to those consumers as
they would no longer be protected. 133 CCPPI distributors said that their consumers
received a credit card statement every month, which included the monthly PPI
premium, and the provision of an annual statement would be superfluous and add to
costs. Similarly, retail PPI providers and the BRC told us that retail PPI policies
allowed cancellation at any time and, in this context, an annual reminder of this was
unnecessary. These parties also told us that an annual statement of cost was not
appropriate for a product with a variable balance and cost, and which itemized PPI
on a monthly basis. 134 The BRC said that most of the issues raised by us could be
addressed by a statement that other policies were available and that existing cover
could be cancelled without penalty and that the remedy was disproportionate to the
retail PPI premiums earned. 135
10.306 We considered that an annual statement would be helpful to consumers and would
increase switching because it would increase transparency and help consumers
compare PPI prices within the market (and with other insurance products such as
critical illness or income protection insurance). We also considered that the remedy
would encourage marketing spend on PPI, which we consider currently to be low,
given the size and profitability of the market (see paragraph 10.44). The annual
statement would also increase consumers’ awareness of PPI, the cost they are
paying and their ability to switch products. We noted those comments that suggested
that an annual statement would unnecessarily duplicate information that is already
provided on a monthly basis, but considered that the benefits we have identified
apply equally to all types of PPI. Whilst we recognized that retail PPI and CCPPI
policies can be cancelled at any time, we decided that the obligation to provide an
annual statement will give consumers an important prompt to consider whether their
PPI policy meets their current needs, and the information needed to compare it with
other PPI policies.
The design and implementation of this measure
10.307 Many parties commented on the particular design and implementation of this aspect
of the remedies package; we have considered these comments in the following
categories:
(a) when the statement should be sent and by whom;
132
See, for example, Citizens Advice response to the Remedies Notice p12, and HBOS response to the Remedies Notice p5. RBSG response to the provisional decision on remedies, pp17&18.
133
See, for example, AMI response to the Remedies Notice, p4, and Defaqto response to the Remedies Notice, p12. Capital One suggested that this might increase cancellations by [] per cent in its response to the provisional decision on remedies, p14.
134
JD Williams response to the Retail PPI Notice of Remedies p12 paragraph 52, SDGFS response to the Retail PPI Notice of Remedies pp34&35, the BRC response to the Retail PPI Notice of Remedies p4, and GMAC response to the provisional decision on remedies, p4.
135
BRC response to the provisional decision on retail PPI remedies, p4. 257
(b) what should be included in the statement;
(c) the appropriate standardization of statements;
(d) whether annual statements are necessary for nil-balance accounts; and
(e) the application of this remedy to the back book of PPI policies.
• When the statement should be sent and by whom
10.308 Respondents were divided as to whether the annual PPI statement should be sent
out with the annual credit statement (CCA 2006 statement). Some parties, mostly
distributors, suggested that it should be provided with the annual credit statement to
reduce costs, although the point-of-sale remedy would cause the appropriate timing
for each to differ. 136 Genworth said that consumers were more likely to read an
annual statement relating to PPI if it accompanied a statement of debt. 137 Other
parties told us that credit and PPI statements should be sent separately to avoid
suggesting to consumers that PPI was only available from the credit provider. 138 The
Post Office said that combining statements would minimize costs for distributors
compared with stand-alone providers who would incur the cost of producing separate
statements. SDGFS told us that there was no requirement to send an annual credit
statement to the vast majority of its retail credit consumers so combining the timing
with an annual PPI statement was not relevant for retail PPI. MBNA and SDGFS
suggested that it would be beneficial to combine the timing of the annual statement
with its monthly statement mail.
10.309 It was suggested by some parties that the statements could all be sent at the same
time in the calendar year to encourage companies to market and compete for PPI
efficiently and with maximum impact. Other parties said that this might confuse some
consumers who might get an annual statement just after taking out a policy and
would possibly be more costly than sending statements on the anniversary of each
policy’s issue. MBNA told us that annual statements might need to be sent out at
times when an account had not yet been fully updated for reverse transactions and
had the potential to confuse a customer (for example, fraudulent activity). We considered that obliging PPI providers to supply statements at the same time of year was
likely to be costly and decided not to require it because the same effect of the
remedy could be achieved by the annual statement being sent on the anniversary of
the consumer taking out the policy. We noted that distributors are able to manage the
risk of statements not being updated for the latest transactions, in the context of
periodic credit statements, and did not consider the risk to be greater in the context of
annual PPI statements.
10.310 Some parties (Barclays, MBNA, Capital One, Paymentshield, a small distributor) told
us that significant additional costs would be incurred to provide annual PPI statements. MBNA suggested that, to address the additional costs of this remedy and to
acknowledge the preferences of many consumers, we should specify that statements
provided in a durable medium were sufficient. 139 A small distributor suggested that
136
See, for example, Abbey response to the Remedies Notice, paragraph 5.2.2, FLA response to the Remedies Notice, p26, and Axa response to provisional decision on remedies, p4, paragraph 21. 137
Genworth response to the provisional decision on remedies, paragraph 3.7. 138
OFT response to the provisional decision on remedies, p6, paragraph 31, Which? response to the provisional decision on remedies, p8, paragraph 6.1, and the Post Office response to the Remedies Notice, p6, paragraph 27. In contrast, MBNA suggested that where a consumer had already decided that purchasing credit and PPI from the same PPI provider was appropriate, it was inappropriate, indirect and inefficient to then attempt to persuade consumers to evaluate these products separately by supplying the documents separately. 139
Durable medium in this context include email and access via a website. 258
annual PPI statements could be sent out with their annual renewal documents.
Nationwide suggested that an annual statement was unnecessary where annual
renewal documentation was already issued with a significant amount of cost information and a 30-day right of cancellation.
10.311 We decided that the annual PPI statement should not be provided together with the
annual credit statement or with any other statement issued in relation to the credit
(such as a monthly credit card or retail credit statement). We noted the observations
that the point-of-sale prohibition could make it impractical to send the two statements
together and agreed with concerns that providing the two statements together would
risk linking them in consumers’ minds, which could discourage consumers from
switching, undermining the effectiveness of this element of the remedy package. The
same considerations also apply to sending the annual statement together with other
credit statements, such as a monthly credit card or retail credit statements. We also
noted that some PPI providers synchronize all annual renewal dates and we do not
intend that this timing will need to be amended by this measure. 140 We decided that,
for annually renewable policies and policies paid for by an annual premium, the statement should be sent during the few weeks leading up to the renewal date (which
would usually be the anniversary of the policy’s inception), so that consumers have
the statement at the point when they are considering whether to renew their policies.
For monthly renewable policies, we considered that the timing of the first statement
was less important, provided that an annual statement cycle was established after
that date. We therefore decided that the first statement should be sent out during the
first 12 months of the policy’s inception and subsequent statements should be sent
out on the anniversary of the first statement. We noted the submissions that we
received on the costs of this remedy and we decided that provision of the annual
statement by electronic means should be permitted where this is requested by a consumer and that the annual statement can be included alongside annual PPI renewal
documents, where relevant.
10.312 We decided that provision of the annual statement will be the responsibility of the
company which sold the PPI policy to the consumer. In the instance of a policy sold
by a distributor or by a stand-alone provider, an ongoing relationship with the consumer is maintained after the initial sale. However, intermediaries do not generally
maintain an ongoing relationship with the consumer in relation to the credit or the PPI
policy and we consider that it would impose an unreasonable burden to require them
to seek the information from each of the credit and PPI provider to complete an
annual statement.
10.313 Many parties told us that an obligation on intermediaries to provide an annual statement would be impracticable, onerous and expensive. We decided that in the case of
intermediaries, provision of an annual statement will be the responsibility of the
underwriter (or distributor or stand-alone provider) with which the consumer has an
ongoing relationship.
• What information should be included in the statement
10.314 We considered that the annual PPI statement should include both financial and
qualitative information to act as a reminder to the PPI consumer of their individual
PPI costs and aid comparison with PPI offered by competing PPI providers. Of the
financial information, most parties suggested that the monthly or annual cost of PPI
was of more interest to a consumer than its lifetime cost. 141 However, many parties
140
See, for example, a small distributor’s response to provisional decision on remedies. See, for example, HBOS response to the Remedies Notice, p5, and MBNA response to the Remedies Notice, p9. 141
259
(MBNA, SDGFS, Lloyds TSB, Otto, Capital One, Express Gifts, JD Williams and the
BRC) told us that a statement of annual cost or whole-life cost would be unnecessary
(and we were told that in the case of whole-life cost impossible to predict) and confusing to consumers for reasons including the fact that credit card and retail credit are
revolving credit. Capital One told us that the inclusion of any credit information risked
confusing customers about the optional nature of PPI.
10.315 Cardif Pinnacle suggested that the statement could contain the average balance
outstanding for CCPPI consumers over the previous period, which would give consumers and stand-alone providers an indication as to the level of protection the
CCPPI consumer would, on average, need. The Post Office suggested that an
average CCPPI balance would encourage switching, and the OFT suggested that the
remaining balance on loans could also be included in the annual statement. However, Barclays considered that it would be very costly to develop the systems to
incorporate credit metrics into the PPI statement. Other credit providers (Lloyds TSB,
JD Williams, Otto) suggested that the provision of an average credit balance had little
value for consumers and Capital One suggested that the concept was equivalent to
requiring an annual credit card statement which could lead to broader adverse implications. Capital One also suggested that average credit balances could lead to some
consumers under-insuring to only the level of their average credit balance. SDGFS
said that the provision of an annual PPI cost conflicted with FSA guidance that
annual costs were not appropriate for policies without a fixed premium. The FSA did
not express this concern to us.
10.316 Of the qualitative information that could be included in the statements, many parties
said that the benefits of PPI and the need to protect repayments against certain
events should be emphasized. 142 Other parties suggested that consumers should be
reminded that they had a choice of PPI supplier and the risks of switching policies.
Some parties suggested that the likely impact of qualification periods and pre-existing
conditions should be noted on the statement. Aviva suggested that consumers
should be provided with a query line for general enquiries. The OFT suggested that
changes to terms and conditions should also be included.
10.317 Cardif Pinnacle suggested that information on a consumer’s claims history could
be included which would remind the consumer of the benefits received during the
year and inform alternative PPI providers of a consumer’s risk or eligibility when
providing a PPI quote. 143 The Post Office suggested that this would be useful information for consumers and stand-alone PPI providers. Other parties (HBOS, MBNA,
JD Williams) were concerned about the cost of providing this information on an
annual PPI statement. Some parties (MBNA, JD Williams, Otto) told us that they did
not deal directly with claims made under PPI policies and this information would have
to be obtained from the underwriter of the PPI.
10.318 We considered that average monthly cost and the annual cost of PPI for the previous
year are the most useful indicators to include in an annual PPI statement for consumers. This information will show consumers how much they regularly pay for PPI
and how much they have paid in the last year, which will assist them to decide
whether it is worthwhile searching for an alternative PPI policy. Including the monthly
price of PPI in an annual statement will facilitate comparisons with alternative PPI
policies. We noted that including monthly cost information in the annual statement
142
See, for example, Assurant response to the Remedies Notice, p4, and Cassidy Davis response to the Remedies Notice, p4, MBNA response to the provisional decision on remedies, p16, paragraph 67, and Genworth response to the provisional decision on remedies, p6.
143
For example, Cardif Pinnacle has offered a stand-alone PPI policy for consumers who had made no claims on their existing PPI policy. Banque PSA Finance also supported the inclusion of this information in its response to the provisional decision on remedies, p3. 260
would make it easier for a consumer to compare their current policy with an alternative PPI policy quote (see paragraphs 10.157 to 10.181).
10.319 Accordingly, we decided that annual and monthly costs of PPI should be provided as
part of the statement.
10.320 We considered the disparate submissions that we have received on whether the
average balance of credit cards and retail credit for the previous year should be provided. We decided to require that this information be provided as we considered that
it would encourage the development of stand-alone provision in these PPI markets
by providing a useful tool for consumers to assess the level of PPI cover that they are
likely to need, on average each month, which will help consumers search for alternative cover. In this context, we noted that Barclays’ ‘Plan B’ CCPPI product allows
Barclaycard consumers to cover the balances across all their credit and store cards.
Consumers choose a fixed amount to cover their typical monthly balance across all
cards (up to £15,000). Inclusion of this information on a statement will also help
stand-alone providers assess the level of cover needed by a consumer and will provide a more effective metric for doing so than alternatives such as the credit limit.
10.321 We decided that the inclusion of a telephone number for a consumer to discuss or
cancel the PPI policy and obtain general information on the extent of the PPI policy
(for example, the elements of cover or, for CCPPI, the percentage of credit balance it
would pay in the event of a claim) should be required.
10.322 We decided not to require that information be provided on the risks of not having PPI
or of switching PPI providers. We considered that the former was unnecessary and
the latter would be counterproductive for a remedy aimed at making it easier for consumers to switch. We also decided, on balance, not to require information about the
claims history of consumers. While, in some circumstances, this might assist consumer switching—by enabling a consumer to demonstrate their recent claims history
to a new provider—we did not consider that its inclusion in the annual statement
would be necessary for our remedies package to be effective in addressing the AEC.
We also acknowledged parties’ representations that obtaining consumer claims
history for inclusion in annual statements would create practical difficulties for some
providers (particularly those which were not vertically integrated) and would add significant additional costs.
10.323 We noted that a requirement to use an APR and a combined APR calculation on an
annual PPI statement would increase its comparability with personal PPI quotes (see
paragraphs 10.157 to 10.181). However, a number of parties (MBNA, GMAC, Lloyds
TSB) and the OFT told us that it was not appropriate to require an equivalent APR for
an annual statement. By definition, a statement is a record of historical costs and in
contrast an APR is a percentage calculation of total expected future costs. The OFT
noted that the APR is fixed at the point a credit agreement is entered into, and creditors are not required to recalculate the APR during the period of the agreement.
Lloyds TSB also said that an APR for CCPPI annual statements could conflict with
the Consumer Credit Act and the CCD. We considered that, because the annual PPI
statement is limited to presenting the monthly and annual cost of PPI, it is not possible to use an annual PPI statement to compare the cost of a bundle of credit and PPI
against those personal PPI quotes which do. We decided not to require inclusion of
an APR or a combined APR on the annual PPI statement.
• Standardization of statements
10.324 Some parties were in favour of standardized information which they considered
would make the statement easier to use as a comparison tool with personal PPI
261
quotes. 144 Which? said that the key comparison information should be presented in a
standardized format in a summary box, while the OFT said that the statement should
contain prescribed information, with rules on format and prominence. Others considered that excessive standardization would make the implementation costs higher
and could decrease innovation. 145 JD Williams told us that the provision of cost per
£100 of benefit would be misleading in the context of retail PPI. 146 These parties said
that the CC should provide a framework detailing the type of information to be provided, rather than prescribe an exact layout or style as this would be more proportionate. 147 The FSA said that any statement should be consumer tested to ensure
that it would be effective.
10.325 We considered that a high level of standardization (ie the same information in the
same format) would increase the ability of consumers to compare products. We
noted that the Mortgage Conduct of Business review by BMRB for the FSA said that
the different formats for the disclosure documents, such as the key facts illustration,
can make it more difficult for consumers to undertake comparisons. 148
10.326 We decided that the format should be similar to that of the personal PPI quote (see
paragraphs 10.161 to 10.163), to aid comparability, and that the statement should
take substantially the form proposed in Appendix 10.2, but that consumer testing by
the CC should be taken into account, prior to the adoption of a final format into any
Order.
• Nil-balance accounts
10.327 Two retail PPI parties (JD Williams and SDGFS) told us that they did not provide
consumers with an annual retail credit statement because many consumers had nil
balances for the duration of the previous year. We were told that a requirement to
provide an annual PPI statement to consumers with dormant accounts would be very
costly and of little benefit as no PPI premiums were being paid. JD Williams told us
that a significant number of consumers who had balances with PPI in 2007 had nil
balances in December 2007 ([]). Express Gifts told us that many customers would
purchase over the Christmas season and pay their balance off over the next three to
six months.
10.328 We were told by a number of credit card PPI providers (Capital One, MBNA, Lloyds
TSB) that there were a large number of credit card consumers who also had nilbalance accounts during the course of a year. In these circumstances, the consumer
was not required to pay any PPI premium. We were also told by some parties that
consumers would be confused by annual statements that showed nil amounts of PPI
premium paid for the previous year.
10.329 We considered that there would be costs to PPI providers of providing an annual
statement in these circumstances which need not be incurred to ensure an effective
remedies package. We decided that annual PPI statements are not required to be
provided to consumers who have not paid any PPI premium in the previous 12
months.
144
See, for example, a large underwriter’s response to the Remedies Notice, pp8&9, Paymentshield response to the Remedies Notice, p3, Defaqto response to the Remedies Notice, p6, and HSBC response to the Remedies Notice, p2, paragraph 2.6. 145
See, for example, Paymentshield response to the Remedies Notice, p3. 146
JD Williams also made this point in paragraph 10.219 in relation to the discussion on the price metric. 147
HSBC response to the Remedies Notice, paragraph 6.7.2. MBNA response to the provisional decision on remedies, p15. 148
Consumers and mortgage disclosure documentation, September 2006, FSA, p9. 262
• Application of this requirement to the ‘back book’ of PPI policies
10.330 We noted a number of submissions (Barclays, MBNA, Lloyds TSB, Capital One,
RBSG) that annual statements should only be required of PPI policies sold after the
application of our Order. Several parties (Lloyds TSB, Capital One, Cardif Pinnacle,
RBSG) said that there would be significant additional costs if annual statements were
required to be provided to all new and existing PPI consumers, because, for
instance, the data could be held by third parties and/or under different IT
platforms. 149 We were told by MBNA that a remedy on the PPI back book would be
disproportionate on entities that had ceased to offer new PPI policies but had
retained existing PPI consumers for the remaining term of the PPI policies.
10.331 We reviewed the arguments for and against applying this remedy to the back-book of
PPI consumers. We considered that it will take significantly longer for the competition
benefits of this element of the remedies package to be realized if it does not apply to
the back-book of PPI consumers. We noted, however, that there would be a number
of additional practical difficulties and additional costs should this remedy be applied
to existing single-premium PPI policies 150 and that this would require additional
investment from distributors in relation to a type of PPI policy that would not be sold
in the future. Moreover, our analysis in paragraphs 5.63 to 5.75 indicates that there is
a currently a high switching cost associated with these policies that the provision of
an annual statement would not address. As we are prohibiting new single-premium
PPI policies (and several large suppliers of PLPPI have already stopped selling
single-premium policies—see paragraph 10.247), we would expect the number of
single-premium policies to decline relatively quickly after implementation of the order,
as providers move to regular-premium policies, as a result, or in anticipation of, the
order, or for other reasons. We decided, therefore, that this remedy should not apply
to the back-book of single-premium PPI policies. In relation to PLPPI, we noted that
PLPPI policies generally have a shorter life cycle than other forms of PPI and that an
exemption for existing single-premium PLPPI but not regular-premium PLPPI could
provide a disincentive for distributors to move to single premiums before they would
be required to do so by the Order implementing this report. This could create a conflict with the FSA’s publicly expressed concerns about single-premium PLPPI. We
therefore decided that this obligation should not apply to existing regular-premium
PLPPI policies. We considered that these factors were less important in relation to
existing regular-premium SMPPI policies.
10.332 We decided that annual statements should be required for existing MPPI, CCPPI and
retail PPI policies and for existing regular-premium SMPPI policies. For these policies, we decided that the first annual statement should be sent to customers within a
year of the commencement of this element of the remedies package and annually
thereafter.
149
RBSG said that if the requirement did apply to the back-book for PPI policies, then such a requirement should not start to run for a period of at least a year following the Order coming into force to allow for update of systems. 150
See, for example, Capital One response to the provisional decision on remedies, p19. 263
Implications of remedies package for providers of stand-alone PPI and shortterm IP 151
10.333 The remedies package we have set out will be most effective if providers of standalone PPI 152 and short-term IP are bound by some of the same requirements as
distributors of credit and PPI. We now review our decisions in relation to the requirements on providers of stand-alone PPI and short-term IP.
10.334 Providers of stand-alone PPI and short-term IP are required to provide all of the
following information in direct marketing materials (such as statement inserts, emails
and direct mail), but only (a) and (c) in regard to their non-direct marketing material
(such as newspaper and radio advertisements):
(a) the cost of PPI per £100 of monthly benefit (if the benefit pays out for less than
12 months; notice of this fact must also be clearly disclosed to consumers alongside the cost of the policy);
(b) that PPI is available from other firms (without specifying those other firms); and
(c) that information on PPI and short-term IP, alternative providers and other forms
of protection products can be found on the FSA’s moneymadeclear website.
10.335 Requiring the provision of this information will ensure that consumers can compare
effectively stand-alone PPI and short-term IP policies with PPI policies offered by
distributors, minimizing search costs for consumers.
10.336 Providers of stand-alone PPI and short-term IP are required to provide comparative
information to the FSA and other parties (see paragraph 10.224). This will minimize
search costs for consumers, providing one source of web-based information on
which consumers can compare all relevant policies. They are also required to provide
the range in which their claims ratio falls to any person on request (see paragraph
10.240).
10.337 To facilitate search and switching and for the point-of-sale prohibition to be most
effective, providers of stand-alone PPI and short-term IP are required to provide a
personal PPI quote to the consumer in a durable medium if the consumer asks the
provider about the cost and/or features of a stand-alone PPI and short-term IP policy
sold by that provider. This will enable consumers easily to compare the offers of
different providers. The personal PPI quote will take substantially the form proposed
in Appendix 10.1, subject to consumer testing by the CC, which will be taken into
account prior to the adoption of a final format into any Order.
10.338 Providers of stand-alone PPI and short-term IP are required to provide an annual
statement to ensure that consumers of stand-alone PPI and short-term IP providers
have, on a regular basis, the information required to facilitate switching. The
statement will take substantially the form proposed in Appendix 10.2, subject to consumer testing by the CC, which will be taken into account prior to the adoption of a
final format into any Order.
151
See paragraphs 2.12–2.14 and Appendix 2.3 with regard to short-term IP policies. We have found that short-term IP policies
sold as a result of a referral during the point of sale of a credit product are PPI products sold at the point of sale. As such, the
remedies package applies to providers of those short-term IP policies in the same way as for other providers of PPI at the point
of sale. In this section, reference to providers of short-term IP is a reference to those providers of short-term IP whose policies
are sold independently of the point of sale of a credit product.
152
As set out in paragraph 2.55, providers of stand-alone PPI includes distributors that offer PPI to insure repayments on credit
supplied by other distributors.
264
10.339 Finally, whilst there are no current providers of stand-alone PPI and short-term IP
policies that offer these policies on a single-premium basis, for the reasons set out in
paragraphs 10.243 to 10.277, stand-alone PPI and short-term IP providers are
restricted to selling only monthly or annual premium policies, and are prohibited from
making separate charges for administration or for the set-up or early termination of a
regular-premium PPI policy. This will ensure that consumers can search across
policies sold by distributors, intermediaries and stand-alone providers on a consistent
basis.
Options we decided not to pursue
10.340 In the Remedies Notice, the Supplementary Remedies Notice and during the
remedies process, we identified a number of other possible remedies which we
decided not to pursue. The reasons for these decisions are set out below.
Further standardization of PPI information given to the consumer at the point of sale
10.341 We considered whether standard information, additional to that already provided,
should be given to consumers at the point of sale.
Views of parties
10.342 Many parties agreed that additional standard information would make the comparison
of different providers’ products and prices easier for consumers. For example, Capital
One was supportive of increased standardized disclosure for PPI products as it considered that it would reduce the risk of providers/distributors misinterpreting principlebased regulation and would improve comparability for consumers. It was put to us
that the message that PPI could be bought from other parties was particularly important. However, a number of other parties did not consider that additional information
and standard forms at the point of sale would address the AEC and had the potential
to cost more, confuse consumers through additional paperwork and reduce provider
innovation in the PPI market. 153 The FSA said that it was unclear what additional
benefits more cost information would provide and that it did not think that further
standardization of documents would create significant additional benefits within the
current market structure.
10.343 JD Williams and SDGFS told us that additional standard information would not be
necessary or appropriate in the context of a revolving credit product such as retail
PPI because it would not be helpful or meaningful for the customer to be provided
with information that was related to a static credit amount.
Our views
10.344 We decided not to require additional standard information to be provided at the point
of sale, although we have required that information is provided in a standardized way
in marketing materials, the personal PPI quote and the annual statement. We considered that solely providing more information at the point of sale was unlikely to be
effective at addressing the AEC and more fundamental change was required to
impact sufficiently customer and supplier behaviour.
153
See, for example, AFB response to the Remedies Notice, p4.
265
All PPI policies to be renewed annually
10.345 We considered the appropriateness of an obligation to renew PPI policies annually
and an associated requirement that customers opt in to obtain continuing PPI cover
for the next year.
Views of parties
10.346 Parties told us that a requirement to renew policies annually on an opt-in basis:
would result in consumers not being covered when they thought they were, as many
would ignore the renewal documents; was different to most other insurance products
which are sold on an opt-out basis; and would be expensive to introduce as it would
be an additional contact with customers which required a change in policy terms and
provider systems. 154 Some parties interpreted this requirement as requiring PPI
policies to be sold with annual premiums, and we were told that this would prevent
the provision of regular-premium policies which allowed consumers to cancel and
switch at any time. 155 Lloyds TSB proposed, as an alternative, a requirement to
renew policies annually on an opt-out basis, which, it said, would address the risk of
PPI cover lapsing accidentally (see paragraphs 10.67(b) and 10.68(b)).
Our views
10.347 We decided not to require PPI policies to be renewed annually on an opt-in basis as
we considered that this would introduce a risk of PPI cover lapsing accidentally, and
that a requirement annually to renew policies would be more onerous on parties than
the package of remedies that we have decided to implement. We decided that the
package of remedies, and in particular the provision of an annual statement in conjunction with the other elements of the remedies package, would be effective in
providing sufficient information and impetus to encourage consumers to consider
whether they should switch, and that a more intrusive measure was not needed to
remedy this aspect of the AEC. This consideration also applies to the proposal by
Lloyds TSB to require an annual renewal on an opt-out basis, which we therefore
decided not to include in our remedies package.
Alternative remedies to a prohibition of single-premium PPI policies
10.348 We considered whether two other remedies might address the AEC instead of prohibiting single-premium PPI policies: requiring equivalent cover regular-premium PPI
policies to be offered alongside any premium policy and/or requiring that any rebate
of a single-premium policy is provided on a pro-rata basis.
Parties’ views on providing an equivalent cover regular-premium policy
10.349 We received a range of views from parties on the appropriateness of requiring the
provision of an equivalent cover regular-premium policy. Some parties considered
that it would provide customers with additional choice. 156 Other parties considered
that it would lead to some firms exiting the PPI market because it would become an
154
See, for example, a large underwriter’s response to the Remedies Notice, p7, MBNA response to the Remedies Notice, paragraph 4.25, Barclays response to the Remedies Notice, p20, and MBNA response to the Remedies Notice, paragraph 4.26, ABI response to the Remedies Notice, p8, and HSBC response to the Remedies Notice, paragraph 6.7.1. 155
See, for example, JD Williams response to the Retail PPI Notice of Remedies, pp8&9. 156
See, for example, Assurant response to the Remedies Notice, p4, and Genworth response to the Remedies Notice, paragraph 2.29; London General Insurance response to the provisional decision on remedies, p2, and Axa response to the provisional decision on remedies, p4. 266
advised sale, which would increase costs leading to some providers leaving the
market, and had greater potential for customer confusion. 157 One party ([]) provided evidence of an unsuccessful pilot in 2006 where it sold both single- and
regular-premium policies in some of its branches. 158
Our views on providing an equivalent cover regular-premium policy
10.350 We considered that a requirement to provide an equivalent cover regular-premium
policy would not address adequately the switching barrier associated with singlepremium policies. We noted the evidence relating to customer confusion about
single- and regular-premium policies and we considered that it would be likely that
many customers would seek the advice of the salesperson in these circumstances.
We considered that the incentives for the salesperson on which product to recommend would be complex and it would be impractical to monitor and enforce without a
significant amount of resource.
Parties’ views on pro-rata rebates of single-premium policies
10.351 All parties were in favour of a fair rebate for customers who cancelled their singlepremium PPI policies. However, parties had different views as to what they considered to be fair to the consumer and to PPI providers. Some parties told us that
rebates were already fair and that there was therefore little need to change current
arrangements significantly. 159 Aviva told us that pro-rata rebates were an incompatible concept with single-premium policies as a rebate in this form ignored the risk
profile of single-premium policies. 160 Other parties considered that an obligation to
provide a pro-rata rebate would be equally effective as a prohibition of singlepremium policies in addressing the switching AEC. 161 These parties considered that
a pro-rata rebate was fair to consumers and more proportionate than prohibiting
single-premium policies as the former would not involve reducing consumer choice or
restricting a firm’s product line. 162 However, one large distributor ([]) said that it did
not believe that pro-rata rebates would be sufficient to remedy the adverse effects
identified by the CC associated with single-premium policies.
10.352 The FSA suggested that we imposed pro-rata rebates for the period after an Order
was made and before the single-premium prohibition came into force. We considered
that this would be a disproportionate cost and complication to the prohibition of single
premiums. We noted in paragraph 10.263 that we do not propose to extend the
single-premium remedy to the ‘back-book’ of single-premium PPI. For similar
reasons, we do not consider that an interim remedy of the type suggested by the
FSA is required in this instance.
157
See, for example, Abbey response to the Remedies Notice, paragraph 5.3.3, HSBC response to the Remedies Notice, p14.
The report noted that widescale confusion existed as customers did not understand the difference between single- and
regular-premium policies.
159
For example, MBNA noted that rebates are already governed by FSA ‘fairness’ rules; see MBNA response to the Remedies
Notice, paragraph 4.32.
160
Aviva said that a single-premium product could not have a pro-rata rebate because any particular refund would be distributed
according to the risk cover provided until the time of cancellation. It said that the imposition of regular-premium policies, or
single-premium policies with pro-rata rebates, would encourage insurers to reduce the duration and level of cover, as when a
policyholder leaves the insurance pool mid-term this will affect the risk profile of the insurance pool overall (this point assumes
that risk is greatest at the beginning of the credit agreement as no amount has yet been paid back). However, we also noted
that at least one party (Nationwide) has previously provided pro-rata rebates with a single-premium PLPPI product.
161
See, for example, Abbey response to the Remedies Notice, pp14&15, and Nationwide response to the Remedies Notice,
pp13&14.
162
See, for example, Nationwide response to the provisional decision on remedies, p5, and RBSG response to the provisional
decision on remedies, p12.
158
267
Our views on pro-rata rebates of single-premium policies
10.353 We considered that the provision of pro-rata rebates on single-premium policies
would only be partially effective in addressing the switching barrier that we found. We
came to this conclusion for three reasons:
(a) It would remove some, but not all, of the financial costs of switching to a regularpremium policy. In particular, a switching cost associated with the rebate on the
interest on the single premium would remain.
(b) It would be likely to require a high degree of financial sophistication on the part of
consumers to calculate the cost of switching from single-premium PPI (albeit with
a pro-rata rebate) and we were not satisfied that enough consumers possess the
required degree of financial sophistication.
(c) It would not contribute to addressing the barriers to search arising from product
complexity and variations in pricing structures. The continued provision of singlepremium PPI policies would make it harder to develop a ‘common currency’ for
disclosing the price of PPI on a comparable basis, inhibiting the effectiveness of
other remedies such as the provision of ‘key messages’ and information to the
FSA comparative website.
10.354 We noted that, in addition to switching costs associated with low rebates, consumers
may also face switching costs arising from administration charges. We considered
that we could have prohibited administration charges from being imposed. We also
considered that we could have required providers to give consumers the option to
obtain their rebate as a cash payment. However, our analysis of switching costs in
Appendix 10.3 suggested that material switching costs would remain because of the
interest that is payable on a single-premium policy. 163 We considered that these
remaining switching costs would act as a disincentive for at least some consumers,
which would reduce the effectiveness of the remedies package.
10.355 We also noted that some consumers may wish to switch both credit and PPI at the
same time, rather than PPI on a stand-alone basis. Our analysis of pro-rata rebates
indicates that the costs of terminating the PPI early would not act as a disincentive to
switching the combination of credit with PPI (this was because the rates charged for
shorter-term single-premium PPI tend to be lower than the rates for longer-term PPI).
However, we considered that for switching to drive competition, it is important to
remove barriers to switching both on the combination of credit with PPI and on a
stand-alone basis.
10.356 In this context, we reviewed instances where consumers face barriers to switching
which are associated with the provision of credit rather than the PPI in a combination
of credit and PPI—that is, although the cost of PPI may be lower when switching to a
new provider, the cost of the new bundle of credit and PPI may be higher in total
(because, for example, the APR of the alternative credit product was higher). In other
situations, the customer may not be able to get credit from an alternative lender. We
concluded that in such instances, switching to a stand-alone PPI provider may represent the consumer’s only or best opportunity for obtaining better-value PPI. In these
instances, the provision or otherwise of pro-rata rebates would not address the
163
We noted that it was not possible for us to address directly the interest element of switching costs because it relates to the
rebate on credit rather than PPI and this allowed the possibility of a significant switching risk (for example, if the credit APR was
increased with the provision of a single-premium policy). Our analysis of switching costs also showed that a pro-rata rebate was
insufficient to buy an equivalent regular-premium PPI policy but, in the absence of administrative charges, was usually sufficient
to buy an equivalent single-premium policy for the remainder of the loan.
268
switching costs that are incurred on the combination of credit and PPI. We noted that
stand-alone PPI is only offered on a regular-premium basis and we have not heard
evidence that any parties would intend to offer stand-alone PPI on a single-premium
basis.
10.357 We concluded that a remedy which did not remove all costs of switching to regularpremium PPI would not fully address the barrier to switching associated with singlepremium PPI.
10.358 As well as looking at the impact on the financial disincentive to switch, we also considered whether there was likely to be any qualitative difference between a prohibition on selling single-premium policies and an obligation to provide pro-rata rebates.
We noted that, in the absence of an explicit charge for cancellation, it would be clear
to consumers that they could switch from regular-premium policy without incurring a
financial cost. In contrast, it would not always be the case that a pro-rata rebate on a
single-premium policy would allow a consumer to switch without financial cost. In any
event, a consumer would require a high degree of financial sophistication to calculate
the correct amount that they would receive on termination of a single-premium PPI
policy. 164 We decided that the inherent simplicity of regular-premium policies for
consumers (when understanding the cost and therefore the benefit of terminating a
regular-premium policy) contributed to the effectiveness of prohibiting single-premium
policies compared with a remedy that required pro-rata rebates to be provided. We
considered that this aspect of simplicity also contributed to the effectiveness of the
package of remedies and, in particular, the personal PPI quote, the requirement to
provide information in marketing material and for use in the FSA comparative
website.
Minimum standards for elements of PPI policies that act as a barrier to switching
10.359 We considered whether it would be appropriate or effective to require minimum
standards in relation to those elements of PPI policies (ie exclusions and qualification
periods for pre-existing conditions—see paragraph 5.39) that act as a barrier to
switching.
Views of the parties
10.360 Some parties were in favour of imposing minimum standards because, in the event of
switching, it would improve a consumer’s ‘peace of mind’ that there was no gap in
PPI cover. Other parties said that there would be cost implications, possibly resulting
in fewer PPI providers, because of a need individually to underwrite PPI policies or
because adverse selection issues would increase. 165 We were told that stand-alone
providers relied on these exclusions to protect themselves from adverse selection
issues when accepting a new consumer. A few parties suggested that a better option
would be for PPI providers to offer to remove or lessen exclusions and pre-existing
conditions for switching consumers only when those consumers switched to an
equivalent or lower-specified PPI policy. 166
164
The calculation would be a combination of the rebate on the PPI and the interest that had not been incurred on the loan funding the PPI premium. Each of these aspects would also be calculated on a different basis. 165
See, for example, ABI response to the Remedies Notice, p13, HSBC response to the Remedies Notice, p14, JD Williams
response to the Retail PPI Remedies Notice, p10, and SDGFS response to the Remedies Notice, p15, paragraph 5.10. 166
See AXA response to the Remedies Notice, paragraph 64, and Cassidy Davis response to the Remedies Notice, p6. 269
10.361 We were told that minimum standards or standard terminology would generally make
comparison between products easier, for example as currently existed for MPPI. 167
However, other parties said that any minimum standards would or could lead to less
product differentiation and maybe less switching as a result, and that minimum standards would limit innovation. 168 JD Williams told us that minimum standards could
reduce consumer choice by preventing consumers selecting products with longer
exclusion and/or pre-qualification periods in return for lower premiums.
Our views
10.362 We considered that having minimum standards for exclusions and qualification
periods for pre-existing conditions would not be necessary in a market where suppliers actively compete for switching customers. In a competitive market we expect
that suppliers would waive or lower qualification periods in order to win customers.
We concluded that the imposition of minimum standards is not necessary to address
the AEC that we have found, given that we expect our remedies package to be
effective in stimulating competition.
Obligation to share information about customer claims
10.363 We considered whether a requirement to share claims data on consumers might
encourage switching by enabling providers to offer a discount to customers who have
made few or no claims.
Views of the parties
10.364 Most parties were against any obligation to share information about consumer claims
because it would be costly and was not a certain indicator of a particular consumer’s
likelihood of making a future claim. 169 The ABI said that some consumers with a
history of making claims might find it more difficult or consider it more difficult to
switch. 170 Some parties were concerned that a database to share claims information
would be difficult to set up given the sensitivity of the data and the requirements of
the Data Protection Act. 171 However, a few parties suggested that consumer claims
information could reduce adverse selection issues for stand-alone providers. Some
parties (Cardif, Barclays, PaymentShield and MBNA) suggested that claims information could instead be added to the annual statement and could be provided to a
consumer on request.
167
See, for example, Aviva response to the Remedies Notice, p25, Coventry Building Society response to the Remedies Notice,
p9, Claim2Gain response to the Remedies Notice, p3, Barclays response to the Remedies Notice, p31, and Defaqto response
to the Remedies Notice, p14. Minimum standards in this context are referred to in Annex A of MPPI: Response by the Council
of Mortgage Lenders to the OFT Report on the Payment Protection Insurance Market Study, 2006.
168
See, for example, Barclays response to the Remedies Notice, p31, Openwork response to the Remedies Notice, p2
(regarding MPPI), and JD Williams response to the Retail PPI Remedies Notice, p10.
169
See, for example, FLA response to the Remedies Notice, p31, HSBC response to the Remedies Notice, paragraph 6.7.5,
and JD Williams response to the Retail PPI Remedies Notice, p10. However, evidence from the parties was mixed on this point
with [] and [] saying that they were predictive. [] said that its MPPI customers (from 2002 to 2006) who claimed were
three times more likely to claim again (7.2 per cent a year vs 2.4 per cent a year). [] said that ‘we may reject a null hypothesis
of no difference in rate of a successful claim in favour of the alternative that there is an increase in the rate of successful claim
amongst those policies claiming more than once‘, while Abbey, AXA and MBNA said that they were not.
170
See ABI response to the Remedies Notice, p14.
171
See, for example, ABI response to the Remedies Notice, p14, and Cassidy Davis response to the Remedies Notice, p6, and
JD Williams response to the Retail PPI Remedies Notice. We note that we were told by one CRA (Experian) that the current
cost of PPI might make it uneconomic for such a system to be set up.
270
Our views
10.365 In paragraph 10.322, we stated our reasons for not requiring that consumer claims
information be included in the annual statement to be provided to consumers. We
also considered the possibility of creating a new database to arrange for the transfer
of consumer claims information between underwriters. We thought that, in some
circumstances, the existence of such a database might help some consumers obtain
better terms when switching—by enabling them to demonstrate their recent claims
history to a new provider. However, we noted that it could make it harder or more
costly for other consumers to switch, for example consumers who have recently
made a claim. We did not consider that development of such a database, which we
thought would be a costly and complex exercise, was necessary for the remedies
package to be effective and we therefore concluded that it should not be included.
Obligation to share information about consumers’ credit card and retail credit
balance with a nominated underwriter
10.366 We considered whether an obligation to share information about a consumer’s credit
card and/or retail credit balance would encourage more tailored stand-alone PPI
products to be developed for consumers.
Views of the parties
10.367 Nearly all parties, including all stand-alone providers, said that they would not be
interested in credit card balance data because it would be difficult and costly to set
up, would expose consumers to a greater risk of fraud and would put the company
that collected the data in a very strong position. 172 Only one party, [], suggested
that it might use this data and/or CRA data to produce a more bespoke product.
However, we were also told that there were a number of technical issues with
requiring CRA data to be shared. 173
10.368 Two stand-alone providers (Paymentshield and the Post Office) said that they would
prefer to offer an income protection product which covered the aggregate credit card
repayments, and as a result some said that that the average annual balance and
minimum repayment amount would be useful, while one said that customers were
aware of their average outstanding balance and did not require this information.
However, it was also suggested that most consumers were aware of their average
outstanding balance and providers would not require the information to be supplied
by the credit provider. 174
172
See, for example, Capital One response to the Remedies Notice, pp6&7, and MBNA response to the Remedies Notice,
pp11&12.
173
APACS said that there were a number of issues which would need to be addressed, including: the governance rules; user
validation; liability for errors; agreement of messaging standards; recovery of costs; and how disputed transactions would be
treated. The CRAs ([] and []) said that the option would be costly, that the data they currently held on credit cards could
not be used for this purpose under the Principles of Reciprocity and that they did not currently hold full data on all credit card
accounts ([] said that it had 60 million accounts but estimated that there were at least 5 million which still needed to be
added). The Principles of Reciprocity is an agreement between the subscribers and the holders of negative, positive and
search data. The agreement regulates the access to data and the use to which subscribers can put data. The body that revises
Principles of Reciprocity agreements is the Steering Committee on Reciprocity (SCOR). SCOR is made up of a number of trade
associations as well as CRAs (Equifax, Experian and Callcredit). Recent joiners include the Consumer Credit Association and
the Credit Services Association.
174
For example, SDGFS told us that it already shared the credit balance with the customer through the monthly statement. It
also told us that the balances on its customers’ home shopping accounts were relatively small compared with credit card
balances and that being able to track the balance was not so important in terms of providing insurance. The customer was
buying a limited amount of security from retail PPI due to the low balance and if the customer wanted security in terms of lifestyle protection then they could put in place other insurance such as income or bill protection.
271
10.369 JD Williams told us that calculating this information for revolving credit was difficult
because credit balances could change on a daily basis. It suggested that stand-alone
providers could better target their PPI products by developing a banded IP product
which would insure the customer balances up to a specified level of cover, and this
would avoid the practical and legal difficulties of supplying credit balance information.
Our views
10.370 We considered that the provision of this information could, in theory, allow businesses to create more tailored products for stand-alone CCPPI and for retail PPI that
tracked a consumer’s actual outstanding balance. However, given the lack of interest
in gaining access to this data, we were not convinced that in practice the information
would be used in this way, and therefore we considered that it was unlikely to be
effective. We also considered whether to require CRA data to be shared but note that
under the Principles of Reciprocity (see first footnote to paragraph 10.367) it would
require both industry and CRA agreement which would make the implementation of
this measure uncertain. We are requiring average balance data to be included on the
annual statement provided to consumers in relation to CCPPI and retail PPI (see
paragraph 10.320), which we consider to be a more effective and proportionate
means of facilitating stand-alone competition in the provision of CCPPI and retail PPI.
Price caps
10.371 We considered the appropriateness of imposing price caps on PPI policies.
Views of the parties
10.372 Three parties (Citizens Advice, the Post Office and []) were in favour of either
temporary or long-term price caps on the basis that the scale of the excess profits
justified the imposition of a price cap. Other respondents did not agree with the
imposition of price caps. Their arguments fell into five main areas:
(a) they would not address the competition issues that the CC had provisionally
found; 175
(b) they were disproportionate and inappropriate, given the absence of a consumer
detriment in the overall markets for PPI and the underlying credit products; 176
(c) they would have other negative effects on the market; for example, they would
decrease innovation and reduce quality, could lead to some consumers, particularly poorer or older consumers, not being offered PPI or some PPI providers
exiting the market because of insufficient profit margins; 177
(d) they would lessen the effectiveness of other remedies; for example, they could
increase the point-of-sale advantage, as consumers could be discouraged from
shopping around as they would assume that they already had a good price; 178
and
175
See, for example, ABI submission covering letter, p2, and Aviva response to the Remedies Notice, p30. See, for example, RBSG response to the Remedies Notice, p17, and SDGFS response to the Retail PPI Remedies Notice, p23.
177
See, for example, Barclays response to the Remedies Notice, p37, and Paymentshield response to the Remedies Notice, p4, JD Williams response to the Retail PPI Remedies Notice, p11, and SDGFS response to the Retail PPI Remedies Notice, p23. 178
See, for example, ABI response to the Remedies Notice, p16, and Axa response to the Remedies Notice, paragraph 84. 176
272
(e) they would be complex to set up and difficult to enforce given the range of PPI
policies in the market, the various risk rating of consumers, underwriting variables
and the lack of any minimum standards. 179
Our views
10.373 We believe that price caps could address the customer detriment of higher prices
and we have not been persuaded by the evidence that price caps would have negative impacts on competition. However, we consider that the package of remedies we
have decided to implement will address the AEC that we have identified in a timely
manner and we do not have to address the customer detriment shown in higher
prices resulting from the AEC. 180 We consider that by addressing the AEC with the
package of remedies, which we have decided on, this aspect of the customer detriment will also be addressed.
Relevant customer benefits
10.374 In this section, we consider whether there are any relevant customer benefits arising
from the adverse features that we have identified.
10.375 In deciding the question of remedies, as noted in paragraph 10.9 the CC may have
regard to the effects of any remedial action on any relevant customer benefits arising
from the adverse feature or features of the market concerned. 181 If it considers that
such relevant customer benefits exist, the CC will consider whether to modify the
remedy that it might otherwise have imposed or recommended; although the consideration of any relevant customer benefits does not involve weighing the benefits
against the AEC and any detrimental effect on consumers. 182 Similarly, the CC will
disregard any benefits that might arise from commitments that the parties may wish
to offer but that do not meet the criteria of a relevant customer benefit. 183 The same
principle would apply to benefits that arise only as a result of remedies imposed by
the CC through its order-making powers. As these benefits result from the remedy,
and not from the features that have been identified, they cannot be considered relevant customer benefits within the meaning of the Act.
10.376 If the CC is satisfied that there are relevant customer benefits deriving from a market
feature that also has adverse effects on competition, when deciding whether to
modify a remedy, the CC will consider a number of factors including the size and
nature of the expected benefit, how long the benefit is to be sustained and the impact
of the benefit on different customers.
10.377 We set out the possible relevant customer benefits put to us, and our views on them,
below:
(a) First, we consider the potential customer benefits arising in PPI markets from
credit providers having a point-of-sale advantage.
179
See, for example, Abbey response to the Remedies Notice, paragraph 6.1.2, Lloyds TSB response to the Remedies Notice, paragraph 7.3, JD Williams response to the Retail PPI Remedies Notice, p11, and SDGFS response to the Retail PPI Remedies Notice, p23.
180
CC3, paragraph 4.22. 181
CC3, paragraphs 4.26–4.41.
182
The CC’s approach to the assessment of relevant customer benefits is discussed in more detail in CC3, paragraph 4.26 and following. 183
CC3, paragraph 4.28. 273
(b) We then consider the potential benefits for PPI policyholders relating to singlepremium policies.
(c) We next consider whether there is a relevant customer benefit associated with
credit providers knowing that a consumer is taking out PPI.
(d) Finally, we consider the potential benefits arising in the credit market.
10.378 Parties mentioned in their responses to the Remedies Notice, the Supplementary
Remedies Notice and in further correspondence various potential harms to consumers that might arise if we imposed particular remedies. These generally related to
the cost of implementing each remedy. We did not think that these were relevant
customer benefits as defined in the Act, but considered the evidence put to us in the
context of each individual remedy and the overall package.
10.379 We consider the effect of each of the alleged relevant customer benefits that were
put to us, in the context of the remedies package, in paragraphs 10.380 to 10.464.
Potential customer benefits of the point-of-sale advantage
Lower PPI prices
10.380 Parties put forward a number of reasons why, all else being equal, distributors may
have incentives to reduce their PPI prices when sold at the point of sale. The key
reasons put forward were:
(a) selling PPI and credit together eliminates double marginalization; 184
(b) selling PPI at the point of sale enables firms to realize economies of scope and
thereby reduce costs; 185 and
(c) selling PPI at the point of sale increases the volume of sales, allowing firms to
realize economies of scale and thereby reduce costs.
10.381 Following the publication of our provisional decision on remedies, some of the parties
provided further reasoning as to why the point-of-sale advantage may result in
efficiencies for distributors and for underwriters. For example, Aviva told us that
‘when customers seek insurance on a standalone basis, separate from the supply of
credit, they are more likely to be a higher risk of making a claim’. 186 If this is the case,
then selling PPI at the point of sale would result in a lower-risk pool of consumers
and therefore lower claims than would be the case away from the point of sale. Aviva
said that this would result in relevant customer benefits in the form of lower prices
and/or higher-quality products as a result of the point-of-sale feature.
10.382 Capital One told us that the point of sale had the highest marketing efficiency of any
distribution channel and that: the point of sale enabled distributors to achieve 100 per
cent reach of new credit customers; it was the point in the process at which customers were most engaged in the credit process; and selling at the point of sale (as
184
Double marginalization occurs where separate firms sell complementary products. Each firm sets prices with regard to the
effect of its own prices on its own sales. However, as the products are complements, the prices that one firm sets will affect the
sales of the other, as well as its own sales. This leads separate firms to set higher prices. When prices are by a single firm, the
effects of the pricing of one product on its complement can be taken into account, and the result is lower prices.
185
The economies of scope put to us included: the ability to save costs by operating a combined sales process, the effect of the
point of sale on adverse selection, the marketing efficiency of selling PPI at the point of sale and the ability to offer PPI products
tailored to the underlying credit product.
186
Aviva response to the provisional decision on remedies, p4, paragraph 2.2.7.
274
opposed to at the point of activation or subsequently) had a near zero cost per sale.
Capital One said that these efficiencies resulted in increased demand for PPI, as well
as lower costs for distributors.
10.383 We concluded that it was unlikely that these claimed efficiencies could constitute
relevant customer benefits, within the meaning of the Act, for two reasons.
10.384 First, given the lack of a competitive threat in the markets for PPI, it appears to us
that these benefits are not passed on to consumers to a sufficient extent to result in
lower PPI prices. 187 In light of our analysis of profitability of PPI policies, which
showed that distributors representing a substantial part of the share of supply of PPI
have earned profits that were persistently and substantially in excess of the cost of
capital and led us to conclude that PPI prices are higher than they would be in a wellfunctioning market (see paragraphs 4.79 to 4.82), we concluded that consumers
were not seeing any benefit in the form of lower PPI prices.
10.385 Lloyds TSB stated 188 that we should take into account customer benefits which did
not presently accrue to consumers, but which could be expected to accrue in the
future. It stated that if the CC were to implement effective remedies to promote
‘system competition’ and/or ‘renewals competition’ (but without implementing a pointof-sale prohibition), then it was to be expected that at least a substantial element of
the costs savings associated with the point-of-sale advantage would feed through
into lower PPI prices within a reasonably short time.
10.386 Whilst it is correct that the Act includes benefits to future customers within the scope
of the definition of relevant customer benefits, 189 it must be the case that the benefits
arise as a result of the feature(s) identified as giving rise to the AEC and would not
accrue, or would be unlikely to accrue, in the absence of those features. In this case,
we have concluded that the alleged benefits do not currently accrue to consumers
due to lack of effective competition in the markets and we have no expectation that
any such benefits would accrue in the future in the absence of a remedies package
that stimulates effective competition. We consider that Lloyds TSB’s suggestion
would require us to assess potential relevant customer benefits to future customers
in a market where a partial remedies package has been implemented. We do not
think that this is compatible with the statutory test, nor with the CC’s guidance on this
issue. We consider that the issue is more appropriately considered in the context of
the assessment of alternative remedies proposals (including the proposal put forward
by Lloyds TSB) and whether these proposals would be sufficient to adequately
address the AEC that we have identified. Accordingly, we consider Lloyds TSB’s
alternative remedies proposal in paragraph 10.68(b).
10.387 However, regardless of whether Lloyds TSB’s suggestions are considered in the
context of preservation of alleged customer benefits for future customers or in the
context of an assessment of the effectiveness of alternative remedies proposals, our
conclusion on the necessary scope of the remedies package would be unchanged.
For the reasons set out in paragraphs 10.466 to 10.476, we consider that all the
elements of the remedies package are necessary in order to provide as comprehensive a solution as is reasonable and practicable to the AEC that we have identified. As
set out in paragraph 10.68, we have concluded that the alternative remedies proposal suggested by Lloyds TSB would not form an effective remedy solution to the
187
This logic will also apply to other ‘non-price’ relevant customer benefits put forward by the parties. In other words, in the absence of a competitive threat, there would be little incentive to pass on any efficiency benefits which accrue to the distributor or underwriter as a result of the point of sale, whether that is in the form of lower PPI prices, by offering more generous policy
terms (see paragraph 10.401) or some combination of lower PPI prices and more generous policy terms. 188
Lloyds TSB response to the provisional decision on remedies, paragraph 1.11. 189
Section 134(8) of the Act. 275
AEC that we have identified. In the absence of a remedies package that stimulates
effective competition, we can have no expectation that the relevant customer benefits
that Lloyds TSB alleges arise from the selling at the point of sale would accrue to
consumers in the future.
10.388 Second, it does not appear that selling PPI at the same time as credit is being sold is
strictly necessary to realize all of these efficiencies—some are also capable of being
realized, to an appreciable level, by selling after a suitable interval. For example, if
PPI is sold after a short interval, distributors can still identify low-risk credit consumers and so limit their exposure to adverse selection risks (we discuss this potential efficiency in more detail in paragraphs 10.427 to 10.441) and they can still enjoy
marketing efficiencies as a result of promoting PPI at the credit point of sale (paragraph 5.107).
10.389 In retail PPI, we note, in particular, that JD Williams does not sell PPI at the initial
credit point of sale but chooses to sell PPI after a suitable interval and appears to do
so successfully. We noted that SDGFS offers PPI not only at the initial credit point of
sale but also to existing consumers when they place subsequent orders on their
account over the telephone. Telephone consumers will not be offered PPI more
frequently than every [three to nine] months. We also noted that SDGFS has recently
started making outbound calls to existing consumers.
10.390 We therefore concluded that lower PPI prices arising as a result of potential cost
savings at the point of sale is not a relevant customer benefit. We consider whether
any benefits are passed on in the form of lower credit prices in paragraphs 10.480 to
10.492.
Increased consumer choice
10.391 Some parties told us that remedies to address the point-of-sale advantage would
reduce consumer choice as they would remove the option from consumers of
purchasing PPI at the point of sale. We were also told that remedies to address the
point-of-sale advantage could lead some distributors and intermediaries to exit the
market altogether and that this would result in a reduction in choice.
10.392 We note that the relevant question, for the purposes of evaluating whether the pointof-sale advantage results in a relevant customer benefit, is whether the point-of-sale
advantage delivers a benefit to consumers in the form of increased choice.
10.393 We do not believe that the point-of-sale advantage results in increased choice for
consumers. Indeed in paragraphs 5.116 to 5.119 we found that the reverse was true
and that the point-of-sale advantage acted as a barrier to other PPI providers competing for customers of any given distributor or intermediary. We found that because
of this lack of competition, the only choice offered to most consumers who wished to
purchase PPI was to purchase PPI from their credit supplier.
10.394 We recognize that remedies to address the point-of-sale advantage may result in
some restriction in consumers’ purchasing freedom, in that there would be a restriction in buying PPI from the credit supplier. We also recognize that some distributors
and intermediaries may decide to exit the market, and we considered this possibility
in paragraphs 10.54 to 10.57. We believe that, by enabling more suppliers to compete for consumers’ business, these remedies will increase consumer choice overall
and that, in our judgement, the restriction in consumers’ purchasing freedom implied
by our remedies package, and in particular the point-of-sale prohibition is necessary
in order to remedy the significant adverse effects on competition identified.
276
Better-quality regular-premium PPI products
10.395 Some parties told us that policies sold at the point of sale were generally of higher
quality than those sold on a stand-alone basis. For example, Aviva told us that
policies sold at the point of sale were capable of providing more exhaustive insurance cover than stand-alone policies. We considered this separately for regularpremium policies and single-premium policies (see paragraphs 10.405 to 10.426 for
our analysis of whether single-premium policies result in lower-priced or higherquality products).
10.396 We note that, in order to find that the point of sale results in a relevant customer
benefit in the form of higher-quality policies, it is not sufficient to show simply that
policies sold at the point of sale are of a higher quality than those sold on a standalone basis. Even if it can be shown that policies sold at the point of sale are of
higher quality, in order to find a relevant customer benefit, it must also be shown that
the suggested benefit (higher quality) has occurred as a result (whether wholly or
partly) of a feature (namely the point-of-sale advantage) and further that this benefit
would be unlikely to accrue without the feature concerned. 190 We assessed the
evidence on whether these criteria were met.
10.397 We looked at the price and quality of PPI policies sold at the point of sale and on a
stand-alone basis, using Defaqto scores as a measure of quality. 191 We did this to
see whether stand-alone policies were in fact of lower quality when compared with
policies sold at a credit point of sale. Our analysis of this is set out in Appendix 10.4.
Figure 10.8 shows the results of this analysis for MPPI policies. 192 We found that,
once differences in price are accounted for, there is no evidence that stand-alone
policies are more highly priced (for a given level of quality), or offer lower quality (for
a given price), than policies offered at the point of sale. We also found that standalone policies offer a wide range of product quality, and that some were of higher
quality than most policies sold at the point of sale of the mortgage.
190
See section 134(8) of the Act. See Appendix 4.2 for a fuller description of the Defaqto data and our analysis of prices and quality based on this data. We chose to compare MPPI policies because both point-of-sale and stand-alone MPPI policies are charged on a regular-
premium basis. Stand-alone PLPPI and point-of-sale PLPPI are more difficult to compare because point-of-sale PLPPI is typically charged as a single-premium policy whereas stand-alone PLPPI tends to be regular-premium policies. 191
192
277
FIGURE 10.8 Comparison of the monthly price per £100 of cover for a 35-year-old with Defaqto rating scores, using the 2008 Defaqto rating methodology,
for stand-alone and non-stand-alone MPPI policies (wide classification) Price of PPI (£ per £100 cover)
6
4
2
0
20
40
60
80
100
Defaqto score
Other
Stand-alone
Source: Defaqto Aequos database December 2008 and CC analysis.
10.398 Table 10.2 provides a comparison of some high-quality stand-alone products with
some policies offered by large distributors at the point of sale. This again shows that
stand-alone policies are not necessarily of inferior quality.
TABLE 10.2 Comparison of the quality of some MPPI policies sold by large distributors at the point of sale with some
policies sold on a stand-alone basis
Provider
Best Insurance (British
Insurance)
NatWest
Nationwide Building
Society
Abbey
ASU direct
Halifax
Paymentcare Ltd
Barclays
Churchill
Iprotect
Product
Type of supplier
Defaqto Rating
Score*
Implied Defaqto
Star Rating†
British Mortgage Payment Protection (ASU)
Stand-alone
100
Not rated
Mortgage Repayments Protector
Mortgage Payment Protection (ASU 30 days/
12 months)
Paymentcare (ASU)
Payment Protection Insurance ASU (30
BTDO)
Mortgage Repayment Cover (MRC—ASU)
Low Rate (ASU)
Mortgagecare ASU
Income Protection
Mortgage Payment Protection ASU 30 days
Point of sale
Point of sale
96
93
5
5
Point of sale
Stand-alone
90
89
5
4
Point of sale
Stand-alone
Point of sale
Stand-alone
Stand-alone
89
89
84
84
82
4
Not rated
3
3
3
Source: Defaqto Aequos database as of 10 December 2008 and CC analysis.
*Defaqto Rating Scores calculated by the CC, based on the 24 2008 Defaqto MPPI rating criteria. †Implied Defaqto Star Ratings calculated by the CC in the cases where Defaqto has indicated that a product is included in its
Star Ratings scheme.
278
10.399 If the point-of-sale advantage gave rise to higher product quality, we may expect that
there would be either no high-quality stand-alone policies or that if there were highquality stand-alone policies they would have to be offered at a higher price to overcome their providers’ lack of access to a point of sale. Our analysis in Appendix 10.4
showed that there is no evidence to support this contention because there are highquality stand-alone policies and because stand-alone policies are not offered at a
higher price when compared with a policy of equivalent quality sold at the point of
sale.
10.400 We therefore concluded from our analysis in Appendix 10.4 that stand-alone products are not necessarily of lower quality, and that there is no evidence that standalone policies are offered at a higher price when compared with a point-of-sale policy
of equivalent quality. There is therefore no evidence that stand-alone products are
necessarily of lower quality because they are sold away from the point of sale.
10.401 Finally, and importantly, in paragraph 5.144 we find that the point-of-sale advantage
is a feature of the supply of PPI that has resulted in a prevention, restriction or distortion in competition. We note that where firms are insulated from competition, they
face weaker incentives to offer higher-quality products for a given price. Therefore,
even if selling PPI at the point of sale of credit delivers benefits to distributors, which
could conceivably allow them to offer higher-quality products, they face little incentive
to pass this benefit on to consumers by offering a higher-quality product (one with
more generous policy terms) for a given price, or a lower price for a given quality of
product.
10.402 We found that PPI policies sold on a stand-alone basis can be of similar quality to
those offered at the point of sale (paragraph 10.400). We note that some stand-alone
policies have relatively low Defaqto quality scores; however, these are generally
offered at a low price. This may simply represent a business choice by those providers rather than any effect of selling PPI away from the point of sale. We also considered that in the absence of a competitive threat, distributors and intermediaries
would have diminished incentives to offer high-quality products at a given price (paragraph 10.401). We therefore concluded that the point-of-sale advantage did not lead
to a relevant customer benefit in the form of higher-quality products.
10.403 RBSG said that we should recognize the customer benefit of the availability of highquality products from integrated providers. However, as our remedies do not preclude sale of PPI products by distributors, we did not consider it necessary to consider whether there was a relevant customer benefit to be maintained.
Potential relevant customer benefits of a single-premium charging structure
10.404 We were given two reasons why a single-premium charging structure might result in
relevant customer benefits. First, we were told that charging a single premium was
more efficient and that consequently distributors’ costs were lower, resulting in lower
prices, higher quality and greater choice for customers. Second, we were told that
there were advantages for consumers of single-premium policies because the singlepremium charging structure allowed distributors to offer a greater continuity of cover
in the event of missed payments. 193
193
HSBC said that the availability of single-premium policies did not give rise to relevant customer benefits that could not be
replicated by regular-premium policies. It told us that HFC had just launched a regular-premium PLPPI policy sold at the point
of sale.
279
Efficiency of charging single premiums
10.405 We were told that single-premium charging was more efficient than a regularpremium charging structure and that, consequently, this allowed providers to offer
lower prices and/or better-quality products than would be the case under a regularpremium charging structure. A similar argument would apply to the rebate structure
for single-premium policies, which we were told was efficient in allowing providers to
offer a better product than would be possible with, for example, a pro-rata rebate
structure.
10.406 The main reason put forward as to why single premiums and their associated rebate
structure offer efficiency benefits, or lower costs, was because of the profile of the
risk on a policy over the life of that policy. Many parties told us that the incidence and
value of claims was higher in the early part of the loan and that therefore the risks
covered by the policy were higher in the early months of the policy.
10.407 The parties told us that, given this uneven claims profile, the simplest and most
logical response was to charge a single premium and provide a rebate for the unexpired cover upon early termination. This rebate is non-linear because of the uneven risk profile, so the value of the cover provided in the earlier months of the policy
is greater than the value of cover provided in later months.
10.408 With an uneven risk profile, under a regular-premium structure, or with pro-rata
rebates, those terminating the policy early would receive greater cover, in proportion
to the premium they had paid, than those holding the policy for longer. The only way
to reduce the impact of this, we were told, would be to reduce the benefits on offer to
try and ‘flatten’ the risk profile to make it more like a pro-rata rebate profile. RBSG
said that the premium could be amortized, so that the premium varied across the life
of the loan (starting higher and falling). Everydayloans told us that it sold a monthly
premium product whereby the premium was calculated as a percentage per month of
the balance outstanding on a fixed-term regular-repayment unsecured loan. This
would generate a reducing premium through the life of the loan. However, other
parties told us that charging a reducing premium for regular-premium PPI was not a
suitable approach. Aviva said that one of its partners had launched a product that
had had a reducing premium. However, it had stopped because it was too complex,
in terms of administration and systems, and there had been a lot of problems around
sales processes and explaining the premium to customers. HBOS said that the point
when customers borrowed money was not the most propitious moment to charge
larger premiums, and that it did not seem terribly customer friendly.
10.409 We analysed the data on claims. Details of this analysis are set out in Appendix 10.5.
We obtained claims profile data for 36- and 60-month single-premium policies from
five large distributors.
10.410 We noted first that claims costs make up a small proportion of the overall income
earned on single-premium policies. In paragraph 4.50 we found that, based on data
from the six largest underwriters, the average claims ratio over the five-year period
2002 to 2006 for PLPPI was 15 per cent. Any cost efficiencies associated with the
claims profile on single premiums are therefore likely to be small relative to the
current prices of single-premium PPI policies.
10.411 Our analysis of the profile of claims on these single-premium policies showed that
there is evidence that there are more claims in the early part of a policy (an effect
which was more pronounced for the 60-month loans than the 36-month loans). There
was a more pronounced skew in the value of claims in the earlier months of a policy
than there was in the incidence of claims. However, since a substantial proportion of
280
loans are settled early, we would expect more claims to occur in the early part of the
original loan period in any case, even if there were no difference in risk profile.
10.412 We therefore weighted the distribution of claims costs by the number of policies that
were still active at different points in the loan, by dividing the total lifetime cost of
claims occurring in each month by the number of policies which were active in that
month. We compared the weighted claims profiles of the five large distributors with
the rule of 78, which is used for assessing risk by many distributors for calculating
refunds (others use actuarial calculations which give very similar profiles—see paragraph 5.67) and to a ‘proportional’ distribution in which the profile of claims is not
affected by the length of time that the consumer has held the loan. The results are
shown in Figures 10.9 and 10.10.
FIGURE 10.9
Cumulative distribution of claim costs, weighted by the number of
active policies, over the life of the loan (%)
Cumulative distribution of claim costs, weighted by the number of active
policies, across the life of the loan: combined sample of 36-month
PLPPI policies
100
90
80
70
60
50
40
30
Actual claims cost distribution, weighted
by the number of active policies
20
Proportional
10
0
Rule of 78
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36
Months elapsed since policy start
Source: CC analysis of data provided by the large distributors.
281
FIGURE 10.10 Cumulative distribution of claim costs, weighted by the number of active policies, across the life of the loan: combined sample of 60-month PLPPI policies Cumulative distribution of claim costs, weighted by the
number of active policies, over the life of the loan (%)
100
90
80
70
60
50
40
30
Actual claims cost distribution, weighted
by the number of active policies
20
Proportional
10
0
Rule of 78
0
2
4
6
8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 52 54 56 58 60
Months elapsed since policy start
Source: CC analysis of data provided by the large distributors.
10.413 We found that the claims profile for 36-month PLPPI policies was between a straightline profile and the rule-of-78 profile—that is to say, while skewed towards earlier in
the loan term, the asymmetry in claims was not as pronounced as implied by a ruleof-78 rebate for these policies. For 60-month PLPPI policies we found that the rule of
78 was a reasonable reflection of the claims profile over the loan term. We also
analysed some data for 60-month SMPPI policies, and concluded that the distribution
of claims costs over the length of the loan most likely lay somewhere between that
implied by a proportional distribution of claims and that implied by the rule of 78 (see
paragraphs 20 to 22 of Appendix 10.5). Finally, we found that the claims profile for
personal loans offered by two large non-standard lenders ([] and []) displayed a
similar degree of asymmetry to standard personal loans, although in one case ([]
36-month PLPPI policies taken out in 2004) the claims profile skew was more
accentuated than that allowed for by the rule of 78 (see paragraphs 23 to 28 of
Appendix 10.5).
10.414 This indicated to us that offering a rebate on single-premium policies based on the
rule of 78 was overly generous to underwriters as an allocation of claims costs over
time for shorter-term policies (36 months), but was a reasonable approximation of the
profile of claims for longer-term policies (60 months).
10.415 To evaluate the scale of any cost efficiencies associated with single-premium policies, we considered how much claims costs and prices might increase if single
premiums were replaced with regular premiums or if pro-rata rebates were given. Our
analysis took into account both the claims profiles we found, and our observation that
claims costs make up a small proportion of the overall income earned on single282
premium policies (see paragraph 10.410). We found that the monthly average of the
lifetime value of claims costs per active policy in the first year of the loan was about
24 per cent higher than the monthly average of the lifetime value of claims over the
entire term for 36-month PLPPI policies, and just over 70 per cent higher for 60month PLPPI policies. Our analysis suggested that the differential could be slightly
greater for non-standard lenders—[>75] per cent for [] 36-month PLPPI. This
would seem to represent the maximum extent to which claims costs could increase
as a result of a move to a regular-premium, or pro-rata rebate, structure,
corresponding to the highly unlikely situation in which all customers switched within
the first 12 months of a loan.
10.416 We calculated an upper bound to the extent to which prices to customers might rise
as a result of any cost increase as a result of increased claims costs. This was
calculated assuming that a switch to regular-premium charging or a pro-rata rebate
would lead to an extreme result whereby all customers switched within a year of
taking out their policy, that all of the additional cost from claims as a consequence
was passed on to customers and that there was no downward pressure on PPI
prices as a result of increased competition. We regard these assumptions as
extremely unrealistic and in practice we would expect much lower levels of switching,
a smaller increase in claims costs and an increase in competitive pressure.
10.417 We found that even under these extreme assumptions the price effect was likely to
be modest. Our calculations showed that this could represent a maximum price
increase of less than 6 per cent for 60-month PLPPI, and slightly above this (up to
7.5 per cent) for non-standard lenders.
10.418 Even if these limited cost efficiencies do exist, the benefit does not accrue to PPI
consumers currently and we had no expectation that it would do so in the absence of
a comprehensive package of remedies to stimulate competition in the market. Without a competitive threat to distributors when selling PPI to their own credit customers,
in our judgement there is little incentive for distributors to pass these benefits on to
PPI consumers. We therefore concluded that the claimed efficiencies do not lead to
relevant customer benefits to PPI consumers and that in the absence of remedies to
stimulate effective competition in this market, we had no expectation that such
benefits would accrue to PPI consumers in the future. To the extent that any cost
efficiencies are currently passed on to consumers, this is more likely to appear as
lower credit prices, which we consider in paragraphs 10.442 to 10.463.
10.419 We were also told that single-premium policies were able to offer enhanced cover
compared with regular-premium policies, such as accident and sickness cover which
lasts for the duration of the policy rather than for a maximum of 12 months. However,
we concluded, and were told by one party (Lloyds TSB 194 ), that if these different
offers were valuable to customers, it would be possible to find a way of pricing
regular-premium policies that would allow them to be offered as part of a regularpremium policy. We noted that Lloyds TSB and Alliance & Leicester had both
recently introduced regular-premium PLPPI policies with exactly or essentially the
same benefits as the single-premium PLPPI policies that they replaced, although as
these products are new to market, we do not yet know how they will perform. As the
claimed benefit could accrue in the absence of single premiums, we concluded that
this alleged benefit does not satisfy the statutory test to be considered a relevant
customer benefit.
194
See also comments in the footnote to paragraph 10.40.
283
Continuity of cover for single and regular premiums
10.420 We considered whether there was a relevant customer benefit associated with continuity of cover guaranteed by having a single premium.
10.421 We asked distributors to tell us the proportion of regular-premium customers that
missed a regular-premium payment. We found that there was a relatively wide range
in estimates of the incidence of missed payments between types of PPI and between
distributors. 195 Estimates of the proportion of MPPI customers that had missed one or
more MPPI payments in 2007 were between 2.3 per cent ([]) and 17.5 per cent
([]). For CCPPI, estimates were generally higher, with, for example, [] and []
reporting that around 30 per cent of their customers had missed one or more
payments in 2007.
10.422 We also found a wide variation between distributors in their approach to customers’
cover in the event of missed payments. [], [] and [] told us that they cancelled
or suspended cover either immediately or within seven days of either the missed
payment or the second request for that payment. Other distributors continued cover
for longer. For example, [] will cancel MPPI cover after a period of around five
weeks from the missed payment. [], [], [], [] and [] said that their cover
would continue from between 30 and 90 days from the first missed payment. Finally,
[] told us that, for MPPI, it treated missed premium payments as arrears, that it did
not cancel or suspend cover and that missed premiums would not prevent the
customer from making a claim.
10.423 As noted in paragraph 2.79, the FOS told us that it would have doubts about whether
it would be fair or reasonable to cancel a policy immediately if only one or two payments were missed.
10.424 It was also suggested to us that single-premium policies prevented companies from
withdrawing cover. For example, London General Insurance noted that in unstable
economic times the insurer remained bound by the terms of the agreed policy and
was unable to withdraw cover, unlike with regular-premium policies. We were told
that there had recently been some changes in regular-premium policies because of
changing economic circumstances. Aviva told us that, in light of a significant increase
in sales of a Paymentshield unemployment-only PPI by British Insurance following
the turmoil in the financial markets from mid-September 2008, it had withdrawn that
policy from the market, and it was considering what to do about the significant
number of ASU policies being sold by British Insurance since the unemployment-only
policy was withdrawn. British Insurance told us that it had been able to source a new
underwriter. Aviva told us that it thought that a significant proportion of customers
buying these policies were doing so in response to a specific trigger event, and would
be at risk of making a claim. More generally, Cattles referred to us some media
articles which it told us showed the benefits of single-premium policies over regularpremium ones.
10.425 With increased competition, we would expect the general level of quality of PPI
policies to increase, as firms seek to win and retain customers by offering better
terms and better prices. We see no reason why this should not also include the
nature of terms such as the continuation of cover in the event of missed payments.
Given this, the evidence that many providers already offer a significant grace period if
195
Cattles told us that it would have been more appropriate to look at the proportion of single-premium consumers that had
missed a loan repayment. Our view is that either analysis would have been sufficient. Having considered the proportion of
regular-premium consumers that had missed a payment, we considered that it was not necessary to carry out a further analysis
of single-premium loan repayment histories.
284
a payment is missed, and the views of the FOS, we concluded that there was not a
relevant customer benefit associated with continuity of cover offered by singlepremium policies. Moreover, as this alleged benefit could accrue in the absence of
single premiums, even if alleged benefit did accrue it does not in our view satisfy the
statutory definition of a relevant customer benefit.
10.426 We noted the views expressed on the ability of regular-premium providers to withdraw cover. We agreed that single-premium policies lock the underwriter in to providing the cover sold, even when there is a change in economic circumstances, and
noted that the underwriter charges a premium taking account of this. However, if this
guaranteed cover were something to which customers attached significant importance, we saw no reason why suitably-priced regular-premium policies could not be
designed with clauses guaranteeing cover for the duration of the policy and that
therefore this benefit could accrue in the absence of single premiums.
Knowledge by credit providers that a consumer will take out PPI
10.427 It was put to us that consumers who take out PPI are, even with PPI, more likely to
go into arrears or default on credit repayments than those who do not take out PPI.
We were told that this could be used to help determine parties’ lending strategies.
10.428 [] told us that its insured portfolio was inherently more risky than its non-insured
portfolio. The net loss rate for [] insured credit card book is projected to be 8.7 per
cent, whilst it is only projected to be 3.5 per cent for its uninsured book. It told us that
any changes to its PPI income or portfolio would impact on its overall business and
lending strategy.
10.429 HSBC said that an important benefit of selling PPI at the credit point of sale was the
additional information that take-up of PPI provided in identifying higher-credit-risk
customers, allowing consumers to benefit from access to affordable protected credit,
which may not otherwise have been available. It said that PPI consumers incurred
significantly higher bad debt than non-PPI consumers despite their PPI cover, and
that its analysis showed that take-up of PPI was a powerful independent indicator of
bad debt propensity. It told us that, in the absence of being able to sell PPI at the
point of sale, lenders would be substantially less able to identify different risk categories, leading to the inability to supply credit to higher-risk consumers at all.
10.430 Finally, Capital One said that PPI revenue was integrated into its pricing methodology
when calculating rates of return. Although this resulted in lower credit prices to all
customers, it said that it typically observed higher losses for consumers with PPI than
those without, and that this would need to be taken into account in examining
whether ultimately there was a material cross-subsidy between these two groups.
10.431 Two potential efficiency gains associated with this difference in credit risk might arise,
related to distributors’ ability to identify which credit consumers intend to purchase
PPI, and to charge these individuals higher prices in order to account for the extra
risk they present. First, the extra information about a customer’s risk of default which
lenders are able to infer from their decision to purchase PPI could allow differentiated
prices to be charged to those credit consumers who intend to purchase PPI, and
those who do not. If so, these prices could be more efficient, better reflecting the
costs associated with providing credit to the two groups. Second, the extra information could serve to offset an asymmetry in the information available to lenders and
borrowers, reducing an adverse selection effect in the credit market and allowing the
lender to provide credit to a greater number of customers.
285
10.432 We consider these issues in more detail in Appendix 10.6. We focus here on whether
credit or PPI prices are at their current levels because distributors are reflecting the
additional risk associated with PPI consumers in higher credit or PPI prices, and, if
so, whether this should constitute a relevant customer benefit.
10.433 We looked first at whether PPI consumers are at higher risk of default on credit
repayments than credit consumers who do not take out PPI. For personal loans, we
found that consumers who took out PPI were more at risk of going into arrears or
defaulting on credit repayments than non-PPI customers. This difference in impairment experience was largely confined to credit consumers with higher-risk scores.
10.434 For mortgages, we were only able to obtain a small sample (most providers were
unable to tell us which of their mortgage consumers had PPI), but analysis of this
sample showed that those mortgage holders with PPI were also more likely to go into
arrears or have their mortgages written off than non-PPI customers.
10.435 For credit cards, we again found that consumers with PPI were more likely to go into
arrears or default on repayments than non-PPI customers, with the value of protected
credit card balances being written off approximately twice as high as the value of
unprotected credit card balances written off, even after controlling for the difference
in risk already observed, captured in credit risk scores.
10.436 We concluded, therefore, that credit consumers who had purchased PPI—and those
with higher credit risk scores in particular—were more likely to enter into arrears or
default on credit repayments than credit consumers without PPI. We considered that
the difference we observed was an overestimate of the difference that would pertain
in a well-functioning market because of certain features of our methodology and also
because a distortion in impairment experience caused by a lack of competition for
PPI may have exaggerated the difference that we observe.
10.437 We considered whether this higher risk was reflected in credit pricing. Nearly all distributors we spoke to told us that whether a consumer was going to take out PPI had
no impact on the credit decision, 196 and those distributors that did factor the decision
to take out PPI into the credit-approval decision said that it was a minor factor. We
noted that if the approval of credit was contingent on taking out PPI, the APR offered
would have to be based on the cost of both credit interest and PPI, and that no distributors currently offer credit contingent on taking out PPI. We concluded, therefore,
that the higher risk associated with PPI consumers is not generally reflected in individualized higher credit pricing for those consumers.
10.438 We also considered whether PPI prices are higher than they need to be to cover
costs of supplying PPI because distributors reflect the increased risk of credit default
among PPI consumers in the price set for PPl.
10.439 We received no clear evidence to suggest that the impairment experience of insured
versus uninsured consumers was taken into account when providers made their
decision about what PPI price to set. 197 Indeed, even []—which out of the three
parties that raised the potential issue of impairments spelt out the argument in most
detail—said that it ‘wouldn’t be right … to say that we explicitly allowed for [the extra
bad debt offered by protected customers] in the pricing of PPI’ when we asked it if it
196
There were some exceptions to this. [] has recently introduced whether or not a consumer requests PPI when signing up
for their credit card as a variable in assigning their risk score. [] told us that the intention to take out PPI affected whether a
consumer would be accepted for credit. One distributor ([]) provided us with a presentation that showed that it had modelled
the probability of default on loans with and without PPI and had considered reflecting this in the pricing of personal loans.
197
One credit card provider ([]) told us that one factor taken into account when setting the price of PPI was the overall
economics of credit cards.
286
priced PPI to take account of default risk. Because of this, there would seem little
reason to believe that current PPI prices have explicitly been set at the level associated with the extra impairment risk presented by these individuals.
10.440 Furthermore, credit providers did not appear to charge differentiated PPI prices on
the basis of individual customers’ risk scores. Given the substantial variation—particularly in the personal loan market—in the additional impairment risk presented by
credit consumers in different risk score deciles, a pricing strategy of charging higher
PPI prices to all PPI consumers is clearly inefficient relative to the theoretical optimum, under which the additional price paid by a given credit consumer also purchasing PPI would vary with that customer’s risk score.
10.441 Whilst we found that taking out PPI can indicate an increased likelihood of defaulting
on credit repayments, especially for customers in higher-risk bands, we concluded
that the informational advantage of being able to identify credit customers who buy
PPI was not as large as would be implied by our analysis. In addition, we found that
in practice the information was not used efficiently by distributors and that therefore
there was little evidence that this alleged efficiency led to benefits for consumers in
practice. Finally, our analysis in Appendix 10.6 showed that much of the difference in
impairment experience between insured and non-insured credit consumers was likely
to have been caused by the interaction of high PPI prices and adverse selection. We
concluded, therefore, that, as this information was not used efficiently by distributors,
and much of the difference in impairment could be explained by high PPI prices,
there was no relevant customer benefit being lost if distributors would not find out
when selling credit whether or not a consumer would take out PPI.
Lower credit prices
10.442 Many parties put to us that prices for credit were lower than they would otherwise be
because profits made on PPI policies were competed away in the form of lower credit
prices. Where firms make profits in secondary markets, and prices in the primary
market affect sales of secondary products, firms have an incentive to discount the
price of their primary market products in order to increase sales of the secondary
product. This is sometimes referred to as a ‘waterbed effect’. A similar argument was
put to us regarding availability of credit; the sale of PPI allows distributors to offer
credit to (higher-risk) customers to whom they would not otherwise offer credit.
10.443 We considered whether credit prices took account of profits made in the sale of PPI,
and if so what the likely scale of the price reduction in credit might be. We expected
that, if a waterbed effect existed, it would have a larger impact on personal loan
prices (in particular, unsecured personal loans) because the price of PPI is often
larger than the interest payable on the loan (see paragraph 2.73) and a higher proportion of credit sales have PPI attached than for mortgages and credit cards.
10.444 We considered evidence from a range of sources: the views of distributors, their
internal documents, what their financial models predicted if PPI incomes were
reduced, an analysis of the profitability of the bundle of credit and PPI, and econometric analysis of distributors’ sales data to see if credit prices were affected by
changes in PPI income.
The views of distributors
10.445 Distributors told us that a reduction in PPI income would result in major changes to
distributors’ (secured and unsecured) personal loan businesses (in terms of higher
prices and/or higher credit score cut-offs); for other forms of credit, we were told that
287
PPI generated a much smaller proportion of overall revenue and so would not result
in such major changes. Details of their views are set out in Appendix 10.7.
Distributors’ internal documentation
10.446 As noted in paragraph 4.88, some distributors’ internal documents showed that PPI
income has influenced pricing decisions for personal loans, has influenced lenders’
determinations of credit score cut-offs, and that some segments of customers within
low credit score bands would be unprofitable were it not for PPI income earned within
that segment.
Distributors’ financial models
10.447 The majority of distributors told us that they used financial models either in setting
their credit prices or in a higher-level evaluation of the profitability of their credit
businesses. Where distributors said that PPI income affected their credit pricing, we
asked them to undertake a modelling exercise, using their existing models to assess
the impact of reductions in PPI income.
10.448 We asked the distributors to estimate, for a given reduction in PPI income, the extent
to which their non-PPI income would need to increase in order to achieve the same
rate of return as they had achieved absent that reduction in PPI income (see
Appendix 10.7 for more details).
10.449 Under a full waterbed scenario, where all of the distributors’ PPI profits are competed
away in the credit market, we would expect distributors’ overall system profits to be
the same regardless of the level of PPI income. The modelling exercise therefore
gives an estimate of the upper bound of the scale of any distortion of credit prices:
(a) Estimates of the increases in APRs necessary to compensate for a 100 per cent
reduction in PPI income, across the personal loans portfolio, 198 ranged between
two percentage points ([]) and five percentage points ([]). 199 Estimates of the
extent to which non-PPI personal loan income would need to rise to achieve the
same level of profitability, given a 100 per cent reduction in PPI income, ranged
from 13 per cent ([]) to 93 per cent ([]).
(b) For credit cards, the estimates of the increase in non-PPI income that would be
required to achieve the same level of profitability were much lower. Estimates of
the increase necessary to compensate for a 100 per cent reduction in PPI income
ranged from 1 per cent ([]) to 22 per cent ([]). 200 In all cases, distributors
gave a lower figure for credit cards than for personal loans in their responses. We
noted that some elements of non-PPI income for credit cards may be inflexible,
implying a slightly larger increase in credit card APRs than would be implied by
the estimate of the necessary increase in non-PPI income alone. 201
198
We were told that some consumers, closest to the credit cut-off point, would face higher increases in APR than the reported
figures for the portfolio as a whole.
[] told us that it did not consider this modelling exercise to be realistic or informative.
200
[] estimated that its credit APR would need to increase by 22 per cent. This did not take into account other non-PPI
income, including revenue streams such as interchange fees.
201
For example, [] told us that many elements of the non-PPI income associated with credit cards were either fixed or were
under regulatory scrutiny. As a result, while the increase in non-PPI income required to achieve the same level of profitability
was small, the increase in the credit APR that this would imply was larger. [] estimated that whilst total non-PPI income would
need to increase by less than 1 per cent, this would be achieved by increasing the credit APR by 4 per cent. We note that, even
taking this into account, the potential waterbed effect on the credit card APR is still small, due to low penetration rates and the
relatively small size of PPI premium income in comparison with interest income on credit cards.
199
288
(c) Only three distributors were able to provide a response relating to second-charge
mortgages—[], [] and []. [] estimated that a 100 per cent reduction in
PPI income would necessitate an increase in its typical APR of between six and
nine percentage points above current levels. This estimate was smaller than its
projected increase for unsecured personal loans. [] estimated that a 100 per
cent reduction in PPI income would necessitate a proportionate increase in its
APR of 8 per cent. This was lower than its estimate for personal loans (13 per
cent). Finally, [] estimated that, for [], a 100 per cent reduction in PPI income
could necessitate an increase in second-charge mortgage APRs of up to 49 per
cent over current levels.
(d) Most distributors did not contend that a reduction in PPI income would have a
significant effect on the pricing of their first-charge mortgage products. For
example, [] estimated that non-PPI income would need to increase by 4.7 per
cent to recover lost PPI income, equivalent to a 0.03 per cent increase in the
APR.
10.450 There are several limitations to this approach which mean that this evidence needs to
be interpreted with care. These figures are based on the price-setting models in
place at the time we asked firms to carry out this exercise, and we noted that firms
may alter their price-setting models or their overall strategy in the event of a major
market-wide change.
10.451 It may also be the case that there is a less than 100 per cent pass-through of PPI
profits to consumers in the form of lower credit prices, in which case the effect of any
intervention in the markets for PPI on credit prices would be mitigated to an extent by
reduced distributor profits.
System profitability
10.452 We created models of market profitability, one for personal loans and one for credit
cards. 202 The models provide a high-level assessment of the profitability of the
personal loan and credit card markets over the last five years (2003 to 2007) using
publicly available information.
10.453 We used the models to identify trends in profitability in the two credit markets and the
relative importance of PPI revenues to that profitability. We summarize the results of
our models in Appendix 4.5.
10.454 In both market models the output was consistent with our previous findings:
(a) In the personal loans market, we saw declining profitability in recent years to the
extent that there was a negative contribution in 2007. We noted that the decline
in profits was due to a squeeze in net interest margins and an increase in impairment costs. We also saw that PPI income appeared to be a relatively small
factor contributing to the decline in profitability.
(b) In the credit card market, we saw that the sector was profitable over the last five
years even before taking into account income from PPI.
10.455 We cannot reliably infer anything about the extent of the waterbed effect from this
analysis. Although the negative contribution from credit appears to be driven primarily
202
We looked at the personal loans and credit card markets only and did not look at the mortgages market; we noted that MPPI
made up less than 10 per cent of the total ‘price’ of the bundle of MPPI and credit and that MPPI is sold on only approximately
15 per cent of mortgages sold by the large distributors.
289
by impairments, we do not know the extent to which PPI profits were passed through
to customers as a result of competition in the credit market.
Econometric evidence
10.456 We undertook an assessment of the distributors’ sales data in order to assess
whether there was any evidence that credit prices were affected by changes in PPI
income. It is difficult to draw firm conclusions from this evidence as there may be
other factors, which we cannot control for, affecting the observed correlations (or lack
thereof) between credit prices and factors affecting PPI income.
10.457 We looked at the correlation between personal loan APRs and PPI prices. We found
that there was a negative correlation over time between PPI prices and credit prices
for personal loans. This means that for an individual product we observe that as PPI
prices rise (fall) the credit price is lower (higher). This would be consistent with a
waterbed effect. However, this was a weak result, as there may be other circumstances and trends that could lead to lower credit prices and higher PPI prices over
time. We also have a large range in size of products and these results could therefore be driven by very small products. To overcome these potential problems, we
included monthly dummy variables to represent wider trends and weighted observations by their volume of sales. Doing this we found no significant correlation. A
fuller discussion of our econometric analysis is provided at Appendix 10.8.
Conclusions on lower credit prices
10.458 The evidence we saw led us to conclude that credit prices, and credit cut-off scores,
are lower than they otherwise would be because of PPI income generated at the
credit point of sale. For credit cards, the effect is small. For mortgages, the effect is
very small.
10.459 We concluded that these lower prices were a direct result of the distributors’ anticipation of high profit margins on PPI. Lower credit prices are therefore a direct result
of the features of the sale of PPI that lead to an AEC in the markets for PPI.
10.460 We conclude, therefore, that the features that we have identified as giving rise to the
AEC result in a relevant customer benefit within the meaning of the Act, namely lower
credit prices for personal loans (unsecured and secured), mortgages and credit
cards. 203 The only credit products on which we thought that such a waterbed effect
might result in an appreciable reduction in credit prices were unsecured and secured
personal loans.
10.461 We note that the incentive to discount credit prices is due, in part, to the intensity of
competition in the credit market. There is some uncertainty over the intensity of
competition for credit customers going forward, for example given recent events in
the credit markets involving the proposed consolidations of some credit providers
and, at least in one case, the running down of some business lines. In this respect
Cardif Pinnacle told us that in the current economic situation banks had made significant changes to their credit policies which had already had an effect on credit
prices, and hence the waterbed effect might not be as marked as our research
suggested.
203
Relevant customer benefits include benefits in markets other than the market to which the feature or features relate
(section 134(8) of the Act).
290
10.462 We note that, although levels of indebtedness in the credit markets have increased,
new gross lending advances have contracted, with credit providers now more
focused on their existing customer base and the quality of their lending book; rising
bad debt levels have forced some providers to review their customer acquisition
strategies by focusing on quality lending, increasing APRs, and lending only to
customers with whom they have a shared knowledge of credit history. The ‘credit
crunch’ has accentuated this trend.
10.463 As a result of these factors, we could not be confident that the scale of the relevant
customer benefit that we observed in the period up to December 2006 would persist
at that level in the future.
10.464 We consider whether we should modify our remedies package to preserve this relevant customer benefit in paragraphs 10.480 to 10.492.
The effectiveness and proportionality of the package of remedies
10.465 Our analysis sets out the following package of remedies for implementation:
(a) a prohibition on selling PPI at the credit point of sale and within a fixed time
period of the credit sale (‘the point-of-sale prohibition’);
(b) an obligation to provide a personal PPI quote (‘the personal PPI quote’);
(c) an obligation to provide information about the cost of PPI and ‘key messages’ in
PPI marketing material (‘information provision in marketing material);
(d) an obligation to provide information to the OFT and the FSA for monitoring and
publication; and an obligation to provide information about claims ratios to any
party on request (‘provision of information to third parties’);
(e) a recommendation to the FSA that it uses the information provided to it under this
obligation to populate its PPI price-comparison tables;
(f) an obligation to offer retail PPI separately from merchandise cover where both
are offered together as a bundled product (‘unbundling retail PPI from merchandise cover’);
(g) a prohibition on the selling of single-premium PPI policies (‘single-premium
prohibition’); and
(h) an obligation to provide an annual statement of PPI cost and a reminder of the
consumer’s right to cancel (‘annual statement’).
The rationale for implementation of all elements of the remedies package
10.466 We considered whether we needed to implement all of these elements of the
remedies package in order to address the AEC and adverse effects identified.
10.467 Many parties said that they were in favour of remedies which improved transparency
and encouraged shopping around by consumers. Some of these parties suggested
that a package of remedies based around providing additional information to consumers to address searching and switching issues would remedy the consumer
detriment that we identified existed in the market. However, it was also apparent that
some parties did not consider that the remedies they favoured would remedy the
291
AEC we have found (and with which they often disagreed). For example, Abbey said
that it ‘did not consider it to be certain that there would be a significant change in
consumer behaviour if consumers were given more information/opportunities to
switch’. 204
10.468 The FSA told us that, whilst further information to enable consumers to search the
market could bring some additional benefits, the incremental benefit would be limited,
as it could not address the barriers to searching and switching we had identified. 205
Cardif Pinnacle told us that [] it did not believe that informational remedies alone
would not have a material effect on the market, and consumer switching in particular.
10.469 We also noted that the recent evaluation of the Extended Warranty Order found that
while the remedies package put in place following the CC’s investigation—comprising
information provision at the point of sale, a cooling-off period for 45 days and pro-rata
rebates beyond that point—has had a net beneficial effect on consumers, the Order
has only resulted in a relatively small reduction in consumer detriment (of
£18.6 million a year) compared with an estimated annual detriment of £366 million. 206
10.470 We concluded that the sale of PPI at the point of sale of credit acts as a barrier to
other PPI providers competing for customers of any given distributor or intermediary. 207 We found in paragraph 5.87 that there are significant barriers to entry for
stand-alone PPI providers seeking to sell PPI products without access to consumers
at the credit point of sale, due to adverse selection, poor consumer awareness and
high marketing costs. 208 These factors would not be addressed by a package of
information remedies.
10.471 The evidence led us to conclude that informational remedies would help remedy the
AEC identified, but that these alone would not be sufficient to remedy the lack of
competition we saw between PPI providers; we would also need to address the
point-of-sale advantage.
10.472 It was put to us that, if we adopted remedies designed to incentivize distributors to
price credit and PPI combinations more efficiently, that would be likely to result in
cost savings associated with selling PPI at the point of sale being passed on, at least
in substantial part, to consumers (see paragraph 10.385). However, if there were
efficiencies to pass on to consumers, they would only be passed on if competitive
markets could be achieved without a point-of-sale prohibition. As set out in paragraph
10.471, we concluded that we could not address the point-of-sale advantage by providing information alone at the point of sale. Further, as set out in paragraph 10.79,
we did not find an effective way of addressing the point-of-sale advantage without
introducing a prohibition on a distributor selling PPI within a certain time period of the
sale of the underlying credit product, whilst allowing consumers at their own initiative
to contact the distributor, intermediary or other business covered by the prohibition to
purchase its PPI. We concluded therefore that it was necessary to prohibit the sale of
PPI at the point of sale as part of the remedies package.
10.473 In the case of retail PPI only, none of the informational remedies would address the
barrier to search created by the bundling of merchandise cover with PPI. While this
barrier to search remained, customers would find it difficult to make comparisons
between the PPI policies offered by the retail credit providers and alternatives that
204
In Abbey’s response to the Remedies Notice, paragraph 2.4. FSA response to the Remedies Notice, p8. 206
Evaluating the impact of the Supply of Extended Warranties on Domestic Electrical Goods Order 2005, prepared for the OFT by LECG, October 2005. 207
Paragraph 5.116. 208
Paragraph 5.87.
205
292
did not include merchandise cover. We therefore concluded that it was necessary to
require PPI to be offered separately from the bundled product.
10.474 Having determined the remedies necessary to address consumer search and the
impact of the point-of-sale advantage, we considered whether we needed any further
remedies to address switching barriers. We considered that the provision of an
annual statement is needed to help consumers to focus periodically on their PPI
policy and assist them to identify whether it still represents the best-value PPI policy
for their circumstances. The remedies to enhance searching will help potential
switching consumers to identify policies which might represent best value for them.
10.475 However, the annual statement would not be sufficient to reduce the switching
barriers associated with single premiums. As set out in paragraph 10.257, a prohibition on the sale of single-premium policies would be the only effective way of
removing this switching barrier, as well as helping to address barriers to searching.
10.476 We concluded, therefore, that it was necessary to implement all the elements of the
remedies package.
Benefits and synergies of the remedies package
10.477 The remedies package will encourage consumers to search by removing many of the
barriers to searching that we identified. In particular, this remedies package will
improve the transparency and comparability of price information, will offer consumers
a clearer understanding of the cost of PPI (and hence the benefits to searching) and
will remove some of the persistent consumer misconceptions that previou