Download Report

Do Insurance Companies Pose
Systemic Risk?
J. David Cummins
2013 China International Conference
on Insurance and Risk Management
Kunming, China
July 18, 2013
Copyright J. David Cummins, 2013, all rights reserved. Not to be reproduced without author’s permission.
Outline of Presentation

What is Systemic Risk?

Systemic Risk: Primary Indicators – Factors Used to
Identify Systemically Important Financial Institutions

Systemic Risk: Contributing Factors – Factors
Exacerbating Vulnerability to Crises

Conclusions: Does Insurance Pose Systemic Risk?


Based on primary indicators and contributing factors
Are Insurers Instigators and/or Victims of Systemic Risk?

An econometric analysis using Granger causality
This Presentation Based on Two Papers



Cummins, J. David and Mary A. Weiss, 2013, “Systemic Risk and the
Insurance Industry,” forthcoming in Georges Dionne, ed., Handbook of
Insurance, 2d ed. (Springer).
Chen, Hua, J. David Cummins, Krupa Viswanathan, and Mary A.
Weiss, 2013, “Systemic Risk and the Inter-Connectedness between
Banks and Insurers: An Econometric Analysis,” forthcoming, Journal of
Risk and Insurance.
Other relevant papers



Chen, Hua, J. David Cummins, Krupa Viswanathan, and Mary A. Weiss, 2013,
“Systemic Risk Measures in the Insurance Industry: A Copula Approach,” working
paper, Temple University, Philadelphia
Cummins, J. David and Mary A. Weiss, 2013, “Systemic Risk and the Regulation of
the U.S. Insurance Industry,” working paper, Temple University, Philadelphia.
Cummins, J. David and Mary A. Weiss, 2012, “Systemic Risk and the U.S.
Insurance Sector,” forthcoming, Journal of Risk and Insurance.
To obtain the papers, please email: [email protected].
What is Systemic Risk?
What Is Systemic Risk?


The risk that an event will trigger a loss of economic value
or confidence in a substantial segment of the financial
system serious enough to have significant adverse effects
on the real economy. Group of 10 (2001).
Systemic financial risk involves




A system-wide financial crisis . . . accompanied by a sharp decline
in asset values and economic activity
The spread of instability throughout the financial system
(contagion)
Sufficient to affect the real economy
World Economic Forum (2008).
Systemic risk is exposure to extreme correlations
Financial Crises and Systemic Risk
 Financial






Crises
Prices of risky assets drop sharply
Prices of safe assets increase (flight to quality)
Asset price volatility increases
Liquidity dries up (rising bid-ask spread & price impact)
Financial institutions become financially distressed
Credit markets dry up, economic activity depressed
 Financial
systemic risk: Financial crisis in which
many institutions become financially distressed,
with a potential impact on real economic activity
Financial distress of one or a few institutions
does not necessarily equal systemic risk!
Too-Big-To-Fail Has Been With Us For a Long Time and Isn’t
Confined to Financial Institutions
Industry/Company
Year
Type of Assistance
Penn Central Railroad
1970
$676.3 million in loan guarantees –
Gov’t spent $19.7 billion and got back
about $4 billion
Lockheed
1971
Government loan which was paid off
New York City
1975
Loans and loan guarantees
Chrysler
1980
Loan guarantees and warrants – Gov’t
earned a profit of about $660 million
Continental Illinois
1984
Government took 80% ownership and
phrase TBTF was coined
Airline Industry
2001
Government bought stock below
market and provided loan guarantees
Automobile industry
20082009
Government takes equity stake in GM
& Chrysler = $80 billion
Systemic Risk:
Primary Indicators
& Contributing Factors
Primary Indicators and Contributing Factors
 The
Question: How to identify systemically risky
markets and institutions?
 Primary indicators: Factors used to identify
systemic markets and institutions
 Contributing factors: Determine the vulnerability
of an institution or market to systemic events

An institution may be systemic in terms of primary
indicators but not vulnerable in terms of contributing
factors
Primary Indicators of Systemic Risk
 Size


– Macroeconomic Importance of Insurers
Size not limited to conventional measures such as
assets
Volume of transactions, exposure to off-balance sheet
positions, and derivatives also play a role
– degree of correlation and
potential for contagion among institutions
 Lack of substitutability –
 Interconnectedness


Are there effective substitutes for an institution’s
products?
Are the products critical to the functioning of the
financial system
Primary Indicators: Size
measures – assets or equity,
absolutely or relative to GDP
 Off-balance sheet exposures and volume of
transactions processed
 Conventional

“Callability” of positions taken – marginability, etc.
 Notional
value of derivatives exposure (e.g.,
AIG’s credit default swaps positions)
 “Too Big to Fail” being replaced by “Systemically
Important Financial Institution (SIFI)”

Reflects inadequacy of conventional size measures
Designating SIFIs
 Financial


Stability Oversight Commission (FSOC)
Established as part of U.S. Treasury in 2010
Can designate banks and non-banks as SIFIs
 Thresholds


for SIFIs
≥ $50 billion of assets, and
Meets or exceeds any one of several thresholds
●
●
●
●
●
$30 billion notional CDS for which firm is reference credit
$3.5 billion of derivative liabilities
$20 billion of total outstanding debt (bonds, etc.)
15-to-1 leverage ratio (assets/equity)
10% ratio of short-term debt to assets
Primary Indicators: Interconnectedness
Could Insurers Cause a “Run on the Bank”?
– extent to which financial
distress at one or a few institutions increases
probability of distress at other institutions
 Interconnectedness


Network or “chain” effects on both sides of balance
sheet or through derivatives exposures, off-balance
sheet commitments, etc.
Existence of “contagion” in the economy
 Interconnectedness
“runs on the bank”
creates conditions that trigger
All Systemic Financial Crises Involve “Runs”
 In
crises, investors seek cash at all costs
 As prices no longer adjust supply, access to
credit becomes central
 Maturity mismatch compounds shock and
spreads runs


Rapid withdrawals lead to “fire sales,” prices crash
Losses induce margin calls, more fire sales
 Runs
are both cause and consequence of
extreme correlation
 Crises that do not spread to general credit market
do not qualify (e.g., “dot-com” bubble collapse in
2000-2001)
Types of “Runs”
 In
the past (e.g., 1930s), financial crises often
involved “retail runs”


Many depositors try to withdraw money from banks
simultaneously
Deposit insurance has virtually eliminated retail bank runs
 Recent



financial crisis involved “wholesale runs”
Banks refused to provide liquidity to troubled institutions
such as Lehman Brothers
AIG counterparties demanded higher margin deposits and
the unwinding of asset lending relationships
Essentially, a “run” on the shadow banking system
“Runs” Triggered by Exposure to Common Shocks
 Common
shock may be exposure to agricultural
depression, real estate, or oil prices
 In the Crisis of 2007-2009, common shock was
the bursting of the housing price bubble
 Bursting of bubble triggered crises in




Inter-bank lending
Commercial paper
Market for short-term repurchase agreements (“repos”)
Essentially a run on the shadow banking system
“Too Big to Fail” & “Too Interconnected to Fail”
 Institutions
that pose significant systemic risk are
viewed as “Too Big to Fail” -- e.g., failure would
cause ripple effects throughout the economy due
to the sheer size of the enterprises
 “Too Interconnected to Fail” -- Firms with multiple
counterparty relationships could trigger a
cascading chain of failures – “domino effect”
Primary Indicators: Lack of Substitutability
Could Unavailability of Insurance Cause a Financial Crisis?
– extent to which other firms or
segments of the financial system can provide the
same services provided by the failed institutions
 For substitutability to be a problem, the services
must be critical to the functioning of other
institutions or the financial system
 Substitutability



Payment system – operational arrangement that
enables individuals and institutions to transfer funds
Settlement system – enables transfer of securities and
cash to settle trades
Liquidity system – inter-bank lending, repos, etc.
Contributing Factors:
Enhance Vulnerability to Systemic Events
 Leverage


Measured conventionally as debt to equity
Off-balance sheet positions, options exposure, and
“marginability” of positions also creates leverage
 Leverage

Declines in asset values erode net worth much faster
than the asset declines themselves (“leverage”)
●

and “loss spirals”
E.g., at 10-to-1 assets/equity ratio, 5% decline in assets means
50% decline in equity
If many institutions are affected at same time, selling
assets puts additional pressure on prices generating a
loss spiral
Contributing Factors:
Liquidity Risk and Maturity Mismatches
risk – vulnerability to shocks is increased
to the extent the institution holds illiquid assets
 Liquidity


Difficulties in obtaining financing may trigger need to sell
assets – problematic for illiquid assets
Especially serious if other institutions also illiquid
 Asset-liability


maturity mismatch raises liquidity risk
In 2007-2009 Financial Crisis, Shadow banks used shortterm commercial paper, overnight lending, and repos to
finance longer-term assets
Disappearance of short-term financing triggered need to
sell illiquid longer-term assets
What Are Repos?
– abbreviation for “sale and repurchase
agreement”
 Repos




Defined: Sale of securities together with an agreement
for the seller to buy back the securities at a later date
Repurchase price > original sale price, the difference
representing interest, the “repo rate”
Seller is a borrower, using securities as collateral for a
cash loan at specified interest rate
The buyer acts as a lender
 During
financial crisis, repo buyers refused to
extend credit to firms such as Lehman
Contributing Factors:
Complexity Enhances Vulnerability to Shocks
 Dimensions



Complexity of organization or group structure – firms
offering banking, insurance, and investment products
more complex than single industry firms
Geographical complexity – multi-national firms face
variety of local and regional risk factors
Product complexity exposes firms to risks that may not
be fully understood
●
E.g., AIG Financial Products CDS transactions
 Complexity

of complexity
aggravated by opacity
AIG’s positions were opaque, preventing market
adjustment for over-exposure
Contributing Factors:
Government Policy and Regulation
 Government
policy and regulation can contribute
to financial system fragility





Deposit insurance and guaranty funds create moral
hazard that may lead to crises – buyers with
government guarantees have no incentive to monitor
“Too Big To Fail” policies also create moral hazard
Complexity of AIG created regulatory blind spot that
led to AIG’s near collapse (nominally regulated by
Office of Thrift Supervision)
US government policy permitted investment banks to
become over-leveraged, helping to precipitate crisis
Lack of regulatory oversight contributed to the housing
bubble and mortgage backed securities crisis
Size Risk:
The Macro-Economic Importance
of Insurers
Total Assets: US Banks and Insurers
Assets: Banks $14.6 trillion, insurers $6.8 trillion.
Assets ($Billions)
20,000
15,000
10,000
5,000
0
Banks
Life Insurers
Source: Federal Reserve Flow of Funds accounts.
PC Insurers
Total US Life and P-C Premiums: % of GDP
Premiums/GDP (%)
9.0%
8.5%
8.0%
7.5%
7.0%
6.5%
Source: A.M. Best Company, American Council of Life Insurance, St. Louis Federal Reserve Bank.
US GDP From Financial Services (Value Added)
9.0%
Contribution to GDP (%)
8.0%
7.0%
6.0%
5.0%
4.0%
3.0%
2.0%
1.0%
0.0%
Insurance
Finance Total
Source: US Department of Commerce, Bureau of Economic Analysis.
Insurance Companies: Share of Total Assets
0%
5%
10%
Corporate Bonds
Municipal Bonds
Agency & GSE Bonds
Treasury Securities
Corporate Equities
%P&L
%Life
Source: Federal Reserve Flow of Funds Accounts.
15%
Conclusions: How Big Are Insurers?
 Insurers




have $6.8 trillion in assets
About 50% as large as commercial banks
Only about 8% of total US financial assets
Insurers do not have large share of any asset market
Insurer insolvencies resolved gradually so even large
insolvency would not lead to liquidity problems
 Insurers
not very important source of GDP (< 3%)
 Therefore, as a sector insurers do not pose
systemic risk due to their size alone
Interconnectedness Risk:
Could Insurers Cause
A “Run on the Bank”?
Interconnectedness:
Insurers and Other Financial Firms I
 Do
US insurers invest heavily in other financial
institutions?

Investment in Banks:
●
●

5.6% of insurer assets are in bank bonds
1% of insurer assets and in bank equities
Investment in Securities firms:
●
●
1.6% of insurer assets in securities firm bonds
1% of insurers assets in securities firm equities
 Conclusion:
Insurers are not vulnerable to stock
and bond declines from financial firms
Interconnectedness:
Insurers and Other Financial Firms II
 Are
insurers a significant source of funds for other
financial institutions?

Life insurers supply:
●
●
9.4% of outstanding bonds for banks
14.1% of outstanding bonds for securities firms
 However,
bonds account for only 10% of
financing for banks and securities firms
 Conclusion: Banks and securities firms are not
dependent on insurers to finance their activities
Interconnectedness Within Insurance Industry
 Reinsurance
is the primary source of intraindustry interconnectedness
 Reinsurance creates risk of counterparty default



Reinsurer failure to pay claims can trigger insurer
insolvencies
Insolvent insurers then default on their reinsurance
counterparties
Result: A Reinsurance spiral
 Reinsurance


counterparty risk present for both
Reinsurance transactions with affiliates
Non-affiliated reinsurance transactions
Measures of Reinsurance Interconnectedness
 Reinsurance
premiums ceded
 Insurance in force ceded (life)
 Reinsurance receivables – funds owed by
reinsurers to ceding company
 Write-down of liabilities due to reinsurance:
Reserve credit taken (life)
Net amount recoverable from reinsurance (P-C)


 If
reinsurers fail:


Default on receivables and ceded premiums
Liability write-downs canceled, increasing leverage
Extent of Reinsurance Interconnectedness
 Reinsurance
premiums ceded to affiliates & non-
affiliates


P-C premiums ceded = 17.2% of surplus
Life premiums ceded = 38.0% of surplus
 Insurance
in force ceded (life) = 55.3% of surplus
 Reinsurance recoverables –


For 25% of P-C insurers, recoverables > 40% of surplus
For 25% of life insurers, recoverables > 100% of surplus
 Write-down


of liabilities due to reinsurance:
Reserve credit taken (life) averages 149% of surplus
Recoverables (P-C) average 39% of surplus
Have Reinsurance Failures Been a
Significant Source of Insurer
Insolvency?
P/C Impairments: Triggering Events
Misc.
8.6%
Reinsurance
Failure
Sig. Change in 3.6%
Business
4.0%
Deficient Loss
Reserves/Inadequate
Pricing
40.3%
Investment
Problems
7.3%
Affiliate
Impairment
7.8%
Catastrophe
Losses
7.1% Alleged Fraud
7.8%
Rapid Growth
13.6%
Source: A.M. Best: 1969-2010 Impairment Review, Special Report, May 2, 2011.
Deficient loss
reserves,
inadequate pricing,
and rapid growth
are the leading
triggers.
Investment,
catastrophe, and
reinsurance losses
play a much
smaller role.
L-H Impairments: Triggering Events
Life insurers more susceptible to affiliate problems.
Sig. Change in
Business
4.6% Reins Failure
Alleged Fraud
8.9%
1.9%
Misc
8.2%
Inadequate
Pricing
28.5%
Inadequate
pricing, affiliate
problems, rapid
growth, and
investments are
primary causes of
L/H insolvencies.
Affiliate
Problems
18.4%
Investment
Problems
15.2%
Rapid Growth
14.3%
Source: A.M. Best: U.S. Life/Health – 1976-2010 Impairment Review, Special Report, May 23, 2011.
Reinsurance Interconnectedness: Conclusions
 Reinsurer
failure traditionally not a major source
of insurer insolvency
 Many insurers do have large reinsurance
counterparty exposure relative to surplus


Therefore, reinsurance failure could threaten solvency
of individual insurers and create industry-wide crisis
However, it is unlikely that even a major intra-industry
crisis would spill over into broader financial markets
 Therefore,
reinsurance causes intra-industry
vulnerability but not systemic risk
Interconnectedness and Non-Core Activities
non-core (“banking”) activities can create
interconnectedness and systemic risk
 Insurer

Example: AIG Financial Products
 Non-core





activities that may be systemic
Credit derivatives transactions
Asset lending programs
Financial guarantees and other off-balance sheet
commitments
Reliance by insurers on short-term financing
Subsidiaries with high exposures relative to capital
 Improved
regulation needed to prevent crises
Substitutability Risk:
Could Unavailability of Insurance
Cause a Financial Crisis?
Lack of Substitutes and Crises
For lack of substitutability to cause a crisis both of
the following must be true:
 The product must be unavailable and have no
substitutes or alternative suppliers
 The product must be essential for the functioning
of other institutions or the financial system
Quantitative Measures of Substitutability
For lack of substitutability to cause a crisis:
 Concentration (market share of top firms) –
highly concentrated markets more likely to
trigger crisis due to lack of substitutes
 Ease of entry into the market


If entry barriers exist, new entrants prevented form
providing vital products or financial services
Ease of entry can mitigate concerns about lack
substitutability
Concentration and Regulation in Insurance
 Concentration


in insurance: US
Top 4 (10) non-life groups have 29% (50%) of market
Top 4 (10) life groups have 24% (45%) of market
 Entry
of new insurers relatively easy (both on
shore and off-shore, e.g., Bermuda)
 Nationally significant insurers reviewed quarterly
by the NAIC – Financial Analysis Working Group
 Therefore, widespread insolvencies causing
insurance unavailability are extremely unlikely

Survivors or new entrants would provide coverage
Do Insurance Products Have Substitutes?
 Life


Insurance
Mostly asset accumulation products rather than
mortality/longevity risk bearing
Many non-insurance substitutes for asset
accumulation and investment products
●


Banks, mutual funds, securities firms, etc.
Many insurers available to fill coverage gaps resulting
from insolvency of one or a few firms
Therefore, lack of substitutes not a problem for life
insurance
Do Insurance Products Have Substitutes?
 Property-Casualty


Mainly provide risk management and risk-bearing
No real substitutes for individual buyers (auto
insurance) and small commercial customers
●


(PC) Insurance
But many insurers are available to fill coverage gaps
resulting from one or a few insolvencies
Large corporate buyers have substitutes – self
insurance, captives, securitization
Therefore, lack of substitutes not a problem for P-C
insurance
Is Insurance Critical to Functioning of Economy?
 Insurance
clearly enables the economy to
function more smoothly by enabling individuals
and businesses to take more risk
 However, it is difficult to argue that insurance is
as important as banking, the payments system,
or the settlement system
 Various insurance markets regularly experience
availability crises without significantly affecting
real economic activity
 Therefore, unavailability of insurance unlikely to
create a systemic crisis
Contributing Factors:
How Risky Are Insurers?
Contributing Factors:
Leverage and Insolvency Rates
Equity Capital-to-Assets Ratios
40.0%
35.0%
30.0%
25.0%
20.0%
15.0%
10.0%
5.0%
0.0%
Banks
Life Insurers
PC Insurers
Source: Federal Reserve Flow of Funds accounts, American Council of Life Insurance, FDIC.
Premiums-to-Surplus Ratios: US Insurers
Insurance leverage ratios have been improving over time.
3
2.5
2
1.5
1
0.5
PC Insurers
LH Insurers
Failure Rates: US Banks & Insurers
Bank failure rate was more strongly affected by the crisis.
3.5%
3.0%
2.5%
2.0%
1.5%
1.0%
0.5%
0.0%
Banks
LH Insurers
PC Insurers
0
Source: A.M. Best; Insurance Information Institute
16
19
21
13
16
19
17
20
30
28
37
34
32
30
33
36
40
16
19
33
41
48
49
47
48
49
50
50
12
15
14
14
13
12
11
55
60
59
60
5
9
9
9
20
15
12
70
7
8
10
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
Number of Impairments
P/C Insurer Impairments: 1969-2011
Financial Crisis Did Not Trigger Many Impairments.
Why Did PC Impairments Increase in 2011?
 Lingering

Crisis appears to have had delayed effect on P-C
insurers
 Near


record catastrophe losses
Insured losses of $116 billion
2nd largest year for catastrophes in recorded history
(largest was 2005 when Hurricanes Katrina, Rita, and
Wilma and other events caused losses of $123 billion)
 “Soft

effects of financial crisis
market” phase of underwriting cycle
PC insurance supply increases and prices decrease
PC Insurer Impairments & Combined Ratio
Impairment rates highly correlated with underwriting performance.
120
Correlation with combined ratio = 64%
2.0
1.8
1.6
1.4
110
1.2
105
1.0
0.8
100
0.6
0.4
95
90
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
0.2
Combined Ratio after Div
Source: A.M. Best; Insurance Information Institute
P/C Impairment Frequency
0.0
Impairment Rate
Combined Ratio
115
0
Source: A.M. Best.
13
10
9
7
5
6
9
9
10
12
26
8
20
18
25
39
47
50
12
11
30
24
27
32
40
16
17
15
16
11
12
13
13
55
60
3
11
11
13
8
10
6
16
82
90
2
10
8
11
20
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
01
02
03
04
05
06
07
08
09
10
11
Number of Impairments
Life/Health Insurer Impairments:1976-2011
Life/health impairments less cyclical than P/C
80
70
LH Impairment Frequency & Profits
3.5
7.0
3.0
6.0
5.0
2.5
4.0
2.0
3.0
1.5
2.0
1.0
1.0
0.5
0.0
0.0
-1.0
L/H FIF
Source: A.M. Best.
A-T Profit Margin
After-tax Profit Margin (%)
Failure Frequency (%)
LH less correlated with profits than PC, Corr = -20%
PC Guaranty Fund Assessments: 1978-2010
Correlation = 81%.
1600
0.50%
GF Assessments
% of NPW
1400
0.45%
Assessments ($Millions)
1200
0.35%
1000
800
600
0.30%
0.25%
0.20%
0.15%
400
0.10%
200
0
Source: A.M. Best Company, National Conference of Insurance Guaranty Funds.
0.05%
0.00%
Assessments: % of Premiums
0.40%
LH Guaranty Fund Assessments: 1988-2010
1,000
0.30%
900
Assessments ($Millions)
700
0.20%
600
500
0.15%
400
0.10%
300
200
0.05%
100
0
0.00%
Assessments
% of Premiums
Source: A.M. Best, National Organization of Life and Health Insurance Guaranty Associations.
Assessments: % of Premiums
0.25%
800
US Insurance Stock Indices vs. S&P 500
PC insurers beat the S&P during crisis, life insurers did not.
Index 12/312004 = 1000
1800
1600
1400
1200
1000
800
600
400
200
0
Life
P&C
S&P
S&P Bank
Insurer Leverage & Solvency: Conclusions

US regulated insurance companies are highly solvent




Life insurers give some cause for concern





Insolvency rates are low
Guaranty fund costs are low
Financial crisis had little impact on insurer insolvencies
More highly leveraged than PC but about same as banks
More interconnected than PC insurers (susceptibility to affiliates)
LH stocks harder hit by crisis than PC stocks
Inter-connectedness does not pose serious solvency
threat for PC insurers based on past experience
Monolines (insurers of bonds) are a different story

Not traditional insurance
Contributing Factors:
Liquidity Risk
and Asset-Liability Mismatches
Liquidity Risk
 Danger


signals for life insurance industry
ABS/MBS = 194% of surplus (only 27% of surplus for
PC insurers)
Privately placed bonds = 204% of surplus (only 10% of
surplus for PC insurers)
 But,
life insurers have significant cash from
operations


43.8% of surplus
28.1% of benefit payments
 Conclusion:
Liquidity risk exists from mortgagebacked securities and private placements for life
but not PC insurers – partly offset by cash flow
Maturity Mismatches
 Asset
and liability maturities tend to be long-term
for insurers (in absolute terms and relative to
banks)
 Property-casualty liabilities not “putable”

Must experience a loss and file a claim to collect
 Most


life insurance long-term and not putable
Exceptions: cash value life insurance, variable life, and
variable annuities
However, usually a penalty for early surrender
 Conclusion:
Maturity mismatch not a problem but
some putability risk for life insurance
Contributing Factors:
Complexity
Complexity
 AIG



prime example of complexity
Complicated group structure
Geographically dispersed
Complex, new financial products
 Large
multi-national insurers common in
insurance industry
 Life insurance more complex than PC

Most life products have embedded derivatives
 Conclusion:
Complexity is a problem for the large,
multi-product, multi-national insurers
Contributing Factors:
Government Policy and
Regulation
Do Guaranty Funds Create Moral Hazard?
 In
theory, mis-priced guaranty fund coverage
provides incentives for excessive risk-taking
 In practice, guaranty funds do not seem to be a
problem


No solvency crisis for US regulated insurance
companies – now or during Financial Crisis
Guaranty fund assessments have been very low
 Possible


rationale:
Risk-based capital (introduced in 1994) blunts insurer
incentives for excessive risk-taking
GF protection is incomplete (low maximums, etc.)
Regulation of Complex Multi-Nationals
 Generally,
complex multi-national financial
service firms lead to gaps in regulation

No one regulator has responsibility for entire firm
●
●

Different national regulators have responsibility for nationally
domiciled subsidiaries
Banking and insurance subs may be regulated by different
organizations
At least in US, regulation of insurers mostly at the
individual insurer rather than the group level
 Conclusion:
Better supervision needed for
insurance groups and multi-national financial
services firms
Conclusions: Does Insurance
Pose Systemic Risk?
P-C Insurance May Not Create Systemic Risk

“Runs” are not possible






To obtain funds, it is necessary to have a claim
Unlike bank deposits, which are instantaneously “putable”
Insurance not involved in liquidity creation, payments
system, or business/consumer lending
Insurers hold only small proportion of total invested
assets in the economy
Insurance claim payments are not a major financial asset
for any economic sector
However, intra-sector reinsurance exposure could cause
“reinsurance spiral” spreading across the P-C industry

Not clear if this would be a true systemic event, i.e., not likely to
affect other financial institutions or the real economy
Does Life Insurance Pose Systemic Risk?
 Why





LI may be systemically risky
Life insurance investment products are susceptible to
“runs” (withdrawals and/or suspension of premium
payments/annuity considerations)
Life insurers are thinly capitalized in comparison with
P-C insurers, but similar to banks
Life insurers hold large amounts of ABS/MBS and
private placements relative to surplus
Insurance guaranty fund system probably not adequate
for a major run or liquidity crisis
Life insurers owned by banks (and vice versa) could
add to fragility of banking system
Does Life Insurance Pose Systemic Risk?
 Why






LI may NOT be systemically risky
Life insurance sector not involved in payments system,
liquidity creation, credit creation, etc.
Life insurers own only small proportion of stocks and
bonds in the economy (about 6%)
Life insurance is a small proportion of household
financial assets (about 3%)
Many substitutes exist for life insurance policies
Life insurers not major employers (< 2% of non-farm
civilian labor force)
Disappearance of the entire sector would be tragic but
sustainable
Systemic Risk In Insurance: Non-Core Activities
 As
AIG debacle shows, the main systemic risk
posed by the insurance industry comes from
insurer participation in “banking” activities, e.g.,
credit default swaps (CDS) and other derivatives
 Swiss Re data shows that insurers and reinsurers
accounted for 33% of CDS market in early 2000s
 As with AIG, most insurers are not adequately
capitalized to sustain large CDS meltdown
 Insurance groups should required to increase
transparency of CDS operations

Tighter regulation of leverage at non-insurance subs
Overall Regulatory Implications

Regulators need to improve capabilities in
group supervision



Regulation of non-insurance subsidiaries to head off
future AIG-type crises
Improved measures of group level solvency risk
Regulators need to improve international
coordination of insurance supervision for multinational insurers

Coordinate national regulators & the International
Association of Insurance Supervisors
Systemic Risk and the
Interconnectedness Between Banks
and Insurers: An Econometric Analysis
Hua Chen, J. David Cummins, Krupa Viswanathan, and
Mary A. Weiss
Presented at: 2013 CICIRM
Kunming, China
July 18, 2013
Copyright J. David Cummins, 2013, all rights reserved. Not to be reproduced without author’s permission.
Prior Literature: Insurers & Banks
 Two
Prior Papers measure systemic risk in
banking and insurance using market data

Billio et al. (2012) – monthly stock returns
●
●
●

Hedge funds, brokers, banks, and insurance companies
Principal components and linear Granger causality test
Conclusion: All four sectors have become highly
interrelated in the past decade, increasing the level of
systemic risk in the banking and insurance industries
Acharya et al. (2010) – daily stock data
●
●
Systemic expected shortfall (SES) – propensity to be
undercapitalized when the system as a whole is
undercapitalized
Conclusion: 9 insurers among the top 50 systemic financial
institutions
Purpose of Our Paper
 Develop
and implement a robust systemic risk
measure for insurance
 Investigate interconnectedness between
banking and insurance during financial crisis
 We use CDS quotes and intra-day equity
returns to estimate systemic risk in the
insurance and banking industries
“Are insurers instigators or victims of
systemic risk?”
Purpose II
 Our


systemic risk measure relies on
Daily-frequency market price data for CDS (Markit)
Intra-day trading data on stock prices (TAQ)
 Systemic
risk measure is risk-neutral, forwardlooking and economically intuitive
 Direction of interconnectedness investigated


Linear and non-linear Granger causality
Correcting for heteroskedasticity
Contribution to Literature
 First
paper to use data on CDS spreads and intraday stock prices to study systemic risk for the
insurance industry
 Different econometric methodology than Billio et
al. (2012) and Acharya et al. (2010)
 New evidence on whether insurers are victims or
sources of systemic risk
Measuring Systemic Risk
 Two
major components that determine risk profile
of sample firms:


Probability of default of each insurer (based on CDS
premiums – Markit.com)
Default correlation (estimated indirectly from underlying
equity return correlation -- TAQ)
 Measure
of systemic risk uses portfolio credit risk
methodology (developed by Huang et al., 2009)
Distress Insurance Premium (DIP)
A Measure of Systemic Risk

Estimate forward-looking, risk-neutral indicator of systemic
risk of insurance industry:
“price of insurance against financial distress
(DIP)”

Define financial distress by choosing a threshold (e.g.,
15%) such that the ratio of portfolio credit losses to total
liabilities of the insurance sector is equal to or above
threshold
A Measure of Systemic Risk II



Construct a hypothetical portfolio
 Consists of debt instruments issued by the sample
banks/insurers, weighted by the liability size of each
firm.
Conduct Monte Carlo simulation (Tarashev and Zhu 2008)
 Probability of joint default (PD)
 Loss given default (LGD)
Systemic risk measure: the price of insurance against
financial distress
A Measure of Systemic Risk III

Systemic risk measure is calculated as risk-neutral
expectation of portfolio credit losses that reach at
least a minimum share (15%) of sector’s total liability
SRtINS = systemic risk measure for insurance industry
Lt = portfolio credit losses
TLt = total liability of insurance sector at time t
A Measure of Systemic Risk IV

Similar estimation procedure is performed for firms in
banking sector to obtain SRtBANK

SRtINS and SRtBANK used to analyze degree of
interconnectedness between insurance and banking
industries.
A Measure of Systemic Risk V
 Advantage
of method is that does not require
large sample of firms
 Huang
et al. (2009) – 12 banks
 Conclusions
apply to relatively large firms since
they have traded CDS
 Large insurers have lower default probabilities
than smaller insurers so results apply more
strongly to small insurers
Granger Causality Tests
 Testing
Granger causality involves using F-tests
to determine whether lagged information on a
variable X provides any significant information
about a variable Y in the presence of lagged Y.

If not, then X does not Granger-cause Y
 Linear
Granger causality tests conducted first
 Then do nonlinear Granger causality tests


Nonlinear Granger causality test uses the residuals
from the linear causality test
Then do Hiemstra-Jones (HJ) Test on residuals
Data and Systemic Risk Measures
 Sample




Sample of banks and insurers (Markit, SIC code)
Check whether the firm is publicly traded on a US
exchange (TAQ)
Focus on 5-year, Senior, No Restructuring CDS
Quotes on Friday
Fill in missing values
●
●
●

Selection
Use other quotes on the same day for conversion
Trace back one (two,…, five) day(s) before
Interpolation
Determine a common time period
 Our
sample: 11 insurers and 22 banks with CDS
quotes over the period Feb 2002 to May 2008
Sample Firms
Average Probability of Default
Default probability begins to spike in 3rd
quarter of 2007.
0.017
0.016
0.015
0.014
Weighted Average PD
0.013
0.012
0.011
0.010
0.009
0.008
0.007
0.006
0.005
0.004
0.003
0.002
0.001
2
0
0
2
Q
1
2
0
0
2
Q
2
2
0
0
2
Q
3
2
0
0
2
Q
4
2
0
0
3
Q
1
2
0
0
3
Q
2
2
0
0
3
Q
3
2
0
0
3
Q
4
2
0
0
4
Q
1
2
0
0
4
Q
2
2
0
0
4
Q
3
2
0
0
4
Q
4
2
0
0
5
Q
1
2
0
0
5
Q
2
2
0
0
5
Q
3
2
0
0
5
Q
4
2
0
0
6
Q
1
2
0
0
6
Q
2
Date
group
Bank Group
Insurance Group
2
0
0
6
Q
3
2
0
0
6
Q
4
2
0
0
7
Q
1
2
0
0
7
Q
2
2
0
0
7
Q
3
2
0
0
7
Q
4
2
0
0
8
Q
1
2
0
0
8
Q
2
2
0
0
8
Q
3
Systemic Risk Measure
 After
measuring probabilities of default and asset
return correlations, the systemic risk measure can
be computed for each week
 Systemic risk measure represents a weekly price
of insurance against distressed losses over the
following three months.
 To make comparisons, unit price of insurance =
ratio of nominal price to total liabilities of sample
firms in each sector
Linear Granger Causality Tests
Insurance Premiums SRtINS and SRtBANK
are not stationary
 Distress
 Differenced
 Linear
time series are stationary
causality tests are performed on the
differenced series
Linear Granger Causality Tests
 The
results imply that systemic risk of insurers
Granger-causes systemic risk of banks
 Also, systemic risk of banks Granger-causes
systemic risk of insurers
 However, results of BDS tests indicate that
nonlinearities are present in the univariate
systemic risk measures for both banks and
insurers
 Therefore, we must conduct nonlinear Granger
causality tests
BDS = Brock-Dechert-Scheinkman.
Control for Conditional Heteroscedasticity
 If
conditional heteroskedasticity exists, causality
test results can be biased
 Residuals from Granger-causality tests reveal


Little autocorrelation
Conditional heteroskedasticity exists
 Therefore,
use GARCH (1,1) model to assess
whether the bi-directional causality changes
 Re-do linear and nonlinear Granger tests using
GARCH
Non-Linear GARCH Models: Main Results
 Nonlinear
effect of insurers on banks is highly
significant at 1 lag

Significance fades after 3 lags
 Banks
in contrast have persistent predictive
power on insurers up to 5 lags


Systemic risk of banks has longer duration of
impact on insurers
Impact is also stronger than insurer effect on banks
Interconnectedness: Stress Testing
 Stress
testing conducted to study the impact of
systemic risk movements in the banking sector on
the insurance sector and then vice versa
A
hypothetical shock in systemic risk of banking
sector is fed into GARCH regression to generate
future dynamic movements of systemic risk in the
banking and insurance sectors

Shocks of 5%, 10%, 15% and 20% are applied
 Systemic
risk of insurance sector fed into GARCH
model as well to generate future dynamic
movements of systemic risk
Inter-Sector Impact of 20% Systemic Shock
Banks to Insurers
Insurers to Banks
14%
12%
10%
8%
6%
4%
2%
0%
-2%
T=1 T=2 T=3 T=4 T=5 T=6 T=7 T=8 T=9 T=10T=11T=12
Weeks After Shock
Non-Linear Tests: Conclusions
 Banks
create economically significant
systemic risk for insurers but not vice versa

Based on linear and non-linear Granger causality
tests correcting for heteroskedasticity
 Therefore,
insurers seem to be victims of
systemic risk rather than instigators
 Banks
are instigators of systemic risk
Systemic Risk: Policy Implications
Regulators should focus on banks to
prevent/ameliorate systemic shocks from banks
 Regulators should focus on non-core rather
than insurance activities of large insurers
 Insurance regulators should focus on mitigating
effect of shocks from banks (e.g., investment
restrictions and tighter capital requirements for
life insurers)

Thank you!
Further Information










American International Group, 2009, AIG: Is the Risk Systemic? Powerpoint presentation (New
York).
De Bandt, Olivier and Philipp Hartmann, 2000, Systemic Risk: A Survey (Frankfurt, Germany:
European Central Bank).
Geneva Association, 2010, Systemic Risk in Insurance: An Analysis of Insurance and Financial
Stability (Geneva, Switzerland).
Group of 10, 2001, Report on Consolidation in the Financial Sector
Harrington, Scott E., 2009, “The Financial Crisis, Systemic Risk, and the Future of Insurance
Regulation,” Journal of Risk and Insurance 76: 785-819.
Kaufman, George G., 1996, “Bank Failures, Systemic Risk, and Bank Regulation,” The CATO
Journal 16: 17-45.
Kaufman, George G., 2000, “Banking and Currency Crises and Systemic Risk: Lessons from Recent
Events,” Federal Reserve Bank of Chicago Economic Perspectives 24: 9-28.
Swiss Re, 2003, Reinsurance – A Systemic Risk, Sigma No. 5/2003 (Zurich, Switzerland).
World Economic Forum, 2009, Global Risks 2009 (Geneva, Switzerland).
Cummins, J. David and Mary A. Weiss, 2012, “Systemic Risk and the U.S. Insurance Sector,”
working paper, Temple University, Philadelphia.
My Research on Systemic Risk





Cummins, J. David and Mary A. Weiss, 2013, “Systemic Risk and the
Insurance Industry,” forthcoming in Georges Dionne, ed., Handbook of
Insurance, 2d ed. (Springer).
Chen, Hua, J. David Cummins, Krupa Viswanathan, and Mary A.
Weiss, 2013, “Systemic Risk and the Inter-Connectedness between
Banks and Insurers: An Econometric Analysis,” forthcoming, Journal of
Risk and Insurance.
Chen, Hua, J. David Cummins, Krupa Viswanathan, and Mary A.
Weiss, 2013, “Systemic Risk Measures in the Insurance Industry: A
Copula Approach,” working paper, Temple University, Philadelphia
Cummins, J. David and Mary A. Weiss, 2013, “Systemic Risk and the
Regulation of the U.S. Insurance Industry,” working paper, Temple
University, Philadelphia.
Cummins, J. David and Mary A. Weiss, 2012, “Systemic Risk and the
U.S. Insurance Sector,” working paper, Temple University,
Philadelphia.
To obtain the papers, please email: [email protected].
Systemic Risk:
Reinsurance Counterparty
Exposure
Reinsurance Exposure:
Recoverables from Non-Affiliates, P-C Insurers
 Reinsurance
Recoverables/Policyholders Surplus
From Non-affiliated Reinsurers






Ceded paid losses
Ceded unpaid losses
Ceded IBNR
Ceded unearned premiums
Ceded commissions
Minus funds held from reinsurers
 Normal
range: 50% to 150%
 Un-weighted average for groups = 43% (2008)
Reinsurance Exposure:
Ceded Reinsurance Leverage from Non-Affiliates
 Ceded
Reinsurance Leverage From NonAffiliated Reinsurers/Policyholders Surplus




Reinsurance recoverables
Ceded balances payable
Ceded premiums written
Minus funds held from reinsurers
 Normal
range: Component of gross leverage.
Gross leverage range is 5 to 7
 Un-weighted average for groups = 70% (2008)
Reinsurance Recoverables/Surplus: Groups
For 26% of groups, non-affiliate recoverables > 50% of surplus
> 100
% of Surplus
51 to 100
41 to 50
31 to 40
21 to 30
11 to 20
1 to 10
0%
5%
10%
15%
20%
Percent of Groups
25%
Reinsurance recoverables from non-affiliated reinsurers, including ceded paid losses, unpaid losses, IBNR losses, unearned
premiums and commissions less funds held from reinsurers. Source: Best’s Key Rating Guide, 2009. Data are for 2008.
30%
Reinsurance Receivables/Surplus: Industry
Non-affiliate receivables = 28% of industry surplus in 2008.
40%
Receivables/Surplus
38%
36%
34%
32%
30%
28%
26%
24%
22%
20%
2004
2005
2006
2007
2008
Reinsurance recoverables from non-affiliated reinsurers, including ceded paid losses, unpaid losses, IBNR losses, unearned
premiums and commissions less funds held from reinsurers. Source: Best’s Key Rating Guide, 2009. Data are for 2008.
Reinsurance Leverage/Surplus: Groups
For 30% of groups non-affiliate leverage > 75% of surplus.
> 200%
176-200%
151-175%
126%-150%
100%-125%
76%-100%
51%-75%
41%-50%
31%-40%
21%-30%
11%-20%
0-10%
0%
2%
4%
6%
8%
10%
Percent of Groups
12%
Reinsurance recoverables, ceded balances payable, and ceded premiums written less funds held divided by
policyholders surplus. Source: Best’s Key Rating Guide, 2009. Data are for 2008.
14%
16%
Reinsurance Leverage/Surplus: Industry
Non-affiliate leverage = 40% of industry surplus in 2008.
Reinsurance Leverage/PHS
55%
50%
45%
40%
35%
30%
2004
2005
2006
2007
Reinsurance recoverables, ceded balances payable, and ceded premiums written less funds held divided by
policyholders surplus. Source: Best’s Key Rating Guide, 2009.
2008
AIG: What Went Wrong?
US Life Insurers: 12 Month Change in Premiums
(as of June 30, 2009)
40
NY Life
30
20
Met
Pru
-50
-60
Source: A.M. Best Company.
AIG
Manulife
-40
Hartford
-30
Axa US
-20
NW Mutual
TIAA
ING USA
-10
Lincoln Fin
0
Aegon USA
% Change
10
AIG: What Went Wrong?
 AIG’s
traditional insurance operations did not
cause its meltdown
 AIG’s problems came from:

Credit default swaps out of AIG Financial Products
●
●

(Supposedly) regulated by Office of Thrift Supervision
US insurance regulators had no jurisdiction
Securities lending program of life subsidiaries
●
●
Indicates need for more regulatory scrutiny in the future
US regulators do have jurisdiction if lending is out of regulated
life insurance subsidiaries
AIG Revenues Before the Crash
Financial
Services
(AIGFP)
9%
Asset
Management
4%
US P&C
33%
Foreign P&C
11%
12 months ending 12/31/2006.
Foreign Life
& Retirement
29%
US
Life&Retirem
ent
15%
AIG’s Credit Default Swaps
 AIG
sold CDS contracts, mostly to European
banks (i.e., writing bond default insurance)
 Banks were using the swaps to reduce regulatory
capital, relying on AIGs overall credit rating
(regulatory arbitrage)
 AIG Financial Products had about $500 billion in
CDS outstanding but virtually no capital


AIG’s models supposedly showed that losses on the
CDS portfolio were virtually impossible
Losses due to model risk and managerial moral hazard
AIG’s Securities Lending Operation
 AIG
loans securities to broker dealer or bank
(e.g., to cover short selling or for diversification)
 Borrower posts collateral in form of cash or high
quality securities
 AIG reinvests the collateral and earns spread
between yield on invested assets and yield on
underlying securities
 AIG had $82 billion in liabilities for securities
lending as of year-end 2007, $69 billion in
August 2008
AIG Securities Lending: What Went Wrong



Declines in value of mortgages and other assets in 20072008 reduced value of reinvested collateral in securities
lending programs
Many of the counterparties in the securities lending
operation were the same institutions holding AIG CDS
As asset values declined, borrowers terminated the
securities lending arrangement to




Improve liquidity
Reduce exposure to AIG’s credit risk
At the same time, AIG had to post additional collateral for
the CDS transactions as underlying “insured” asset
values declined
Essentially, a “run on the bank” by AIG’s counterparties
The Bailout: Payments to AIG Counterparties
Counterparty
Goldman Sachs
CDS Trans. Asset Lending Total
8.10
4.80
12.90
11.00
0.90
11.90
Deutsche Bank
5.40
6.40
11.80
Barclays
1.50
7.00
8.50
Merrill Lynch
4.90
1.90
6.80
Bank of Amer
0.70
4.50
5.20
UBS
3.30
1.70
5.00
0
4.90
4.90
Others
14.70
11.60
26.30
Total
49.60
43.70
93.30
Societe Gen.
BNP Paribus
US Insurer Assessments vs. AIG
 Total
US life and P-C assessments: 1988-2010
$18.8 billion
 Federal assistance to AIG (as of June 30, 2009):
$136 billion


Not necessarily a net loss
But . . . .
 Total


assets of largest insurers: 2011
Met Life:
State Farm:
$612.8 billion
$135.2 billion
Source: A.M. Best Company, Harrington (2009).
Failure Costs: Met Life and State Farm
Cost: % of Ind Assets
(% of Life & PC Industry Assets, resp.)
6.0%
5.0%
4.0%
3.0%
2.0%
1.0%
0.0%
5% 10% 15% 20% 25% 30% 35% 40% 45% 50%
Impairment (% of Insurer Assets, 2011 Data)
Met Life
Source: A.M. Best Company, author’s calculations.
State Farm
What Is Systemic Risk Policy Trying to Prevent?
 Runs

on banks – traditional problem
Contagion – information asymmetries
 Banking

system collapse –
Continental bank
●
●
Correspondent bank collapse – counterparties
Jobs would be lost
 Threat
to settlement system
 Threat to payments system
 Fear that Infrastructure of short-term money
market and OTC derivatives would not handle
failure of significant counterparty

Cast doubt on soundness of other counter parties (e.g.,
case of Bear Stearns)
Systemic Risk Policy Is Trying to Prevent II
 Panic
due to loss of confidence
 Risks to system due to failure of “highly
interconnected” firms
 “Unpredictable consequences of a failure for
broader financial system”
 Reaction of counterparties of other firms that might
come under future government control

AIG was large, complex and interconnected whose
failure would impose losses on counterparties and also
endanger the entire world’s financial sector (BernankeMorehouse University)
Preview of Results
 After
adjustment for heteroscedasticity, impact of
banks on insurers found to be stronger and of
longer duration than impact of insurers on banks
 Stress tests indicate that banks create significant
systemic risk for insurers & not vice versa
 Insurers are primarily victims rather than
instigators of systemic risk
Measuring Systemic Risk: Methodologies

Huang et al. (2009): Distress Insurance Premium

Acharya et al (2010): Systemic Expected Shortfall

Adrian and Brunnermeier (2010): CoVAR

Zhou (2010): Multivariate extreme value theory

Billio et al (2012): Principal Components, Causality tests

Insurers can be a source of systemic risk to some
extent.
124
Nonlinear Granger Causality Tests II

Diks and Panchenko (2005, 2006)
 HJ test could produce spurious results in the presence
of conditional heteroskedasticity

Ross (1989), Andersen (1996)
 Volatility of a time series can measure the rate of
information flow.

Use the GARCH model to assess whether the bidirectional causality changes.
Sample Firms II
 Of
11 insurers, 8 are classified as property-liability
insurers
 Of remaining three,



Lincoln National had only life-health operations
MetLife group has property-liability operations
Prudential divested property-liability operations, but
owned these operations at the beginning of sample.