Download Economic Bulletin - May 2017

Economic Bulletin
May 2017
Lisbon, 2017 • www.bportugal.pt
Economic Bulletin | May 2017 • Banco de Portugal Av. Almirante Reis, 71 | 1150-012 Lisboa • www.bportugal.pt • Edition
Economics and Research Department • Design and printing Communication Directorate | Image and Graphic Design Unit •
Print run 25 • ISSN 0872-9794 (print) • ISSN 2182-0368 (online) • Legal Deposit no. 241772/06
Index
I The Portuguese economy in 2016 | 5
1. Overview | 7
Box 1 | Assessment of projections for 2016 | 10
2. International environment | 13
3. Monetary and financial conditions | 21
3.1. Euro area | 21
Box 3.1.1 | The impact of the Corporate Sector Purchase Programme | 24
Box 3.1.2 | Dispersion of interest rates in the euro area money market | 27
3.2. Portugal | 30
Box 3.2.1 | Market dynamics and the deleveraging of Portuguese firms | 39
4. Fiscal policy and situation | 43
Box 4.1 | Structural developments in tax revenue in 2016 | 50
Box 4.2 | Analysis of deviations in budget execution in 2016 | 52
5. Supply | 54
Box 5.1 | Capital per worker and productivity | 59
6. Demand | 62
Box 6.1 | Developments in unit values of Portuguese exports of goods | 71
7. Prices | 75
8. Balance of payments | 80
Box 8.1 | Direct investment flows into the Portuguese economy | 86
II Special issue | 91
Distribution mechanisms of monetary policy in the Portuguese economy | 93
I
The Portuguese
economy in 2016
1. Overview
2. International environment
3. Monetary and financial conditions
4. Fiscal policy and situation
5. Supply
6. Demand
7. Prices
8. Balance of payments
The Portuguese economy in 2016
1.Overview
In 2016 output growth in the Portuguese econ-
of firms has been important, especially for small
omy stood at 1.4 per cent, which compares with
firms and for those with activity in the construc-
1.6 per cent in the previous year. Albeit mod-
tion sector (Box 3.2.1. ‘Market dynamics and the
erate, these GDP developments had a marked
deleveraging of Portuguese firms‛). Although the
intra-annual profile, accelerating strongly in the
exit of firms is part of a natural process of restruc-
second half of the year, signalling that the recov-
turing and reallocation of productive factors in
ery process of the economy is expected to con-
the economy, the potential non-recovery of debts
tinue. In 2016 Portuguese output growth stood
raises difficulties to credit institutions, especially in
0.3 percentage points below that observed in
terms of profitability and impact on own funds.
the euro area, and its level is still 4 per cent below
that registered in 2008, the year that marked the
outbreak of the latest international economic
and financial crisis.
The continued reduction of the indebtedness
levels of the Portuguese economy requires the
creation of favourable conditions to economic
growth, against a background of social cohe-
In terms of the fundamental macroeconomic bal-
sion and sustained improvement of citizens’
ances, there was a slight increase in the current
living conditions. Investment is one of the key
and capital account surplus (0.5 p.p.) and a signifi-
variables for Portuguese economic growth. In
cant decline in the total general government defi-
2016 this variable declined in volume by 0.8 per
cit (2.3 p.p.), against a background of stabilisation
cent, after an increase of 4.7 per cent in 2015,
of the structural balance and decline in the struc-
although with an intra-annual recovery profile
tural primary balance (0.3 p.p.). The improvement
and an acceleration in the business compo-
in these balances is a prerequisite to ensure
nent. However, investment developments, are
macroeconomic stability and the credibility of the
insufficient, not only because the fall during the
Portuguese economy among international inves-
crisis period was very sharp, but also because
tors, and should therefore be a guideline for eco-
capital per employee in the Portuguese econ-
nomic policy decisions adopted at the national
omy is still at low levels, when compared with
level. In effect, the high indebtedness levels pre-
the euro area average (Box 5.1 ‘Capital per
vailing in the different institutional sectors make
worker and productivity‛). The low capital levels
the Portuguese economy particularly vulnerable
contribute to the under-performance of labour
to any developments implying an increase in the
productivity in the Portuguese economy, which
interest rates applied to external financing.
declined again in 2016.
The deleveraging process of the Portuguese
The resident population and the labour force con-
economy continued in 2016, reflecting a decline
in households’ and corporations’ indebtedness
ratios, which nonetheless continued to be very
high. Public debt ratio, in turn, virtually stabilised
in net terms, in parallel with a reduction of external indebtedness in the total economy. In this
context, there was an improvement in the international investment position of 7 percentage
points (p.p.) of GDP, which nonetheless stands
at a negative level of 105 per cent of the product.
tinued to decline in 2016. These declines maintain the trend observed since 2011, although
slightly less marked than in previous years. In
2016 resident population in the 25-34 age group
declined by 2.5 per cent, while the labour force
declined by 2.7 per cent. The sharp demographic
trends in the Portuguese economy are not easily
revertible and are aggravated by migration phenomena. Although the ageing of the population
is evident in several European countries, this situ-
In the case of non-financial corporations, the con-
ation is an important constraint to Portuguese
tribution to the decline in debt due to the exit
economic growth and puts greater pressure on
7
8
BANCO DE PORTUGAL • Economic Bulletin • May 2017
the dynamics of public expenditure with pensions
non-resident entities, almost exclusively focus-
and health.
ing on the issue of debt securities. Therefore,
In 2016 labour market developments were characterised by an increase in employment, exceeding that of Gross Value Added and maintaining
the recovery profile observed since the second
it was possible to offset the new decline of the
stock of bank loans granted by resident banks,
which indicates an alternative means to finance
investment projects.
quarter of 2013. Moreover, although remain-
The less buoyant economic activity in 2016 also
ing at very high levels, the unemployment rate
reflected the deceleration in exports of goods
declined by 1.3 p. p., in a context of stronger
and services, against a background of lower
wage dynamics than in past years. This may have
growth of the external demand directed towards
contributed to the slight increase in the inflation
the Portuguese economy. The deceleration in
rate in the course of 2016, via an increase in unit
exports of goods and services in 2016 reflected
labour costs. The inflation rate in Portugal, meas-
a lower growth of exports of non-tourism ser-
ured by the change in the Harmonised Index of
vices and chiefly the fall in exports of energy.
Consumer Prices, stood at 0.6 per cent in 2016,
with zero price growth in goods and 1.5 per cent
price growth in services.
The fall in investment contributed to less robust
growth of domestic demand than that observed
in 2015. Also, private consumption increased
slightly less than in the previous year, as a result
of less buoyant consumption of non-food current goods and durable goods. In turn, public
consumption, determined by the dynamics of
compensation of employees and consumption
of goods and services in the general government,
increased slightly less in 2016 than in 2015.
Developments in investment and consumption
took place in a context of gradual improvement
of the financing conditions of private economic
agents, in line with the maintenance of an expansionary stance of the monetary policy in the euro
Excluding this category, growth of exports of
goods was 1.8 percentage points higher than in
2015. In turn, the buoyancy of exports of tourism services continued to be high, in tandem
with the significant increase in nights spent by
non-residents in Portuguese hotels.
The performance of Portuguese exports in 2016
translated into a new gain in market share, which
underlines the competitive capacity of Portuguese firms in international markets. The stance
of the Portuguese economy, gradually turning
towards the production of goods and services
tradable abroad is a very positive feature of the
adjustment in the productive structure, which
was started before the latest international economic and financial crisis. Only by strengthening
its export capacity may the Portuguese economy accommodate growth of imports resulting
from a desirable strengthening of the invest-
area. This improvement, visible in the course
ment rate and an increase in consumption lev-
of 2016, was reflected in a decline of the inter-
els. The re-emergence of current account defi-
est rate spread between Portugal and the euro
cits, even if financed via foreign investment, may
area, for both corporations and households. New
be perceived as a return to imbalances, which
loans to households reinforced the growth trend
will tend to adversely affect refinancing condi-
observed in early 2016, especially in consump-
tions of existing debt. The relevance of exports
tion. Notwithstanding the increase in new credit,
performance for developments in the Portu-
the households’ deleveraging process continued.
guese economy and the prevailing high indebt-
The annual rate of change of total credit to
edness levels highlight the importance of devel-
corporations was also positive at the end of
opments in the international economy.
the second half of 2016. Relevant, however,
In 2016 world economic activity decelerated,
was the important contribution of credit from
reaching the lowest growth since the latest
The Portuguese economy in 2016
international economic and financial crisis. In
in goods and services, shall be complementary
turn, growth in the euro area was slightly lower
to internal debate on strengthening growth con-
than in 2015, with an increase in output more
ditions. The future of the Portuguese economy
based on domestic demand than in the past. In
depends heavily on its external partners, either
2016 financial market developments were con-
as destination of exports and source of produc-
ditioned by the UK referendum in June and the
tive investment, or within the scope of the ongo-
US presidential election in November, with vola-
ing financial integration process. This stresses
tility peaks and an increase in uncertainty follow-
the need to increase national agents’ standards
ing both events. Stock market indices, however,
regarding the improvement of the country’s com-
registered gains, chiefly in the second half of the
petitiveness. Only in this vein shall it be possible to
year and in the banking sector, partly reflecting
resume convergence towards European Union’s
the strong liquidity creation via non-standard
average levels of well-being.
monetary policies.
In a context of low economic growth, the existence of high volatility and strong valuations in
financial markets may lead to disturbances in
the international economy, which will tend to be
amplified in most indebted economies. In addition, the rise in inflation in the euro area from
mid-year onwards, triggered by energy prices, as
well as the increase in the federal funds rate target contribute to maintain concerns regarding
the development of the interest rate at which
the Portuguese economy is financed.
The Portuguese economy has shown a remarkable macroeconomic adjustment capacity and
a sectoral restructuring based on the internationalisation of Portuguese firms. Continuing
economic growth and increasing citizens’ wellbeing require the strengthening of firms’ productive investment and the availability of skilled
human resources, in a context of financial stability and legal and fiscal certainty. In this context, the obstacles created by the low savings
rates in the Portuguese economy and the unfavourable demographic developments, as well
as cyclical difficulties associated with the high
indebtedness levels and the elevated uncertainty in the international framework, require
a consistent and forward-looking approach to
economic policies.
The discussion on the deepening of European
integration, which has always been a catalyst for
reforms in the Portuguese economy, allowing for
gains through intensified capital flows and trade
9
10
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Box 1 | Assessment of projections for 2016
This box analyses the gap between Banco de Portugal’s projection for GDP in 2016 and the actual
figures, trying to identify the main factors explaining that gap. The analysis is based on the projection published in March 2016, which contains for the first time comprehensive information for
2015.1
The above-mentioned projections for the Portuguese economy are based on Banco de Portugal’s
quarterly econometric model (‘M’ Model). In addition to this main model, other satellite models
are considered, such as the MIMO model for inflation projection, and bridge models of shortterm activity developments.2 Forecasts depend on a range of assumptions regarding the external
environment and a set of rules for the incorporation of fiscal variables. In the first case, as usually
mentioned in the projection articles of the Economic Bulletin, assumptions correspond to those
of the Eurosystem’s projection exercises. Fiscal variables follow the rule used within the scope
of these exercises, i.e., incorporating the policy measures that have already been approved (or
that are highly likely to be approved) and that have been specified in detail. As mentioned in the
note published in March 2016, the projection incorporates most measures included in the State
Budget for 2016, which are sufficiently detailed.
Projection errors are the result not only of the fact that assumptions incorporated in the forecast
exercise have not materialised, but also of factors related to the model and judgement elements
incorporated in the projection.
Comparing the annual growth rate of real GDP in 2016 published in the Quarterly National Accounts
by Institutional Sector of 24 March with the projection published by Banco de Portugal in early 2016,
it may be concluded that the projection error was -0.1 p.p., i.e., Banco de Portugal projected GDP growth
at 1.5 per cent, while actual GDP growth was 1.4 per cent.
Based on elasticities implied in the ‘M’ model, it is possible to analyse the contributions to projection errors made by changes in assumptions, in particular financial conditions, foreign exchange
rates, and the international scenario (external demand, oil prices and external prices). In this
context, Table 1 presents the revisions of the external environment assumptions – defined as
the difference between the actual value and the value underlying the March exercise – and their
impact on the annual GDP growth rate in 2016.
Table 1 • Revisions on the external environment and its impact on the real annual growth rate
of GDP in 2016 | Revision on the annual growth rate in p.p., except interest rate, compared to the projection
of March 2016
Revisions on the external
environment
Impact on GDP
0.0
0.0
Euro effective exchange rate
-0.3
0.0
Competitors' prices in national currency
-1.0
0.0
Oil prices in euros
18.2
-0.1
External demand
-1.9
Interest rate
Impact of revisions on the external environment
-0.4
-0.5
Source: Banco de Portugal.
Note: A ‘-‛ sign in the effective euro exchange rate corresponds to a higher depreciation or a lower euro appreciation than expected in March.
The Portuguese economy in 2016
Expectations in early 2016 pointed to a more favourable international scenario than that subsequently observed. In particular, external demand for Portuguese goods and services was anticipated to have an annual average growth of 3.9 per cent, i.e., 1.9 p.p. above that actually registered
in 2016. This significant downward revision of external demand was due to both intra-euro area
countries and third countries (Section 2: ‘International environment’) and implied a downward
revision of the GDP growth rate projected at 0.4 p.p. for 2016. In addition, the technical assumption for developments in the effective exchange rate of the euro, which assumes the maintenance, over the projection horizon, of the average levels observed in the two weeks prior to the cut-off date, was revised slightly downwards, implying an appreciation of the euro slightly below that
expected in March 2016. This was reflected in a slight improvement in the price competitiveness
of the Portuguese economy against the main competitors located outside the euro area. There
was also a slight decline in competitors’ prices in external markets, with a negative impact on the
price competitiveness of Portuguese exports. The impact on the GDP projection of the revision of
exchange rates and external prices is negligible.
The deterioration of the international scenario in the course of 2016 was offset by a more favourable than anticipated behaviour of the export market share in 2016. The very significant gains of
the Portuguese export market share were registered in both goods and services, particularly in
the second half of the year (Section 6: ‘Demand’). Therefore, contrary to any conclusion that might
be withdrawn from the broadly downward revision of the international environment assumptions, exports of goods and services were higher than projected in the March exercise.
As regards monetary and financial conditions, the technical assumption for the 3-month Euribor
rate remained virtually unchanged from March projection (Section 3.1: ‘Monetary and financial
conditions in the euro area’). The increase in oil prices in the course of 2016 was higher than that
expected and implied an upward revision from the assumption considered in the March exercise,
with a negative impact of approximately 0.1 p.p. on the GDP growth rate. Therefore, taking into
account the multipliers implied in the ‘M’ model, revisions of the external environment led to a
significant downward revision of GDP by around 0.5 p.p.
Table 2 presents the projection errors in GDP and its components, as well as the respective gross
and net contributions of the imported content. Analysing the components net of imports, it may
be concluded that the projection error of -0.1 p.p. in GDP was due to the overestimation of the
public finance variables, in particular to the lower growth of public consumption and to the stronger fall of public investment (Box 4.2: ‘Analysis of deviations of budget execution in 2016‛). These
effects were partly offset by the underestimation of exports, largely reflecting the already mentioned unanticipated gains in market share. In spite of an underestimation of private consumption
of 0.3 p.p. in gross terms, the contribution of this component net of the imported content for the
GDP error was virtually zero, given that the projection error of this variable largely reflected car
purchases.
Moreover, the annual average growth rate of GDP and its components in 2016 implied a sharp
intra-annual profile from the first to the second half of the year. The first half of the year was
characterised by low growth, with quarter-on-quarter rates of change of 0.2 per cent, whereas in
the second half there was an acceleration in economic activity, with quarter on quarter rates of
change of 0.9 and 0.7 per cent in the third and fourth quarters respectively.
The projection published in March 2016 did not anticipate these intra-annual GDP dynamics,
and projected a smoother profile, with a slight acceleration in the second half of the year. In this
11
12
BANCO DE PORTUGAL • Economic Bulletin • May 2017
context, there was an overestimation of GDP and its main components in the first half of the year
and an underestimation in the following half year.
Table 2 • Projection error in real GDP growth rate and in major components in 2016
| Observed– projected in March 2016
GDP
Weights 2015 in %
GDP
Projection error
Gross contributions
to GDP growth
Net contributions
to GDP growth
-0.1
100.0
-0.1
-0.1
Private consumption
65.6
0.5
0.3
0.0
Public consumption
18.2
-0.6
-0.1
-0.1
Investment
15.5
-1.0
-0.2
-0.2
Exports
40.6
2.2
0.9
0.3
Imports
39.8
2.3
-0.9
Source: Banco de Portugal.
Note: The demand aggregates net of imports are obtained by subtracting an estimate of the imports needed to meet each component. The calculation
of import contents was based on data for 2005. For more information, see the Box entitled ‘The role of domestic demand and exports in economic
activity developments in Portugal’, in the June 2014 issue of the Economic Bulletin.
The Portuguese economy in 2016
2.International environment
World economic activity
and trade decelerated
further in 2016, dropping
to the lowest level since
the last international
economic and financial crisis
Taking the pre-crisis period as reference (2002-07),
international trade recorded an average growth
much higher than global activity (around 8 and
5 per cent respectively), while world trade and
GDP grew by around 3 per cent, on average, from
2012 to 2016, i.e. a decline in world trade elasticity of around 0.7 p.p. (Chart 2.2). This decrease in
In 2016 world economy decelerated, recording the lowest growth since the last international
economic and financial crisis (Table 2.1 and
world trade elasticity is expected to be associated,
inter alia, to slower developments in global value
chains and a geographical change in trade benefiting emerging market economies, which have
Chart 2.1). This slow recovery from the crisis
shown a slower response of trade to changes in
has been different across regions. In 2016 as a
output.
whole, contrary to the previous year, advanced
economies slowed down and emerging market
and developing economies decelerated slightly.
In 2016 developments in financial markets were
affected by the referendum in the United Kingdom in June and the presidential election in the
After recovering from its major collapse, world
United States in November, which were followed
trade growth has been weakening since 2012.
by volatility peaks and increased uncertainty.
Table 2.1 • Gross Domestic Product – Real year-on-year rate of change | Percentage
2002-07
2013
2014
2015
2016
World
4.8
3.4
3.5
3.4
3.1
Advanced economies
2.6
1.3
2.0
2.1
1.7
USA
2.7
1.7
2.4
2.6
1.6
Japan
1.4
2.0
0.2
1.2
1.0
Euro area
2.0
-0.2
1.2
1.9
1.7
Germany
1.4
0.6
1.6
1.5
1.8
France
1.8
0.6
0.7
1.2
1.1
Italy
1.1
-1.7
0.2
0.7
1.0
Spain
3.5
-1.7
1.4
3.2
3.2
2.7
1.9
3.1
2.2
1.8
Emerging and developing economies
7.2
5.1
4.7
4.2
4.1
Emerging and developing Europe
5.9
4.9
3.9
4.7
3.0
Commonwealth of Independent States
7.6
2.1
1.1
-2.2
0.3
7.1
1.3
0.7
-2.8
-0.2
United Kingdom
Russia
Emergind and developing Asia
China
Latin America and the Caribbean
Brazil
Sources: Eurostat, IMF and Thomson Reuters.
9.0
6.9
6.8
6.7
6.4
11.2
7.8
7.3
6.9
6.7
4.1
2.9
1.2
0.1
-1.0
3.9
3.0
0.5
-3.8
-3.6
13
14
BANCO DE PORTUGAL • Economic Bulletin • May 2017
In June, the result of the referendum raised con-
from July onwards, reflecting reduced concerns
cerns about the UK’s economic outlook, and, in
about the world economy outlook, the percep-
November, the result of the US elections sig-
tion that monetary authorities might be less
nalled shifting expectations regarding the new
reactive to the first signs of an inflation hike,
administration’s policies. In the first half of the
and the anticipation of a more expansionary fis-
year, long-term interest rates followed a down-
cal policy and an increase in investment in infra-
ward trend in the United States, the United King-
structures after the US elections.
dom and Germany, amid doubts about growth
At the start of the year, equity indices record-
in global economic activity, the situation in the
ed losses, given the more negative outlook for
euro area financial sector and continued (or
the world economy, particularly in China, which
strengthening) accommodative monetary poli-
were followed by a recovery, amid rising oil pric-
cies by major central banks (Chart 2.3). In con-
es. After reaching a trough of 28 USD/barrel in
trast, long-term interest rates began increasing
January, oil prices increased gradually, before
12.0
Chart 2.1 •
Gross Domestic
Product - annual
growth between
2010 and 2016
| Percentage
10.0
8.0
6.0
4.0
2.0
0.0
-2.0
World
Advanced
economies
Euro area
US
Emerging economies
China
Pre-crisis growth (2002-2007)
| Percentage and
percentage points
13.0
2.6
7.0
2.1
1.0
1.6
-5.0
1.1
-11.0
1980
1984
1988
1992
GDP growth
Pre-crisis GDP average growth
Trade elasticity (rhs)
1996
2000
2004
2008
2012
Trade growth
Pre-crisis trade average growth
Sources: IMF and Banco de Portugal calculations.
Note: Trade elasticity is calculated over 5-year periods. Pre-crisis period: 2002-2007.
2016
0.6
Percentage points
Chart 2.2 •
Growth of world
GDP and trade
volumes
Percentage
Sources: Eurostat, IMF and Thomson Reuters.
The Portuguese economy in 2016
stabilising at around 50 USD/ barrel at the end
Advanced economies grew by 1.7 per cent
of the summer. At the end of the year, the stock
in 2016, compared with 2.1 per cent in 2015,
market in the United States reached historical
improving in the second half of the year. For
highs due to the anticipation of expansionary
most economies, domestic demand and, in par-
policies, which influenced the various sectors,
ticular, private consumption continued to be the
in particular banks (Chart 2.4).
drivers of growth. Investment, as measured by
15
gross fixed capital formation (GFCF), continued
to grow moderately and accelerated in the sec-
Advanced economies grew
less strongly, while emerging
economies accelerated slightly,
despite differences across
countries
ond half of 2016, making a marginal contribution to GDP.
In the United States, economic activity slowed
down in 2016, with GDP growing by 1.6 per cent,
i.e. 1 p.p. less than in the previous year. The US
6
Chart 2.3 •
10-year sovereign
debt rates
5
4
| Percentage
3
2
1
0
-1
Jan. 11
Jan. 12
Jan. 13
Euro area
Jan. 14
US
Jan. 15
Jan. 16
UK
Japan
Sources: Thomson Reuters and ECB.
200
175
125
Chart 2.4 •
Major stock
markets indexes,
total and banks
100
| Index Jan. 2014=100
150
75
50
25
Jan. 11
Jan. 12
Eurostoxx
Banks: euro area
Source: Thomson Reuters.
Jan. 13
Jan. 14
S&P500
Banks: US
Jan. 15
Jan. 16
Footsie
Banks: UK
16
BANCO DE PORTUGAL • Economic Bulletin • May 2017
economy recorded weak growth during the first
to reinforce the pass-through of monetary policy
half of the year, largely due to developments in
to the real economy.
private business investment, recovering in the
Japan gradually recovered throughout 2016,
second half of the year owing to private consumption. At the same time, labour market conditions remained favourable, with the unemployment rate remaining below 5 per cent, for
the first time since 2007. Inflation increased, in
particular from the middle of the year onwards,
due to an increase in commodity prices, but
continued below the Federal Reserve’s target.
In terms of the average annual rate of change,
inflation reached 1.3 per cent (0.1 per cent in
2015), while inflation excluding food and energy
remained slightly above 2 per cent. Against this
background, in December 2016, the Federal
Reserve decided to increase the target for the
federal funds interest rate to a 0.50-0.75 per
cent range.
The United Kingdom grew above the forecasts
released after the June referendum, which resulted in the United Kingdom leaving the European
Union. Economic activity grew by 1.8 per cent in
2016, i.e. 0.4 p.p. less than in 2015, supported by
private consumption, which, contrary to expectations, did not show signs of slowing down.
However, in the next few years, real household income may decelerate, negatively affecting private consumption, amid moderating wage
growth and higher import prices associated
with the depreciation of the pound sterling. The
year-on-year rate of change in the Harmonised
Index of Consumer Prices (HICP) remained below
2 per cent, but increased over the year, reaching
1.6 per cent (0.2 per cent in December 2015), as
a result of the depreciation of the pound sterling, which was more pronounced immediately
after the referendum’s unexpected result. In this
owing to domestic demand and supported by
accommodative monetary and fiscal policies
and favourable financial conditions. The Japanese GDP grew by 1.0 per cent, compared with
1.2 per cent in 2015. In 2016, the Bank of Japan
made two changes to its Quantitative and Qualitative Monetary Easing programme, by introducing a negative interest rate to banks’ current accounts in January, and by conditioning
the purchase of Japanese government bonds, in
September, so that 10-year interest rates would
remain close to zero. In addition, the Bank committed itself to expanding the monetary base
until the observed inflation rate exceeds the target (inflation-overshooting commitment).
The situation of emerging market and developing economies continued to be quite mixed. In
China, economic activity remained robust in
2016, with GDP growing by 6.7 per cent yearon-year (6.9 per cent in 2015), benefiting from
the economic policy stimulus. Nevertheless, economic activity was weaker than expected in a
number of Latin American countries, which have
been in a recession since 2015, specifically Brazil. In Russia, economic growth was slightly better
than expected, partly owing to developments
in oil prices.
Euro area economic activity
continued to recover
gradually, with a greater
support from domestic
demand than in the past
context, in August, the Bank of England adopt-
Euro area GDP grew at a relatively stable rate
ed a package of measures to support the econ-
throughout the year. Growth of 1.7 per cent
omy, which included a cut in the Bank Rate, an
reflected a positive contribution from domes-
increase in the stock of purchased UK govern-
tic demand and a negative contribution from
ment bonds, a programme to purchase non-
net exports. Above all, this pattern was driven
financial investment-grade corporate bonds and
by consumption, associated with favourable
a Term Funding Scheme for eligible institutions
developments in households’ real disposable
The Portuguese economy in 2016
income, amid growing employment and declin-
Unemployment continued to decline in the euro
ing oil prices. The recovery in activity in the cur-
area, with some cross-country differences. The
rent economic cycle has a greater support from
larger economies experienced declines, with
domestic demand than in the previous cycle,
the exception of Italy. In parallel, employment
with services playing a greater role (Chart 2.5).
continued to grow moderately, above the lev-
The recovery seen after 2013 is also more broad-
els observed in 2015.
17
ly-based across countries, with particularly positive developments in Spain (3.2 per cent growth)
(Chart 2.6). GDP grew by 1.8 per cent in Germany and slightly above 1 per cent in France. In
External demand for Portuguese
goods and services decelerated
turn, GDP growth stood at 1.0 per cent in Italy,
External demand for Portuguese goods and
standing considerably below pre-crisis levels.
services decelerated in 2016, growing by 2 per
Chart 2.5 • Euro area two latest economic recoveries – contributions by sector | Index 2009 Q2=100
Contributions to GDP
Contributions to GVA
110
110
110
108
108
106
106
104
104
108 110
106 108
102
102
100
100
9898
2009
2010
2012
2013
2009
Q2Q2 2010
Q4Q4 2012
Q2 Q2 2013
Q4 Q4 20152015
Q2 Q220162016
Q4 Q4
Domestic
demand
Domestic
demand
NetNet
Exports
Exports
GDPGDP
cumulated
growth
sincesince
20092009
Q2 Q2
cumulated
growth
104 106
104
102
102
100
100
98
98 Q2
2009
2009 Q2
2011 Q2
2011 Q2
Industry
2013 Q2
2013 Q2
2015 Q2
2015 Q2
Industry
Services
Services
Construction
Construction
Gross
value added cumulated growth since 2009Q2
Gross value added cumulated growth since 2009Q2
Sources: Eurostat, CEPR and Banco de Portugal calculations.
Note: Two latest economic recoveries - since the 2009 Q2 and 2013 Q1 troughs, according to the CEPR Dating Commitee.
120
Chart 2.6 •
GDP evolution
since trough
115
110
| Index 2009 Q2=100
105
100
95
90
2009 Q2
2010 Q4
Euro area
2012 Q2
Germany
Sources: Eurostat, CEPR and Banco de Portugal calculations.
2013 Q4
France
2015 Q2
Italy
2016 Q4
Spain
18
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Increase in euro area
inflation from the middle
of the year
cent year-on-year, after around 4 per cent in
2015 (Table 2.2). These developments resulted from lower growth in demand from euro
area trade partners (3.5 per cent, compared
with 5.6 per cent in 2015), and a contraction
in external demand from economies outside
the euro area (-0.4 per cent, after 1.1 per cent
in 2015). When external demand from Angola
In 2016 developments in HICP inflation mostly
reflected the path of energy prices (Chart 2.7).
This influence translated into a low or nega-
is taken into account, the effect is even more
tive inflation in the first months of the year and
pronouced.3 External demand for goods and
subsequent increases, which were heightened
services is expected to have grown by 1.3 per
by the base effect of the falls observed in 2015
cent in 2016, compared with 2.6 and 5.4 per
and related to an increase in energy prices.
cent in 2015 and 2014 respectively.
Table 2.2 • External demand of goods and services – Real year-on-year rate of change
| Percentage
Weights(b)
2013
2014
2015
2016
100.0
1.9
5.0
3.9
2.0
66.3
0.9
5.2
5.6
3.5
External demand (ECB)(a)
Intra euro area external demand
of which:
Spain
27.1
-0.5
6.5
5.6
3.3
Germany
13.7
3.2
4.0
5.0
3.6
France
12.5
2.2
4.8
6.4
3.6
3.9
-2.3
3.1
6.7
3.1
33.7
3.7
4.6
1.1
-0.4
Italy
Extra euro area external demand
of which:
United Kingdom
5.6
3.4
2.5
5.5
2.8
USA
3.5
1.1
4.4
4.6
1.1
Goods and services imports from Angola (c)
9.0
11.4
-21.8
-27.8
Adjusted external demand (d)
2.3
5.4
2.6
1.3
Comércio mundial de bens e serviços (FMI)
3.7
3.7
2.7
2.2
Importações mundiais de mercadorias (CPB)
2.0
2.6
1.9
1.2
Memo:
Sources: ECB, CPB, IMF, Thomson Reuters and Banco de Portugal calculations.
Notes: (a) External demand is computed as weighted average of the imports volume of Portugal's main trading partners. Each country/region is
weighted by its share in Portuguese export. (b) Shares computed using 2012-2014 data. (c) The weight refers to the weight of nominal goods and
services exports to Angola on portuguese exports. (d) External demand indicator adjusted for the importance of the foreign trade with Angola.
Corresponds to the weighted average (by the exports weight) between the external demand indicator calculated by the ECB and the volume of the
goods and services imports of the Angolan economy. The IMF forecasts (Word Economic Outlook) for the growth of the volume of imports of Angola
in 2016 are used.
The year-on-year rate of change in HICP exclud-
countries. As in the past, differentials between
ing energy and food remained stable at relatively
groups of countries reflected divergences in
low levels, ranging between 0.7 and 1.0 per cent
inflation excluding energy and food (Charts 2.8
throughout the year (Chart 2.7). Since mid-2016,
and 2.9). At the end of the year, the year-on-year
the increase in euro area inflation was boost-
rate of change in HICP excluding energy and
ed by energy prices and broadly-based across
food stood at 1.4 per cent in Germany, 0.9 and
The Portuguese economy in 2016
19
1.5
Chart 2.7 •
Contributions
to year-on-year
HICP rate of change
in the euro area
1.0
0.5
0.0
| Percentage points
-0.5
-1.0
-1.5
Jan. 14
Jul. 14
Jan. 15
Jul. 15
Excl. food and energy
Energy
Jan. 16
Food
Jul. 16
Total
Source: Eurostat.
3
Chart 2.8 •
Inflation differentials
between high-rated
and low-rated
countries
2
1
0
| Percentage points
-1
-2
Jan. 99
Jan. 03
Jan. 07
Total
Jan. 11
Jan. 15
Excluding food and energy
Sources: Eurostat, ECB and Banco de Portugal calculations.
Note: High rated countries: Austria, Belgium, France, Finland and Germany. Low rated countries: Cyprus, Greece, Ireland, Italy, Portugal
and Spain.
4
Chart 2.9 •
Euro area HICP
3
| Year-on-year rate
of change
2
1
0
Jan. 99
Jan. 03
Jan. 07
High rated countries
Jan. 11
Jan. 15
Low rated countries
Sources: Eurostat, ECB and Banco de Portugal calculations.
Note: High rated countries: Austria, Belgium, France, Finland and Germany. Low rated countries: Cyprus, Greece, Ireland, Italy, Portugal
and Spain.
BANCO DE PORTUGAL • Economic Bulletin • May 2017
0.7 per cent in Spain and Italy respectively, and
in services prices in high rated countries and a
0.4 per cent in France.
slight acceleration in lower rated countries. Long-
Since 2013, the inflation differential between
term inflation expectations increased from September onwards, standing above the levels seen
tries has been negative, in contrast to the pre-
at the start of 2015, but below historical aver-
crisis period. This reflects lower growth in pric-
ages. Long-term inflation expectations seem to
es in the services component in lower-rated
have become rather more sensitive to changes
countries, associated with wage adjustments.
in short-term expectations and oil prices both in
Since the second half of 2015, the differential
the euro area and in the United States and the
has decreased, reflecting continued low growth
United Kingdom (Chart 2.10).
Chart 2.10 •
Euro area inflation
expectations and
oil prices
| Percentage and USD
dolars per barrel
3.0
175
2.5
150
2.0
125
1.5
100
1.0
75
0.5
50
0.0
Jan. 06
25
Jan. 08
1-y, 1-y ahead
Jan. 10
Jan. 12
Jan. 14
5-y, 5-y ahead
Sources: Thomson Reuters and Banco de Portugal calculations.
Note: Expectations implicit in inflation swaps. Vertical line indicates the APP anouncement.
Jan. 16
Oil price (rhs)
US dollars
high rated countries and lower rated coun-
Percentage
20
The Portuguese economy in 2016
3.Monetary and financial conditions
3.1. Euro area
The Eurosystem maintained
an accommodative monetary
policy in 2016
Euro area financial markets were influenced by
the accommodative monetary policy stance. The
downward path in euro area government bond
yields reflected the effect of the APP purchases
and other monetary policy measures taken by
the ECB. Overall effects were also relevant, in
2016 saw the adoption of non-standard mon-
particular an increase in uncertainty about the
etary policy measures on the part of the ECB in
outlook for global growth at the start of the year
March and December. At the start of the year,
and the result of the UK referendum in June.
economic and financial conditions deteriorated
From the summer onwards, a decline in con-
and downside risks to inflation increased, lead-
cerns about developments in the global econ-
ing the Governing Council to reduce the rates
omy and the incorporation of an inflation risk
on the main refinancing operations, the deposit
premium led to a partial reversal in the decline
facility, and the marginal lending facility to 0 per
seen in interest rates.
cent, -0.4 per cent and 0.25 per cent respectively,
Euro area equity indices improved slightly through-
and to adopt a number of other measures. This
package included expanding the monthly purchases under the Expanded Asset Purchase
Program (APP) from €60 billion to €80 billion,
and extending it until March 2017, and expanding the purchase programme to debt securities
issued by non-bank corporations established in
the euro area (Box 3.1.1: ‘The impact of the Corporate Sector Purchase Programme‛). The ECB
also announced a new series of targeted longer-term refinancing operations (TLTRO II), to
support lending to the economy. These operations are different from previous TLTROs mostly because the bank will pay a rate equal to the
deposit facility rate if it grants loans above a specific limit. At the end of the year, the Governing
Council considered there was a need for further
substantial monetary stimulus to the economy in
order to ensure a sustained adjustment in the
path of inflation towards the inflation aim. The
Governing Council therefore decided to extend
the APP until December 2017, but to reduce the
purchases from €80 billion to €60 billion.
Monetary and financial
conditions in the euro area
continued to evolve favourably
out the year, although the intra-annual pattern
showed marked movements, consistent with
the previously mentioned factors. Equity prices in
the European financial sector fell by around 8 per
cent in 2016, amid prospects of lower yields in the
future and the negative impact of non-performing
loans on bank profitability.
In foreign exchange markets, the euro appreciated slightly, in effective terms, in the course of
the year, reflecting in particular the result of the
UK referendum, the world economic outlook
and monetary policy differences across jurisdictions. The effective appreciation of around
2 per cent was the result of an appreciation of
17 per cent against the pound sterling and
4 per cent against the Chinese renminbi, and a
depreciation of 3 per cent against the US dollar
and 6 per cent against the Japanese yen. Money
market interest rates continued to decline, but,
since the summer of 2016, the differential has
been widening between the EONIA and deposit
facility rates vis-à-vis the secured money market rates and rates on government debt, which
might indicate a less effective transmission of
monetary policy decisions (Box 3.1.2: ‘Dispersion of interest rates in the euro area money
market‛).
21
22
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Firms’ financing costs continued to decline in
Against this background, the net issuance of euro-
2016. In particular, the negative deposit facil-
denominated bonds increased considerably from
ity rate and the new TLTRO II announced in
the second quarter of 2016 onwards (Chart 3.2).
March are expected to have helped reduce the
The cost of bank loans to firms decreased by
interest rates on bank loans to firms. In addition,
around 30 b.p. throughout the year, slightly
the introduction of a purchase programme for
below the decline seen in loans to households
investment-grade bonds issued by corporations
for house purchase, standing at around 1.8 per
established in the euro area might also partly
cent in both cases. In addition, the differential
explain the decline in the respective interest
between the cost of loans to firms in lower rated
rates, which have stood at levels similar to those of
and high rated countries returned to the levels
highly-rated banks’ bonds since June (Chart 3.1).
observed at the start of 2011.
6
Chart 3.1 •
NFC's financing
costs
| Percentage
5
4
3
2
1
0
Jan. 11
Jan. 12
Jan. 13
Jan. 14
High rated NFCs bonds
NFC loans below 1 million
Jan. 15
Jan. 16
NFC loans over 1 million
High rated Banks' bonds
Sources: ECB, Bank of America Merrill Lynch, Blooomberg and Banco de Portugal calculations.
Note: Vertical line indicates the announcement of CSPP and TLTRO II.
Chart 3.2 •
External financing
of NFC in
the euro area
275
15
225
10
175
5
| Billion euros
125
Jan. 11
0
Jan. 12
Jan. 13
Bank loans (new operations)
Source: ECB.
Note: 12-month moving average.
Jan. 14
Jan. 15
Jan. 16
Net issues of bonds in euros (rhs)
The Portuguese economy in 2016
Bank lending to firms grew considerably in the
improvement in lending conditions for firms and
first half of the year both for high rated countries
households in 2016. Banks have also reported
and lower rated countries (Chart 3.3). In annu-
an increase in demand for loans to both firms
al average terms, loan growth stood at 2.3 per
and households, which was broadly-based across
cent in December 2016, compared with 0.5 per
all loan categories. Banks have used the liquidity
cent in December 2015, accelerating 1.4 p.p. up
obtained from the APP for granting loans, for refi-
to June. Loans to households recovered more
nancing purposes and, to a lesser extent, for pur-
slowly, particularly in high rated countries, from
chasing assets. As regards the negative deposit
1.4 to 2.0 per cent during the same period. The
facility rate, banks have indicated a positive impact
improvement in financial and monetary condi-
on lending volumes, but a negative impact on net
tions is in line with the results of the euro area
interest income and loan margins.
23
Bank Lending Survey, which have indicated an
6.0
Chart 3.3 •
Anual rate
of change
of bank loans
in the euro area
3.0
0.0
| Percentage
-3.0
-6.0
-9.0
Jan. 11
Jan. 12
Jan. 13
High rated countries
Jan. 14
Low rated countries
Jan. 15
Jan. 16
Euro area
Source: ECB.
Note: Loans adjusted for sales and securitization. Full lines represent firms and dashed lines represent families.
24
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Box 3.1.1 | The impact of the Corporate Sector Purchase Programme
The Corporate Sector Purchase Programme (CSPP) was announced by the Governing Council of
the European Central Bank on March, 10 2016 as an extension of the Expanded Asset Purchase
Programme (APP)4 to the purchase of debt securities issued by non-financial corporations established in the euro area. The programme aims to further strengthen the pass-through of the
Eurosystem’s asset purchases to the financing conditions of the economy, to provide further
monetary policy accommodation, and to contribute to a return of inflation rates to levels below,
but close to, 2 per cent in the medium term. As the CSPP is part of the APP, these purchases are
expected to continue until the end of 20175 or beyond, if necessary, and in any case until the Governing Council of the ECB sees a sustained adjustment in the path of inflation consistent with its
inflation aim. This box describes the CSPP and analyses its impact on the funding costs of euro
area non-financial corporations and on their funding structure.
Following the announcement of the programme, the technical parameters were published on
April, 21 2016 and purchases began on June 8. Eligible bonds under the CSPP have a minimum
remaining maturity of 6 months and a maximum remaining maturity of 30 years and 364 days at
the time of the purchase. The lower limit was established in order to ensure that bonds issued by
smaller firms, usually with shorter maturities, were included in the purchases. Moreover, bonds
must be denominated in euros and they must have high credit quality (rating of BBB or higher), obtained from an external credit assessment institution. Finally, bonds must be marketable
instruments eligible as collateral for Eurosystem liquidity-providing operations. Bonds must be
issued by an institution established in a euro area Member State which is not a credit institution, does not have a parent undertaking which is a credit institution, and is not an eligible issuer
under the public sector purchase programme (PSPP) in secondary markets. When the technical
parameters of the programme were announced, the purchase of bonds with negative yields was
permissible provided these were above the deposit facility rate. At the December 2016 meeting,
the Governing Council of the ECB announced that purchases below the deposit facility rate would
also be permitted.
In order to limit risk exposure and to ensure the diversification of the Eurosystem’s portfolio,
a number of restrictions are applied on the amounts and on issuer entity’ groups. Regarding the
amount, the purchase of any type of bond may not exceed 70 per cent of its outstanding amount.
Nevertheless, this limit may be lower for bonds issued by public undertakings. Lastly, the programme is coordinated by the ECB, but only six Eurosystem central banks carry out the purchases in both primary and secondary markets. Each central bank is responsible for purchases of a
particular geographical allocation of the euro area and is responsible for publishing a list of all
purchased bonds at the end of each week, as these will become available in the securities lending market.
At the end of 2016, the Eurosystem had purchased a total of €51,069 million under the CSPP, of
which 86.4 per cent were purchased in the secondary market and 13.6 per cent in the primary
market. At that time, the CSPP accounted for 3.3 per cent of total purchases under the APP. Even
though the programme is small in volume, its importance within the APP has been increasing
since its implementation. This is particularly relevant in a context where, historically, firms tend to
fund themselves through bank loans, as is the case of the euro area. Overall, the CSPP seems to
have helped reduce debt issuance financing costs of non-financial corporations and to increase
the use of this type of funding.
The Portuguese economy in 2016
A preliminary assessment shows that apparently the programme had some impact on corporate
bond yields and on the yields of bonds issued by other sectors. Economic literature assessing the
APP has identified the impact of its announcement as the most significant event. The same seems
to apply to the CSPP. The first vertical line in Chart 1 marks the date of the CSPP announcement
and it is clear that, at that moment, bond yields dropped considerably for both non-financial corporations and banks. In the case of non-financial corporations, this downward trend continued,
being only interrupted by some volatility in May and June, as a consequence of the outcome of the
UK referendum on leaving the European Union. Following the start of the programme (second
vertical line in Chart 1), developments in yields resumed the downward trend. Simultaneously,
interest rates on long-term government bonds remained fairly constant, contributing to the perception that the implementation of the programme has helped reduce bond yields. High-yield
bonds and yields on bonds issued by financial corporations also declined immediately after the
announcement of the programme, particularly in peripheral countries (Table 1).
Table 1 • Change in yields following the CSPP announcement | Basis points
Investment-grade bonds
Non-financial corporations
High yield bonds issued by
non-financial corporations
Banks
Euro Area
-14
-8
-58
Germany
-14
-4
-56
France
-13
-8
-72
Italy
-19
-17
-63
Spain
-19
-16
-73
Portugal
-65
-
-64
Sources: Bloomberg, Bank of America Merril Lynch and Banco de Portugal calculations
Notes: Event study: change in yields within a 3-day window following the CSPP announcement on 10 March 2016.
Chart 1 • Euro area bond yields
5
400
350
4
Percentage
250
2
200
150
1
Basis points
300
3
100
0
50
-1
Jan. 12
0
Jan. 13
Asset swap spread (rhs)
Jan. 14
Jan. 15
Jan. 16
Non-financial corporations
Banks
Sources: Bloomberg, Bank of America Merril Lynch and Banco de Portugal calculations.
Note: Effective yields on euro-denominated investment-grade bonds. Asset swap spread: difference between the bond yield
and a low-risk swap rate of the same maturity. Vertical lines: CSPP announcement; start of the program.
25
26
BANCO DE PORTUGAL • Economic Bulletin • May 2017
The programme must also be assessed in terms of volume. Financing of non-financial corporations through debt issuance has been increasing since the first quarter of 2016. In addition,
after a slow down that began in 2013, the average gross issuance of bonds between June and
December recovered in 2016 despite remaining at levels below those of previous years (Chart 2).
This period is particularly interesting as it compares the issuance during the months when the
programme was already being carried out with the same period in previous years. This suggests
that the CSPP may have led firms to be more likely to prefer funding via debt securities.
Although other factors may have influenced the financing costs of non-financial corporations
and were possibly taken into consideration when choosing between debt issuance and bank
loans, the developments described above seem to indicate that the CSPP has had the intended
effect on the financing conditions of the economy. On the one hand, the financing costs of debt
issuance have declined and remain low by historical standards. On the other hand, since the start
of the programme, activity in this market seems to have gained some momentum.
Chart 2 • Gross issuance of debt securities by euro area non-financial corporations
| Millions of euros
100,000
80,000
60,000
40,000
20,000
0
Jan.10
Jan.11
Jan.12
Gross issuance
Sources: Thomson Reuters and Bloomberg.
Jan.13
Jan.14
Jan.15
June to December average
Jan.16
The Portuguese economy in 2016
Box 3.1.2 | Dispersion of interest rates in the euro area money market
In a money market operating perfectly, interest rates on all financial instruments should fully reflect
the central bank’s reference rate – perceived as a risk-free interest rate – plus premia associated,
inter alia, with credit and liquidity risk. Money market interest rates should therefore stand at a
level close to the central bank reference rate and, ceteris paribus, any change in this rate resulting
from monetary policy decisions should mostly pass through to negotiated rates for granting and
obtaining short-term funding.
At present, in the euro area, the deposit facility rate, which stands at -0.4 per cent since March
2016, would be expected to represent the effective lower bound of money market interest rates,
and assets perceived to be safe and liquid would be expected to provide a yield close to this
benchmark. However, since the start of 2015, the downward trend seen in the interest rates of
short-term secured instruments and sovereign bonds with a high credit quality has been more
marked than the declines in the deposit facility rate and the EONIA (Chart 1). This spread widened
from the summer of 2016 onwards, contributing to greater dispersion around the ECB’s monetary policy interest rate, which should provide an anchor for euro area money market interest
rates, under normal circumstances.
High dispersion of interest rates in the money market may indicate market segmentation, which, as a
result of the lower importance given to the central bank’s rate, translates into difficulties in the monetary policy transmission. Against this backdrop, it is crucial to analyse the reasons behind the evidence illustrated in Chart 1.
First, the downward trend in interest rates on short-term instruments with a high credit quality may reflect anticipation by financial market participants of further cuts in the ECB’s policy
rates. Nonetheless, since the summer of 2016, agents’ expectations about further reductions in
Chart 1 • Evolution of representative euro area money market rates since 2015
| Percentage
0.25
1
0.00
0
-0.25
-1
-0.50
-2
-0.75
-3
-1.00
-4
-1.25
Jan. 2015
-5
Apr. 2015
Jul. 2015
Oct. 2015
Jan. 2016
Apr. 2016
Deposit facility rate
EONIA
Germany 3-month sovereign debt rate
Euro area 3-month GC Polling rate
Germany 3-month repo rate (rhs)
Sources: Thomson Reuters and Bloomberg.
Jul. 2016
Oct. 2016
Jan. 2017
27
28
BANCO DE PORTUGAL • Economic Bulletin • May 2017
the deposit facility rate were rather contained and consequently insufficient to explain the movement verified in some short-term interest rates in the euro area market.
Furthermore, the increase in the dispersion of euro area money market interest rates occurred
at the time of the referendum on whether the UK should remain in the European Union. Once
the results were known, behaviours typical of a decline in risk appetite were observed, such as a
search for safe and liquid assets (safe havens), which exerted downward pressure on their interest rates. However, most of these movements were reversed the following weeks, which was not
the case with euro area money market interest rates.
In addition, the misalignment of short-term interest rates and the ECB’s deposit facility rate may
signal the existence of arbitrage opportunities that agents are not exploiting. In effect, to carry out
a direct arbitrage operation on the basis of the current levels of short-term interest rates in the
euro area, at least two conditions must apply: (i) accessing to the deposit facility rate6 and (ii) holding securities with yields lower than the deposit facility rate. This fact leads to two considerations.
On the one hand, the persistence of potential arbitrage opportunities may indicate that monetary
financial institutions (hereinafter called banks) in the euro area, which have access to the ECB’s
deposit facility rate, face constraints in their capability to explore the apparent opportunities for
profit persisting in the euro area money market. For example, the need to hold securities which
may be used as collateral may justify why arbitrage operations are not conducted. Similarly, the
fact that banks face non-negligible costs in adjusting their balance sheets may discourage changes to the composition of their portfolios.
On the other hand, the importance of non-bank institutions in the euro area has increased over
the past few years. Likewise, a significant share of Eurosystem purchases under the expanded
asset purchase programme (APP) comes from the non-bank sector and non-resident investors.
Taking into account that, in the euro area, only banks may make deposits (without risk) with the
ECB and assuming agents who do not belong to the banking sector may be unwilling to keep
large amounts deposited (among other reasons, large amounts are not covered by a deposit
insurance), these investors will tend to look for safe alternatives for their funds. In this context,
an investment possibility would be to purchase instruments with high credit quality, which exerts
downward pressure on their yields and may help explain the dispersion of interest rates in the
euro area money market.
In parallel, in the recent past, there has been an increasing scarcity of short-term securities with
low credit risk and high liquidity, which adds to the downward pressure on the level of their interest rates. This trend may be justified by at least three factors: (i) agents’ concerns arising from the
considerable uncertainty patterns at international level; (ii) lesser availability of securities owing
to the implementation of non-standard monetary policy measures on the part of the Eurosystem,
namely the APP; and (iii) growing demand for liquid and safe securities, driven by regulatory
requirements.
The scarcity of short-term securities is more visible in instruments associated with jurisdictions
which are perceived as having a high credit quality. Specifically, in the past few years, German
sovereign bonds have strengthened their status as a safe haven, are considered special,7 and
therefore strongly sought after by financial market participants.
In addition to these arguments, the dispersion of short-term interest rates in the euro area may
also reflect regulatory issues. In particular, the Basel III requirements may be affecting the smooth
functioning of the money market. Admittedly, in the current regulatory framework, banks must
The Portuguese economy in 2016
comply with a set of requirements which have an impact on liquidity deposited with the central
bank. These guidelines tend to influence money market activity as a whole, thereby influencing short-term interest rates, and ultimately monetary policy transmission. Moreover, a stricter
regulatory framework also leads to higher balance sheet adjustment costs.8 Consequently, the
reversal of positions taken during times of lower risk appetite is harder than in the past and, in
this sense, the correction of certain atypical phenomena, such as those resulting for the UK referendum, tends to be more gradual.
To sum up, greater dispersion of interest rates in the euro area money market may be the result
of a set of conditions interacting with each other, such as (i) greater uncertainty in global financial
markets privileging low-risk assets; (ii) lower interbank activity on the part of euro area banks,
which, in a demanding regulatory framework, do not take advantage of potential arbitrage opportunities and face higher costs in adjusting their portfolios, which, in turn, may lead to temporary
shocks having a more prolonged effect in time; (iii) growing importance of non-bank institutions
which do not have access to the deposit facility rate and seek to allocate their resources to assets
with a high credit quality; and (iv) increasingly lower availability of short-term securities perceived
as safe havens. Although all of these factors may contribute to downward pressure on short-term
interest rates, additional analysis is needed to assess their relative importance.
In this vein, it is relevant to continue to tackle the scarcity of liquid and safe securities, which
might be accomplished, for example, through the Security Lending Scheme by Eurosystem central banks. Similarly to the initiative for reverse repurchase agreement operations, proposed by
the US Federal Reserve, broadening the base for counterparties with access to the deposit facility
rate might, in a context of ample liquidity, strengthen the role of the ECB’s policy interest rate.
Hence, from the perspective of the stance and transmission of monetary policy and aspects
related to financial stability, it is crucial to ensure the smooth functioning of the euro area money
market, by avoiding distortions in price formation.
29
30
BANCO DE PORTUGAL • Economic Bulletin • May 2017
3.2. Portugal
levels seen in interbank reference rates, which
continued to decline throughout 2016.
Monetary and financial
conditions continued
to improve throughout 2016
Monetary and financial conditions in the nonfinancial private sector improved in 2016,
helped by the impact of the non-standard
monetary policy measures adopted by the ECB
(Section 3.1: ‘Monetary and financial conditions
in the euro area‛). In the course of 2016, the
differential between interest rates in Portugal
and the euro area decreased both for firms
and households. The current differentials are
comparable to those seen in the period before
2009 (Chart 3.4).
New loans to households
reinforced the growth trend
seen at the start of 2016,
particularly in consumer credit
The growth rate of new bank loans to households increased in the second half of 2016.
Monthly volumes of new loans for house purchase continued to increase, despite remaining at levels far from those seen in the period before the financial crisis (Chart 3.6). The
share of loans with an initial rate fixation over
1 year continued to grow, accounting for more
than 40 per cent of new loans granted in this
Interest rates on new loans
to households remained
on a downward trend
Interest rates on new loans to households for
house purchase and consumption remained
on the downward path observed since 2012
(Chart 3.5). In addition, December 2016 figures
for these segments correspond to minimum lev-
segment at the end of 2016. This trend occurs
amid very low interest rates – with the accommodative conditions expected to continue for
an extended period – and the existence of longer-term financing instruments for banks.
The recovery in new loans for consumption continued both in personal loans and loans for car
purchase, with loans for the purchase of used cars
gaining importance in this segment (Chart 3.7).
els in the series (which started in 2003). These
On the basis of the results of the Bank Lending
developments are associated with the negative
Survey throughout 2016, the recovery in loans
5.0
Chart 3.4 •
Interest rate
differencials
between Portugal
and the average of
the euro area
| In percentage
4.0
3.0
2.0
1.0
0.0
Dec. 07 Jun. 08 Dec. 08 Jun. 09 Dec. 09 Jun. 10 Dec. 10 Jun. 11 Dec. 11 Jun. 12 Dec. 12 Jun. 13 Dec. 13 Jun. 14 Dec. 14 Jun. 15 Dec. 15 Jun. 16 Dec. 16
-1.0
-2.0
Differencial (Housing)
Differencial (Consumption)
Sources: Consesus Economics, Thomson Reuters and Banco de Portugal.
Differencial (non financial corporations)
31
The Portuguese economy in 2016
to households for house purchase and consumption was in line with a sustained increase
in demand, supported, in particular, by improvements in housing market prospects and the general level of interest rates. According to the survey, although credit standards are relatively stable, competitive pressure between banks and
Household deleveraging
continued, particularly
in the housing segment
At the end of 2016, households’ total indebtedness stood at 75.4 per cent of GDP, compared
with 79.4 per cent one year earlier, continuing the
improvements in housing market prospects con-
deleveraging process that started in mid-2011.
tributed to an easing of credit standards for loans
The annual rate of change in loan stocks for
to households. In addition, several banks reported
house purchase increased slightly, but never-
a positive effect of TLTROs on an increase in loans
theless remained at negative levels. In consumer
to households across segments.
credit, the upward trend continued, with a rate
14.0
Chart 3.5 •
Interest rates
on new loans
granted by
resident banks
to households
| Percentage and
12.0
10.0
8.0
6.0
percentage points
4.0
2.0
0.0
Dec. 07
Dec. 08
Dec. 09
EAPR – housing
Dec. 10
Dec. 11
Spread – housing
Dec. 12
Dec. 13
Dec. 14
EAPR – consumption
Dec. 15
Dec. 16
Spread – consumption
Sources: Thomson Reuters and Banco de Portugal.
Notes: Average interest rates are based on new loans by initial fixation period and weighted by new loan amounts in each period. In
the case of loans for consumption, the 6-month Euribor, the 1-year Euribor and the 5 year swap rate were considered as reference
interest rates for loans with initial fixation period of less than 1 year, 1 to 5 years and more than 5 years, respectively. In the case of
housing, the reference interest rate is the 6-month Euribor.
2,500
Chart 3.6 •
New loans granted
by resident banks
to households
2,000
1,500
| 3-month moving
average, EUR millions
1,000
500
0
Dec. 07
Dec. 08
Dec. 09
Housing (<1 year)
Dec. 10
Dec. 11
Dec. 12
Housing (>=1 year)
Dec. 13
Dec. 14
Consumption
Source: Banco de Portugal.
Note: In the case of housing, new loan amounts are disaggregated by interest rate fixation period.
Dec. 15
Dec. 16
32
BANCO DE PORTUGAL • Economic Bulletin • May 2017
of change of 5.7 per cent at the end of 2016. In
loans for other purposes, the loan stock continued to decline at a rate of around 2.5 per cent for
the year as a whole (Chart 3.8).
Nevertheless, the differential against the EURIBOR stabilised over the course of the year. Together, these factors show the considerable impact of the ECB’s monetary policy on the reduction
in bank interest rates.
Interest rates on new loans
to firms reached new lows
at the end of 2016
The distribution of interest rates
on new loans shows a gradual
increase in risk differentiation
Interest rates on new loans to firms remained on
a downward trend in 2016, reaching a new his-
On the basis of estimates on the probability of
torical low of 2.8 per cent in December (Chart 3.9).
default for Portuguese non-financial corporations
500
Chart 3.7 •
New loans to
households for
consumption by
credit category
| 3-month moving
375
250
average, EUR millions
125
0
Dec. 09
Dec. 10
Dec. 11
Dec. 12
Personal credit
Dec. 13
Dec. 14
Car loans – new
Dec. 15
Dec. 16
Car loans – second hand
Source: Banco de Portugal.
Notes: New loan amounts for consumption granted by financial institutions. The analysis excludes credit cards, current accounts,
and overdraft facilities.
15.0
Chart 3.8 •
Loans granted
by resident banks to
households
| Annual rate
of change, percentage
10.0
5.0
0.0
-5.0
-10.0
-15.0
Dec. 2007
Dec. 2008
Dec. 2009
Total
Dec. 2010
Dec. 2011
Consumption
Dec. 2012
Dec. 2013
Housing
Dec. 2014
Dec. 2015
Dec. 2016
Other purposes
Source: Banco de Portugal.
Notes: Annual rates of change are based on the relation between end-of-month outstanding amounts (adjusted for securitisation operations) and monthly transactions. Monthly transactions correspond to the difference in the end-of-month outstanding amounts adjusted for
reclassifications, write-offs/write-downs, exchange rate and price revaluations, and any other variations that do not correspond to financial
transactions. Whenever relevant, figures are additionally adjusted for sales of credit portfolios.
33
The Portuguese economy in 2016
in 2016, the distribution of interest rates on new
the average. This analysis may suggest the market
loans to lower-risk firms continued to move to
is close to stabilising, with no further reductions
lower levels, particularly through the reduction in
expected in interest rates.
the volume of loans with high interest rates for
The annual rate of change
in total credit to firms
recorded positive levels
in the second half of 2016
the low risk profile of this class (Chart 3.10).9 Panel
b) in Chart 3.10 shows that interest rates on new
loans to risky firms continue to show a considerable dispersion. Nevertheless, a comparison of the
distribution of interest rates in the last quarter of
2016 with those of the same period in 2015 in the
The annual rate of change in total credit to non-
segment for risky firms shows a greater concen-
financial corporations recorded a positive level
tration of new loans with interest rates close to
of 0.5 per cent at the end of 2016 (Chart 3.11).10
8.0
Chart 3.9 •
Interest rates
on new loans
granted by
resident banks
to households
6.0
4.0
2.0
| Percentage
and percentage
points
0.0
-2.0
Dec. 07
Dec. 08
Dec. 09
Dec. 10
Dec. 11
Dec. 12
Average interest rate
Dec. 13
Dec. 14
Dec. 15
Euribor (3-month)
Dec. 16
Differencial
Sources: Consesus Economics, Thomson Reuters and Banco de Portugal.
Notes: Average interest rates are based on new loans by initial fixation period, weighted by new loan amounts in each period.
Chart 3.10 • Density distribution of interest rates on new loans granted by banks to private corporations
by credit risk profile
Panel b) – high risk
0.6
0.6
0.4
0.4
Density
Density
Panel a) – low risk
0.2
0.0
0.2
0.0
0
1
3
5
6
8
9
11
12
14
0
1
3
5
2014 Q4
2015 Q4
6
8
9
11
12
14
Interest rate
Interest rate
2016 Q4
2014 Q4
2015 Q4
2016 Q4
Source: Banco de Portugal.
Note: Interest rates weighted by loan amounts. The sample includes pro-profit corporations. High (low) risk firms lie in the first (last) two deciles of the credit
risk distribution. Credit risk is measured by the Z-score estimated according to Antunes, Gonçalves and Prego, ‘Firm default probabilities revisited‛, Banco de
Portugal Economic Studies, Vol. 2, No 2, April 2016.
34
BANCO DE PORTUGAL • Economic Bulletin • May 2017
These developments were affected by a con-
Survey, which indicate a slight increase in loans
siderable increase in credit from non-resident
to large firms and a narrowing of margins on
entities, which contributed with 3.2 p.p. to the
average loans.
annual rate of change. This increase in credit
Conversely, in the second half of 2016, differ-
was concentrated almost exclusively in debt
ing loan dynamics observed across econom-
securities issuance. Conversely, the annual rate
ic sectors diminished, mainly explained by a
of change in the stock of credit granted by resi-
stabilisation in the rate of change in lending
dent banks remains negative, worsening slight-
to retail and wholesale trade, which fell in line
ly in the last quarter of 2016. In the resident
with manufacturing figures (Chart 3.13). Finally,
financial sector, lending experienced a par-
lending to construction and real estate contin-
ticularly marked contraction, with holdings of
ued to experience a contraction, with an annu-
debt securities by resident entities recording a
al rate of change of almost -7 per cent.
slightly positive rate of change in 2016.
Developments in lending by resident and nonresident institutions are also asymmetric when
breaking down developments in credit by firm
size (Chart 3.12). This asymmetry became more
Bank loans continue to be
channelled to firms
with a better risk profile
marked in the last quarter of 2016. During this
Bank loans by resident banks to firms contin-
period, lending to large firms and holding com-
ue to show mixed developments, according to
panies posted a rate of change of 0.4 per cent,
risk profile (Chart 3.14). Only firms in the lower
as these entities have greater access to non-res-
risk quartile continue to show positive rates of
ident financial institutions, and lending to micro,
change of loans. However, although experienc-
small and medium-sized firms recorded a rate
ing a decline, the total stock of non-performing
of change of -1.8 per cent. This is consistent with
loans accounts for 18 per cent of the total stock of
the results of the January 2017 Bank Lending
loans in 2016 (Chart 3.15).
Chart 3.11 •
Total credit granted
to non-financial
corporations
| Annual rate
of change and
contributions,
percentage
and percentage points
6
6
4
4
2
2
0
0
-2
-2
-4
-4
-6
-6
-8
-8
2010 Q1 2010 Q3 2011 Q1 2011 Q3 2012 Q1 2012 Q3 2013 Q1 2013 Q3 2014 Q1 2014 Q3 2015 Q1 2015 Q3 2016 Q1 2016 Q3
Loans and debt securities held by resident financial institutions
Loans and debt securities held by other resident sectors and trade credit from residents
Non-residents (loans, debt securities and trade credit)
Total credit (rhs)
Loans and debt securities – resident banks (rhs)
Source: Banco de Portugal.
Notes: Annual rates of change are based on the relation between end-of-month outstanding amounts (adjusted for securitisation operations) and monthly transactions. Monthly transactions correspond to the difference in the end-of-month outstanding amounts adjusted for
reclassifications, write-offs/write-downs, exchange rate and price revaluations, and any other variations that do not correspond to financial
transactions. Whenever relevant, figures are additionally adjusted for sales of credit portfolio.
35
The Portuguese economy in 2016
Different credit standards according to risk pro-
loans (Chart 3.16). By contrast, other services con-
file have helped reduce the concentration of loan
tributed with nearly 0.2 p.p.
stocks in certain sectors. For example, the share of
Contrary to 2015 and in particular 2014, loans
(performing) loans to construction and real estate
continued to decline (from a peak of 40.4 per cent
of the total in 2007 to 24.6 per cent in 2016).
granted to large firms with a credit event in the
past again posted a negative rate of change
during 2016 (Chart 3.17). As for loans granted to
In terms of the stock of non-performing loans, con-
micro, small and medium-sized firms with a cred-
struction and real estate contributed with around
it event in previous years, the decline in loans
remained stable during the course of 2016.
0.6 p.p. in 2016 to the decline in non-performing
2.0
Chart 3.12 •
Credit granted
by resident banks
to non-financial
corporations
by firm size
0.0
-2.0
-4.0
| Annual rate
of change, percentage
-6.0
-8.0
2010 Q1 2010 Q3 2011 Q1 2011 Q3 2012 Q1 2012 Q3 2013 Q1 2013 Q3 2014 Q1 2014 Q3 2015 Q1 2015 Q3 2016 Q1 2016 Q3
Micro, small, and medium sized firms
Large firms and holdings
Source: Banco de Portugal.
Notes: Credit includes bank loans and debt securities held by resident banks. Figures are adjusted for securitisation operations, reclassifications, write-offs/write-downs, and exchange rate and price revaluations. Whenever relevant, the figures are additionally adjusted
for sales of credit portfolio.
20.0
15.0
Chart 3.13 •
Loans granted by
resident banks
to non-financial
corporations by
sector of activity
| Annual rate of
10.0
5.0
0.0
-5.0
change, percentage
-10.0
-15.0
Dec. 07 Jun. 08 Dec. 08 Jun. 09 Dec. 09 Jun. 10 Dec. 10 Jun. 11 Dec. 11 Jun. 12 Dec. 12 Jun. 13 Dec. 13 Jun. 14 Dec. 14 Jun. 15 Dec. 15 Jun. 16 Dec. 16
Total
Manufacturing
Construction and real estate activities
Trade
Source: Banco de Portugal.
Notes: Annual rates of change are based on the relation between end-of-month outstanding amounts (adjusted for securitisation
operations) and monthly transactions. Monthly transactions correspond to the difference in the end-of-month outstanding amounts
adjusted for reclassifications, write-offs/write-downs, exchange rate and price revaluations, and any other variations that do not correspond to financial transactions. Whenever relevant, figures are additionally adjusted for sales of credit portfolio.
36
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Average risk for firms
declined further in 2016
line with the recovery in the Portuguese economy. Although the decline in 2016 was smaller in size, all economic sectors made a positive
contribution, in particular wholesale and retail
trade, and electricity, gas and water.
Average risk for firms with performing loans
declined further in 2016, with an aggregate probability of default of 5.9 per cent, slightly below
Corporate indebtedness
decreased slightly, in parallel
with a recovery in profitability
the level observed in 2007 (Chart 3.18). After
growing in 2012 and 2013, average risk posted
a decline, in 2016, for the third year in a row, in
20.0
Chart 3.14 •
Loans granted by
resident financial
institutions
to private
non-financial
corporations by
credit risk quartile
| Year-on-year rate of
change, percentage
15.0
10.0
5.0
0.0
-5.0
-10.0
-15.0
Dec. 07
Dec. 08
Dec. 09
Dec. 10
st
Dec. 11
Dec. 12
nd
1 quartile (less risk)
Dec. 13
Dec. 14
rd
2 quartile
Dec. 15
Dec. 16
th
3 quartile
4 quartile (more risk)
Source: Banco de Portugal.
Notes: Credit risk is measured by the Z-score estimated according to Antunes, Gonçalves and Prego, ‘Firm default probabilities revisited‛,
Banco de Portugal Economic Studies, Vol. 2, No 2, April 2016. The year-on-year rate of change is the annual rate of change of outstanding
amounts in each month.
120
100
120
80
60
40
20
0
Chart 3.15 •
Performing and nonperforming loans
stock
| Thousands millions
euros
100
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
80
60
40
20
0
2006
2007
2008
2009
2010
Trade e Utilities – performing
Other services – performing
Tourism, transports and consulting – performing
Construction and real estate activities – performing
Manufacturing, Mining and Agriculture – performing
2011
2012
2013
2014
2015
2016
Trade e Utilities – non-performing
Other services – non-performing
Tourism, transports and consulting – non-performing
Construction and real estate activities – non-performing
Manufacturing, Mining and Agriculture – non-performing
Source: Banco de Portugal.
Notes: Credit risk is measured by the Z-score estimated according to Antunes, Gonçalves and Prego, ‘Firm default probabilities revisited‛, Banco de Portugal Economic Studies, Vol. 2, No 2, April 2016. The year-on-year rate of change is the annual rate of change of
outstanding amounts in each month. Annual rates of change are based on the relation between end-of-month outstanding amounts
and monthly transactions.
37
The Portuguese economy in 2016
The deleveraging process of non-financial cor-
Statistics on non-financial corporations in the
porations continued in 2016, with the level of
Central Balance Sheet Database, for 2015 as a
indebtedness standing at 105.5 per cent of GDP
whole, confirm that their indebtedness ratio con-
at the end of the year, compared with 109.6 per
tinued to improve (Chart 3.19). With the excep-
cent in the previous year. Although internation-
tion of services and accommodation, this decline
al comparisons of indebtedness ratios face sta-
was broadly-based across economic sectors
tistical challenges, Portuguese firms have delev-
and is partly explained by firms exiting the
eraged more strongly than their euro area coun-
market (Box 3.2.1: ‘Market dynamics and the
terparts.
deleveraging of Portuguese firms‛). In parallel,
11
4
Chart 3.16 •
Non-performing
loans evolution
by industry sector
3
2
|Percentage points
annual contributions
1
0
-1
2007
2008
2009
2010
Tourism, transports and consulting
Other services
2011
2012
2013
Construction and real estate activities
Trade e Utilities
2014
2015
2016
Manufacturing, Mining and Agriculture
Source: Banco de Portugal.
Notes: Credit risk is measured by the Z-score estimated according to Antunes, Gonçalves and Prego, ‘Firm default probabilities revisited‛, Banco de Portugal Economic Studies, Vol. 2, No 2, April 2016. The year-on-year rate of change is the annual rate of change of outstanding amounts
in each month. Annual rates of change are based on the relation between end-of-month outstanding amounts and monthly transactions.
30
Chart 3.17 •
Credit to firms
with a past default
event
20
10
| Percentage points
0
-10
-20
-30
Sep. 07
Sep. 08
Sep. 09
Sep. 10
Sep. 11
Micro and SMEs
Sep. 12
Sep. 13
Sep. 14
Sep. 15
Sep. 16
Large firms
Source: Banco de Portugal.
Note: This analysis includes the evolution of banking loans to firms that registered a credit default event during the period of 2006 to 2016.
38
BANCO DE PORTUGAL • Economic Bulletin • May 2017
a broadly-based increase was observed in the cor-
construction and real estate continue to show
porate profitability ratio. This improvement was
high levels of indebtedness and low levels of prof-
experienced by all economic sectors, with prof-
itability. By contrast, manufacturing, mining and
itability levels in 2015 standing at levels above
quarrying and agriculture posted the lowest lev-
those of 2007 in wholesale trade and transpor-
el of indebtedness and the highest level of prof-
tation, and services and accommodation. Devel-
itability. In addition, the level of indebtedness in
opments in the ratio of indebtedness to profita-
services and accommodation grew consider-
bility continued to differ across economic sectors.
ably, without a significant change occurring in the
Despite improvements compared with the past,
sector’s profitability.
1.5
Chart 3.18 •
Average risk for
firms with only
performing loans
and industry
contributions
| Annual rate
of change and
contributions,
percentage and
percentage points
8%
1.0
6%
0.5
0.0
4%
-0.5
2%
-1.0
-1.5
0%
2006
2007
2008
2009
2010
2011
2012
2013
Tourism, transports and consulting
Manufacturing, mining and agriculture
Trade e utilities
2014
2015
2016
Construction and real estate activities
Other services
Average risk (rhs)
Source: Banco de Portugal.
50
| In percentage
2011
40
2015
Indebtedness
Chart 3.19 •
Firm's indebtedness
and profitability
2007
2011
2015
2011
30
2015
2015
2011
2007
2007
2007
20
0
2
4
6
8
10
12
14
16
18
Profitability
Manufacturing, Mining and Agriculture
Construction and Real estate activities
Construction and Real estate activities
Trade and transports
Source: Banco de Portugal.
Note: Firm's indebtedeness ratio calculated as the sum of total debt over total assets. Firm's profitability ratio calculated as EBITDTA
(which stands for earnings before depreciations and amortizations, interest expenses, and income tax) divided by the sum of equity
and total debt.
The Portuguese economy in 2016
Box 3.2.1 | Market dynamics and the deleveraging of Portuguese firms
The economic and financial crisis of 2008 highlighted the negative effects of the excessive indebtedness of Portuguese firms on financial stability and economic growth. The Economic and
Financial Assistance Programme identified the private sector deleveraging, including the financial
system, as an important factor for the adjustment of the Portuguese economy. In this context, the
indebtedness of Portuguese non-financial corporations as a proportion of total assets, is estimated to have declined by approximately 4 p.p. to 49 per cent in the period between 2012 and
2015, accounting for 118 per cent of GDP by the end of this period.12
Simultaneously to the deleveraging process observed in the past few years in the private sector,
the number of firms ceasing their activity increased.13 Therefore, it is important to understand
to what extent market dynamics contributed to explain the deleveraging process of Portuguese
firms.
The firm balance sheet information reported in the survey Simplified Corporate Information (Informação Empresarial Simplificada – IES) was considered in the analysis.14 This dataset allows the identification of firms exiting and entering the market in the period between 2010 and 2015. A firm exit in
year t is defined by its absence from the IES in that year, provided that this absence does not constitute
a reporting gap.15 The entry of firms in the market is identified by the first year reporting to the IES. Chart
1 shows the number of new firms and firms that have ceased their activity in the period between 2010
and 2015.
In the analysis, debt includes the funding obtained (loans and debt securities) by the firm and
reported in the IES. According to this information, the total debt of non-financial corporations
declined by around 3.4 percent in 2014 and 6.0 percent in 2015 (Chart 2).
Chart 3 depicts the contributions to the change in the debt of non-financial corporations, considering the set of new firms, the set of firms that have ceased their activity, and the set of firms that
remained active in the period between 2011 and 2015. The contribution of the set of new firms to
the aggregate change in debt is small, in contrast to the substantial contribution of firms that have
Chart 1 • Entry and exit of firms
20,000
18,000
16,000
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
2011
2012
Entrants (1ˢᵗ report in t)
2013
2014
2015
Exits in t (do not report in t)
Source: Banco de Portugal.
Note: The sample includes private non-financial corporations of all economic activity sectors except insurance and financial
activities.
39
BANCO DE PORTUGAL • Economic Bulletin • May 2017
ceased their activity. The negative contribution of firms exiting the market to changes in debt was
particularly evident in 2015.16 In turn, the debt of firms that remained active increased considerably
in 2011 and less significantly in 2012, and decreased in the following years.
Nevertheless, the exit of firms from the market does not necessarily mean that their debt amounts
are removed from the debt outstanding amounts recorded in the balance sheets of financial institutions, which are the main source of information when analysing the growth of credit granted by the
financial system. In fact, the existence of assets and personal guarantees allows the repayment of
part of the debt even after the firm ceases its activity. This box does not analyse this difference between information reported to the IES and information in the balance sheets of financial institutions.
Chart 2 • Total debt of private non-financial corporations | Annual rate of change
220,000
7
200,000
5
180,000
160,000
3
140,000
1
120,000
100,000
-1
Percentage
EUR millions
40
80,000
-3
60,000
40,000
-5
20,000
-7
0
2010
2011
Total debt
2012
2013
2014
2015
Annual rate of change (right-hand scale)
Source: Banco de Portugal.
Note: Total debt includes loans and debt securities. The sample includes private non-financial corporations of all economic activity sectors except insurance and financial activities.
Chart 3 • Contributions to debt change by firms' dynamics | EUR millions
10.000
5.000
0
-5.000
-10.000
-15.000
-20.000
2011
Entrants (1ˢᵗ report in t)
2012
2013
Stayers (report in t-1 and t)
2014
2015
Exits (do not report in t)
Source: Banco de Portugal.
Note: Total debt includes loans and debt securities. The sample includes private non-financial corporations of all economic
activity sectors except insurance and financial activities.
The Portuguese economy in 2016
Chart 4 shows the contributions to changes in debt by firm size. The debt reduction resulting from the
exit of firms is concentrated in smaller-sized firms. Large firms that remained in the market in 2014 and
2015 also registered a reduction in debt.
The decomposition of the contributions to changes in debt by sector of economic activity shows that
the exit of firms in the construction sector had a substantial impact on debt reduction, especially in the
last year (Chart 5). Firms in the trade sector that remained in the market in 2012 and 2013 significantly
reduced their debt. The debt reduction resulting from the exit of firms in this sector of economic activity is particularly relevant in 2015. In the manufacturing sector, it is particularly relevant the debt reduction of firms that remained in the market in 2014 and firms that ceased their activity in 2015.
Chart 4 • Contributions to debt change by firms' dynamics
| By firm size, EUR millions
SME
Large corporations
6.000
4.000
2.000
0
-2.000
-4.000
-6.000
-8.000
-10.000
-12.000
5.000
4.000
3.000
2.000
1.000
0
-1.000
-2.000
-3.000
-4.000
-5.000
2011
Entrants (1ˢᵗ report in t)
2012
2013
Stayers (report in t-1 and t)
2014
2015
2011
Exits (do not report in t)
2012
Entrants (1ˢᵗ report in t)
2013
2014
Stayers (report in t-1 and t)
2015
Exits (do not report in t)
Source: Banco de Portugal.
Note: Total debt includes loans and debt securities. The sample includes private non-financial corporations of all economic activity
sectors except insurance and financial activities.
Chart 5 • Contributions to debt change by firms' dynamics | By sector of economic activity,
EUR millions
Construction
Trade
2.000
2.000
1.000
0
-1.000
-2.000
-3.000
-4.000
-5.000
-6.000
1.000
0
-1.000
-2.000
-3.000
-4.000
-5.000
-6.000
2011
Entrants (1ˢᵗ report in t)
2012
2013
Stayers (report in t-1 and t)
2014
2015
Exits (do not report in t)
2011
Entrants (1ˢᵗ report in t)
2012
2013
Stayers (report in t-1 and t)
2014
2015
Exits (do not report in t)
Manufacturing
2.000
1.000
0
-1.000
-2.000
-3.000
-4.000
-5.000
-6.000
2011
Entrants (1ˢᵗ report in t)
2012
2013
Stayers (report in t-1 and t)
2014
2015
Exits (do not report in t)
Source: Banco de Portugal.
Note: Total debt includes loans and debt securities. The sample includes private non-financial corporations of all economic activity
sectors except insurance and financial activities.
41
Neste contexto de desalavancagem das empresas, importa aprofundar de que forma
A análise
empírica
decompõe
a evolução
do end
A análise
empírica
abaixoabaixo
decompõe
a evolução
do endividam
42
a dinâmica de mercado contribuiu para o endividamento e rendibilidade das empresas.
das empresas
portuguesas
no período
compreendido
das empresas
portuguesas
no período
compreendido
entre e
BANCO DE PORTUGAL • Economic Bulletin • May 2017
A análise empírica abaixo decompõe a evolução do endividamento e da rendibilidade
a dinâmica
de mercado
e
idade.
A análise
catego
com
acom
dinâmica
de
mercado
edas
idade.
A análise
categoriza
das empresas portuguesas
no
período
compreendido
entre
2010
e 2015,
de
acordo
Neste
contexto
de
desalavancagem
das
empresas,
importa
aprofundar
de
Neste
contexto
de
desalavancagem
das
empresas,
importa
aprofundar
de
qu
Neste
contexto
de
desalavancagem
empresas,
importa
aprofundar
de
qq
In this context of corporate deleveraging, it is important to know to what extent the market dynamics
com atodinâmica
mercado
e idade.
A análise
categoriza
as
empresas
de
contributed
corporatede
indebtedness
and
profitability.
The
empirical
analysis
presented
below
adinâmica
dinâmica
de
mercado
contribuiu
para
oendividamento
endividamento
eque
rendibilidade
das
a adinâmica
de
mercado
contribuiu
para
oo
endividamento
eacordo
das
em
mercado
contribuiu
para
erendibilidade
rendibilidade
das
eeo
ade
idade:
empresas
novas
são
asdetêm ou
dois
com
acom
idade:
empresas
novas
são
as
que
têm
dois
me
and 2015, according
to market
dynamics and
firm age. of
InPortuguese
the analysis,
classified
to
composes changes
in indebtedness
and profitability
firmsfirms
in theare
period
betweenaccording
2010
com a idade: empresas
novas
são
as que
têmdecompõe
dois
ou menos
anosdo
no
mercado,
A
análise
empírica
abaixo
decompõe
a
evolução
do
endividamento
eda
da
ren
A
análise
empírica
abaixo
a
evolução
endividamento
e eda
rendi
A
análise
empírica
abaixo
decompõe
a
evolução
do
endividamento
rend
their age: young
firmsaccording
are those
have
been
in the
market
for
years
ortêm
less,
vintage
firms
are
and 2015,
to that
market
dynamics
and
firm
age.
Insão
the two
analysis,
firms
are
classified
empresas
vintage
são
as
que
têm according
mais
dethose
dois anos
e
empresas
vintage
as
que
mais
de
dois
anos
e meno
to their
age:
young
firms
are
those
that
have
been
inde
thedois
market
for
years
or
less,
vintage
firms
with more than
two
years
and
less
than
five,
and
firms
are
those
aged
more
than
five
years.
empresas
vintage
são
as
que
têmestablished
mais
anos
etwo
menos
de
cinco,
e empresas
das
empresas
portuguesas
no
período
compreendido
entre
2010
2015,
das
empresas
portuguesas
no
período
compreendido
entre
2010
e ee
2015,
de
das
empresas
portuguesas
no
período
compreendido
entre
2010
2015,
dd
are thoseequation
with morewas
thanestimated:
two years and less than five, and established firms are those aged more
Then, the following
estabelecidas
são
astêm
quemais
têm de
mais
de
cinco
De s
estabelecidas
são
as
que
cinco
anos.anos.
De seguida
sãofollowing
as que
têm
mais
deestimated:
cinco
anos.
De
a seguinte
com
dinâmica
demercado
mercado
eidade.
idade.AAestimou-se
Aanálise
análisecategoriza
categoriza
asempresas
empresasde
com
a aadinâmica
de
e eseguida,
asas
com
dinâmica
de
mercado
idade.
análise
categoriza
empresas
dd
than estabelecidas
five years. Then, the
equation
was
equação:
com
idade:empresas
empresasnovas
novassão
sãoasas
asque
quetêm
têmdois
doisou
oumenos
menosanos
anosno
nom
a aaidade:
com
idade:
empresas
novas
são
que
têm
dois
ou
menos
anos
no
equação:
�� � �� � �� � �� �equação:
��com
����
��� �
�� � �� ��������� � �� ������� � �� ��������� � ��� empresas
vintage
são
as
que
têm
mais
de
dois
anos
menos
de
cinco,
empresas
vintage
são
as
que
têm
mais
de
dois
anos
e ee
menos
de
cinco,
e ee
em
empresas
vintage
são
as
que
têm
mais
de
dois
anos
menos
de
cinco,
e
where ‫ݕ‬௜௧two
assumes
indicators:
of debt
total
to assets)
profitability
ൌ
ߚindicators:
൅ ߛ௦ ൅ ߚindebtedness
‫ܽ݀ܽݎݐ݊ܧ‬
ߚଷassets)
ܰ‫ܽݒ݋‬
ߚand
‫ݑ‬௜௧ of
଴ ൅ ߙtwo
௜ ൅ߜ
௧ indebtedness
ଵ ܵܽÀ݀ܽ(ratio
௜௧ ൅ ߚof
ଶ(ratio
௜௧ ൅debt
௜௧ ൅
ସ ܸ݅݊‫݁݃ܽݐ‬
௜௧ ൅ (ratio
where ��� assumes
total
to
and
profitability
‫ ݕ‬൅
ൌߙߚ ൅൅ߜߙ ൅൅ߛߜ ൅൅ߚߛܵܽÀ݀ܽ
൅ ߚ ܵܽÀ݀ܽ
൅ ߚ ‫ܽ݀ܽݎݐ݊ܧ‬
൅ߚ
‫ ݕ‬ൌߚ
൅ ߚ ‫ܽ݀ܽݎݐ݊ܧ‬
൅ ߚ ܰ‫ܽݒ݋‬
௜௧. The
௜ Exit
௧mais
௦ de
ଵ ௜௧anos.
variables
Entry
are
binary
and ଶ௜௧
equal
toଶ
(ratio of EBIT to assets) of firm
௜௧in year଴
௜଴ as
௧que
௦and
ଵde
௜௧
ଷ௜௧
estabelecidas
são
as
que
têm
mais
de
cinco
anos.
De
seguida,
estimou-se
estabelecidas
que
têm
cinco
De
seguida,
estimou-se
a asaଷ
estabelecidas
são
as
têm
mais
cinco
anos.
De
seguida,
estimou-se
EBIT17 to assets)
of firm i in year t. The
variables
Exitsão
and
Entrant
are
binary
variables
equal
to
one for
one for
firms
exiting
and
entering
the
market
respectively.The
variables
Exit
and
Entrant
are
binary
ementering
que ‫ݕ‬௜௧ assume
dois respectively.
indicadores: endividamento
(rácio and
da dívida
total
ativo)
e
firms exiting and
the market,
The variables Young
Vintage
arenoalso
binary
equação:
equação:
variables equal to one for firmsequação:
exiting
and entering the market, respectively. The variables Young
variables equal rendibilidade
to one for firms less
two
old
firms with
two
to
fivevariáveis
years,
respectively.
17 ‫ݕ‬years
em
‫ݕ‬for
assume
indicadores:
endividamento
em
que
assume
dois
indicadores:
endividamento
(rácio(rá
d
௜௧to
dothan
EBIT
noque
ativo)
daand
empresa
idois
notwo
ano
t. As
Saída e
and Vintage are also (rácio
binary variables
equal
one ௜௧
firms
less than
years
old and firms with
Finally, the terms �� , �� , and �� denote a set of firm, time, and sector of economic activity fixed effects,
two to five years, respectively. Finally,
the
,൅
a set
time, and
൅
൅
ܵܽÀ݀ܽ
‫ܽ݀ܽݎݐ݊ܧ‬
ܰ‫ܽݒ݋‬
൅
ܸ݅݊‫݁݃ܽݐ‬
ൌ
ߚ଴terms
ߙߙ
ߜ௧ߜߜand
ߛ௦ߛߛ௦൅
ߚଵߚߚܵܽÀ݀ܽ
ߚfirm,
ߚଷߚߚܰ‫ܽݒ݋‬
ߚସߚߚ
ܸ݅݊‫݁݃ܽݐ‬
‫ݕ‬௜௧
ൌൌ
ߚߚ଴൅
൅൅
൅൅
ߚߚ‫ܽ݀ܽݎݐ݊ܧ‬
‫ܽ݀ܽݎݐ݊ܧ‬
൅൅
‫ݕݕ‬௜௧௜௧
଴൅
௧൅
௦denote
ଵܵܽÀ݀ܽ
௜௧of
௜௧sector
௜௧൅
௜ߙ൅
௜௧ ௜௧൅
ଶ
௜௧no
௜௧ ௜௧൅
௜௧ ௜
௜ ௜൅
௧൅
ଵ
ଶଶentram
௜௧൅
ଷଷܰ‫ܽݒ݋‬
ସସܸ݅݊‫݁݃ܽݐ‬
Entrada
binárias
e effects
iguais
aaccount
um
para
empresas
que
mercado,
17
respectively. Firm
and são
sector
fixed
for
their
specific
(observed
rendibilidade
(rácio
dosaem
EBIT
noand
ativo)
da empresa
in
rendibilidade
(rácio
do EBIT
noe17
ativo)
daunobserved)
empresa
i no ano
17
of economic activity fixed effects, respectively. Firm and sector fixed effects account for their spe-
characteristics,
provided they remain
constant
over
time.
Insão
turn,
time effects
attempt
to capture
the
cific (observed and unobserved)
characteristics,
provided
they
remain constant
over
time. In(rácio
turn,
respetivamente.
As variáveis
Nova
eassume
Vintage
também
binárias
e assumem
o valor
em
que
assume
dois
indicadores:
endividamento
(rácio
da
dívida
total
n
em
que
‫ݕ‬௜௧
dois
indicadores:
endividamento
da
dívida
total
no
em
que
‫ݕݕ‬௜௧௜௧assume
dois
indicadores:
endividamento
(rácio
da
dívida
total
no
cyclical position
of
the
economy.
In
this
case,
the
comparison
group
are
the
established
firms
that
Entrada
são
binárias
e
iguais
a
um
para
empresas
que
Entrada
são
binárias
e
iguais
a
um
para
empresas
que
saem
time effects attempt to capture the cyclical position of the economy. In this case, the comparison
17
1717
de umIn no
caso
de rendibilidade
empresas
novas
eobservations
com
idades
entre
dois
e percentile
cinco
anos,
rendibilidade
(rácio
do
EBIT
no
ativo)
da
empresa
no
ano
As
variávei
ativo)
da
empresa
i no
ano
t. t.t.
As
variáveis
(rácio
do
EBIT
(rácio
do
ativo)
da
empresa
i i no
ano
As
variáveis
remained ingroup
activity.
order
to avoid
extreme
values,
above
the
99th
and
are the established
firms rendibilidade
that
remained
in
activity.
InEBIT
order no
tono
avoid
extreme
values,
obser-
st As variáveis
below the 1st
percentile
notrespetivamente.
considered.
The
was
performed
using
the
fixed-effects
respetivamente.
As
variáveis
Nova
eanalysis
Vintage
são
també
e Vintage
são
bin
vations
above thewere
99th percentile
and
below
the 1analysis
were
notNova
considered.
The
respetivamente.
Finalmente,
os
termos
ߙ௜percentile
, ߜe௧ eee
ߛiguais
um
conjunto
de
efeitos
datambém
Entrada
são
binárias
um
para
empresas
que
saem
entram
no
Entrada
são
binárias
iguais
a aa
um
para
empresas
que
saem
e ee
entram
no
m
Entrada
são
binárias
iguais
um
para
empresas
que
saem
entram
no
௦ denotam
estimator for
panel
data. using the fixed-effects estimator for panel data.
was
performed
empresa, tempo e setor,
respetivamente.
Os
efeitos
da
empresa
e do
setor
respetivamente.
As
variáveis
Nova
eVintage
Vintage
são
também
binárias
eassume
assume
respetivamente.
As
variáveis
Nova
e eVintage
são
também
binárias
e eassumem
respetivamente.
As
variáveis
Nova
são
também
binárias
deinno
um
no
caso
de
empresas
novas
e
com
idade
de
um
caso
de
empresas
novas
epermitem
com
idades
ent
of the estimation
are
reported
in Table
According
to
estimates,
firms
exiting
the
The resultsThe
of results
the estimation
are reported
Table
1.1.According
tothe
the
estimates,
firms
exiting
the
are the
most
and
the least
profitable.
In turn,
firmsthat
that have
have
recently
entered
market are market
the controlar
most
indebted
and
least
profitable.
In
firms
recently
entered
thedois
paraindebted
as the
respetivas
características
específicas
(observadas
e entre
não
de
umno
no
casode
deturn,
empresas
novas
comidades
idades
entre
doise eecinco
cin
de
um
caso
empresas
novas
e ee
com
de
um
no
caso
de
empresas
novas
com
idades
entre
dois
cin
respetivamente.
Finalmente,
termos
e ߛ௦ denot
respetivamente.
Finalmente,
termos
ߙ௜ , firms
ߜexiting
e௜ ,ߛ௦ߜthe
um
are less
and less
profitable
older incumbent
firms.
Inos
particular,
௧denotam
௧ߙ
market are the
lessmarket
indebted
andindebted
less profitable
than
olderthan
incumbent
firms. os
In
particular,
firms
observadas),
desde
que
sejam
constantes
ao
longo
do
tempo.
Por
sua
vez,
os
efeitos
respetivamente.
Finalmente,
os
termos
ߙ
,௧ߜߜprofitable
e௦older
denotam
um
conjunto
de
respetivamente.
Finalmente,
osos
termos
,௜௜ߜ
um
conjunto
de
efe
respetivamente.
Finalmente,
termos
,than
ߛߛ௦denotam
um
conjunto
de
the market33.9
are approximately
33.9 p.p.and
more
indebted
and
4.93
p.p.ߙߙ
than
௦ denotam
௜less
௧e௧ eߛ
market are exiting
approximately
p.p. more
indebted
4.93
p.p. less
profitable
incumbent
older incumbent firms, on average.
This
selection
dynamics
suggests
that the
in the
empresa,
tempo
e da
setor,
respetivamente.
Os efeitos
da
empresa,
tempo
ethe
setor,
respetivamente.
Os
efeitos
da empr
firms, on average.
This selection
suggests
that
improvement
inimprovement
the
aggregate
financial
temporais
têm emdynamics
consideração
atempo
posição
cíclica
economia.
Note-se
que
neste
empresa,
esetor,
setor,
respetivamente.
Os
efeitos
da
empresa
edo
do
setor
empresa,
tempo
e esetor,
respetivamente.
Os
efeitos
da
empresa
e edo
setor
pe
empresa,
tempo
respetivamente.
Os
efeitos
da
empresa
setor
aggregate financial position of Portuguese firms is partly due to the fact that the most indebted
position of Portuguese firms is partly due to the fact that the most indebted and less profitable firms
and less
have
exited
the
market.
casoprofitable
o grupofirms
de comparação
são
as
empresas
estabelecidas
que
permaneceram
no específicas
controlar
para
as respetivas
características
espe
controlar
para
as
respetivas
características
controlar
para
as
respetivas
características
específicas
(observadas
controlar
para
asas
respetivas
características
específicas
(observadas
controlar
para
respetivas
características
específicas
(observadas
have exited the market.
mercado. Para evitarobservadas),
a presençadesde
de valores
extremos,
nãoao
se
consideraram
observadas),
desde
que
sejam
constantes
ao
longo
do
tempo.
Por
sua
vez,
que
sejam
constantes
longo
do
tempo.
Por
sua
vez,
oso
observadas),
desde
que
sejam
constantes
ao
longo
do
tempo.
Por
sua
vez,
Table 1 • Fixed effects estimates
observadas),
que sejam
constantes
ao longo
do t
observadas),
desdedesde
que sejam
constantes
ao longo
do tempo
12. See Box “Developments in corporate indebtedness in Portugal and the euro area”, Economic Bulletin, May 2016, Banco de Portugal. The analysis considers
temporais
têmem
em
consideração
posição
cíclica
daeconomia.
economia.Note-se
Note-sequ
temporais
têm
consideração
cíclica
temporais
têm
em
consideração
aaposição
posição
cíclica
da
economia.
Note-se
qq
data from the national financial accounts and the measure of
indebtedness
considered
is the
ratio of debt (loansaand
debt
securities)
to totalda
assets.
Indebtedness
Profitability
temporais
têmconsideração
emde consideração
a posição
da e
temporais
a posição
cíclicacíclica
da econo
13. Félix, S. (2017) “Firm creation and survival in Portugal”,
Banco de
Portugaltêm
Economicem
Studies,
Banco
Portugal.
caso
ogrupo
grupo
de
comparação
são
as
empresas
que
permane
caso
o ogrupo
de
comparação
são
asas
empresas
estabelecidas
permanece
caso
de
comparação
são
empresas
estabelecidas
que
permane
14. The survey IESEntrant
includes economic, financial, and accounting
information
of-0.0332***
virtually
all
the Portuguese non-financial
corporations
and
isestabelecidas
available for the que
-0.0413***
period between 2005 and
2015. Nevertheless, changes in the accounting
standards system in 2010 have led to a loss of data comparability.
caso
oPara
grupo
de
são
as
empresas
estabel
caso
o
grupo
deevitar
comparação
são
as
empresas
estabelecidas
(0.0031)
(0.0018)
mercado.
Para
evitar
presença
devalores
valores
extremos,
nãose
seconsid
con
Para
a aacomparação
presença
de
extremos,
não
mercado.
de
valores
extremos,
não
se
cons
17 EBIT é o acrónimo para earnings before interest and taxes. is absent
15. A reporting gap is observed
when the firm reports to themercado.
IES
in year t-1 and
year t+1,evitar
but
in presença
year t.
Exit
0.3386***
-0.0493***
16. Some of the absences that are considered exits from the market in 2015 may constitute reporting gaps. This can only be verified when the IES for 2016
65 mercado.
Para
evitar
a (0.0024)
presença
de valores
extre
Paragaps
evitar
aunderpresença
de valores
extremos,
becomes available. However, only about 0.2 per cent ofmercado.
the observations
are reporting
in the period
analysis.
(0.0047)
17. EBIT stands for Earnings Before Interest and Taxes.
Young
-0.0603***
-0.0543***
(0.0037)
(0.0026)
Chart3.2.1Entryandexitoffirms
17 1717 EBIT é o acrónimo para earnings before interest and taxes. EBIT é o acrónimo para earnings before interest and taxes. EBIT é o acrónimo para earnings before interest and taxes. -0.0223***
-0.0067***
Entrants(1stVintage
yearint)
Exitsint(absentint)
(0.0025)
(0.0016)
Source:BancodePortugal.
Observations
2,001,156
1,063,059
Note:Thesampleincludesprivatenon‐financialcorporationsofalleconomicactivitysectorsexcept
17 EBIT é o acrónimo para earnings before interest and taxes. 17 EBIT é o acrónimo para earnings before interest and taxes. insuranceandfinancialactivities.
Note: *, **, *** denote statistical significance at 1, 5, and 10 percent, respectively. Both specifications include firm, sector of economic activity, and time
fixed effects. Robust standard errors clustered at firm level in parentheses.
Chart3.2.2Totaldebtofprivatenon‐financialcorporations|Year‐on‐yearrateofchange
Totaldebt
Year‐on‐yearrateofchange(right‐handscale)
Source:BancodePortugal.
Note:Totaldebtincludesloansanddebtsecurities.Thesampleincludesprivatenon‐financialcorporations
ofalleconomicactivitysectorsexceptinsuranceandfinancialactivities.
The Portuguese economy in 2016
4.Fiscal policy and situation
The general government deficit
declined significantly, against
a background of a stabilisation
in the structural balance and
a reduction in the primary
structural balance
In 2016, the general government deficit stood at
2.0 per cent of GDP, which is a historically low
level (Table 4.1). This figure stands below the
official target included in the State Budget for
2016 (2.2 per cent of GDP), the revised estimate
included in the State Budget for 2017 (2.4 per
cent of GDP), and the limit laid down by the
European Union Council in August 2016 (2.5 per
cent of GDP).
the impact of temporary measures (Chart 4.1).
In 2015 the general government balance was
negatively affected by the resolution of BANIF
(-1.4 per cent of GDP). In 2016 the effect of
temporary measures corresponds to revenue
obtained from full payments made within the
scope of a special programme for the settlement of overdue tax and social contributions
(Programa Especial de Redução do Endividamento ao Estado – PERES) and the refund by
the European Financial Stability Facility (EFSF)
of amounts previously paid by Portugal under
the Economic and Financial Assistance Programme. 18 Overall, these effects contributed
to increase the fiscal balance in 2016 by 0.4 per
cent of GDP. Therefore, the decline in the fiscal deficit adjusted for temporary measures
The deficit improved by 2.3 p.p. vis-à-vis the pre-
stood at 0.5 p.p. of GDP, 0.3 p.p. of which refer
vious year, chiefly as a result of the change in
to a drop in interest expenditure.
Table 4.1 • Main fiscal indicators | In percentage of GDP
2010
Overall balance (1)
Temporary measures and special
factors (a) (2)
Overall balance excluding temporary
measures and special factors (3 = 1-2)
Cyclical component (4)
Structural balance(b) (5 = 3 - 4)
Interest expenditure (6)
2011
2012
2013
2014
2015
2016
Change
2016-2015 (a)
-11.2
-7.4
-5.7
-4.8
-7.2
-4.4
-2.0
2.3
-2.8
-0.2
0.0
0.3
-3.6
-1.4
0.4
1.8
-8.4
-7.2
-5.6
-5.2
-3.6
-3.0
-2.4
0.5
1.5
0.6
-1.7
-2.3
-1.4
-0.6
-0.1
0.5
-9.9
-7.9
-3.9
-2.9
-2.2
-2.4
-2.3
0.0
2.9
4.3
4.9
4.9
4.9
4.6
4.2
-0.3
Structural primary balance (7=5+6)
-7.0
-3.5
1.0
2.0
2.7
2.2
1.9
-0.3
Structural revenue (percentage
of trend GDP) (b)
39.9
42.0
43.1
44.9
44.9
44.1
42.5
-1.6
Structural primary expenditure
(percentage of trend GDP) (b)
47.1
45.6
42.2
43.0
42.3
42.0
40.6
-1.3
Public debt
96.2
111.4
126.2
129.0
130.6
129.0
130.4
1.4
Change in public debt (in p.p.)
12.6
15.2
14.8
2.8
1.6
-1.6
1.4
–
(-) Primary balance
8.2
3.1
0.8
0.0
2.3
-0.2
-2.2
–
Differential between the effects
of interest and of GDP growth
0.8
6.4
10.0
3.5
2.8
-0.1
0.5
–
Deficit-debt adjustments
3.5
5.8
4.0
-0.6
-3.5
-1.3
3.1
–
Sources: INE and Banco de Portugal.
Notes: (a) Special factors are operations with an one-off effect on the general government deficit which cannot be treated as temporary measures
according to the definition adopted in the Eurosystem. (b) Structural figures are adjusted for the impacts of the cycle, temporary measures and
special factors. The cyclical component and temporary measures are computed by Banco de Portugal according to the methodologies adopted
in the Eurosystem.
43
44
BANCO DE PORTUGAL • Economic Bulletin • May 2017
According to the cyclical adjustment methodol-
terms, both the tax and contributory burden and
ogy adopted in the Eurosystem, economic activ-
the non-tax revenue declined as a ratio to trend
ity gave a positive contribution to fiscal develop-
GDP (by -0.9 p.p. and -0.7 p.p. respectively).
ments, accounting for 0.5 p.p. of the deficit
improvement. Therefore, the structural balance
is estimated to have stabilised in 2016.19 As
regards the structural primary balance, which
excludes the impact of interest expenditure, it is
estimated to have declined for the second year
in a row (by 0.3 p.p. of trend GDP).
The drop in structural revenue
largely stemmed from tax
revenue developments
In actual terms, tax and contributory revenue rose by 2.2 per cent in 2016, reflecting an
The decline in the structural primary balance
increase in revenue from taxes on production
in 2016 reflects a drop in structural revenue by
and imports and social contributions. Econom-
1.6 p.p. of trend GDP, which more than offset the
ic activity is estimated to have given a positive
1.3 p.p. contraction recorded by structural primary
contribution to these developments. They also
expenditure (Chart 4.2). In particular, in structural
reflect revenue collected within the scope of
6.0
Chart 4.1 •
Breakdown of
the change in the
general government
balance
4.0
| In percentage points
-2.0
of trend GDP
2.0
0.0
-4.0
-6.0
2011
2012
2013
Structural primary expenditure ( - )
Temporary measures
Structural balance (p.p. of GDP)
2014
2015
Interest expenditure ( - )
Structural revenue
2016
Cyclical component
Overall balance (p.p. of GDP)
Sources: INE and Banco de Portugal.
50
Chart 4.2 •
Structural revenue
and primary
expenditure
developments
| In percentage of
45
40
trend GDP
35
30
1995
1997
1999
2001
2003
Structural revenue
Sources: INE and Banco de Portugal.
2005
2007
2009
2011
2013
Structural primary expenditure
2015
The Portuguese economy in 2016
The decrease in general
government expenditure chiefly
reflected a decline in investment
PERES, part of which, as previously mentioned,
has a temporary nature. The structural tax and
contributory burden, which is adjusted for the
impact of the cycle and temporary measures,
stabilised in 2016, implying a decline by 0.9 p.p.
of trend GDP (Chart 4.3). These developments
are chiefly explained by a drop in revenue from
the taxes on income and wealth (-0.8 p.p.),
reflecting the impact of policy measures implemented not only in 2016, but also in previous
years (Box 4.1: ‘Structural developments of tax
revenue in 2016’).
45
The structural decline in primary expenditure was
chiefly driven by a decrease in capital expenditure. Adjusting for the effect of temporary measures with an impact on 2015, the decline in capital
expenditure was 0.9 p.p. of trend GDP.20 These
developments were largely due to the decline
in public investment (Chart 4.4). In 2016, in particular, public investment declined by 30.9 per
In 2016 non-tax revenue declined by 6.3 per
cent in nominal terms, which partly mirrored a
cent, resulting from a drop in other current rev-
base effect resulting from the recording of real
enue (including sales of goods and services) and
estate acquisitions by Oitante in 2015, within the
capital revenue. The 3.7 per cent decline in oth-
scope of the resolution of BANIF, and the record-
er current revenue reflects decreases in most
ing of military material deliveries in 2016.21 The
of its items, particularly as regards interest and
change in public investment adjusted for these
dividends. Capital revenue, in turn, declined by
factors stands at -24.8 per cent. This decline may
27.9 per cent in actual terms. This occurred in
have partly reflected a decrease in funds received
spite of the refund of the margins previous-
from the European Union and it was particularly
ly paid to the EFSF, which, as previously men-
relevant in the Autonomous Services and Funds
tioned, has a temporary impact. Therefore, the
subsector. Turning to other capital expenditure
change in capital revenue adjusted for tempo-
adjusted for the impact of temporary measures,
rary measures would be more negative in 2016,
there was also a decline as a ratio to trend GDP
at -51.0 per cent, as a result of the significant
(0.3 p.p.), reflecting several of base effects that
fall in capital transfers from the European Union
contributed to a one-off growth of this expendi-
matched by expenditure in the same year.
ture item in 2015.22
3.0
Chart 4.3 •
Breakdown
of the change
in structural
revenue
2.0
1.0
| In percentage
points of trend GDP
0.0
-1.0
-2.0
2012
2013
Taxes on households' income
VAT
Social contributions
2014
2015
2016
Taxes on firms' income
Other taxes on production and imports
Other revenue (a)
Sources: INE and Banco de Portugal.
Note: (a) Other revenue encompasses ‘other current‛ revenue, including sales, and ‘capital revenue‛.
46
BANCO DE PORTUGAL • Economic Bulletin • May 2017
In 2016, expenditure on social benefits in cash
pension expenditure increased by 1.5 per cent.
increased by 1.3 per cent, both in actual and
Whereas pension expenditure paid by CGA (pub-
structural terms, and declined by 0.2 p.p. as a
lic employees’ pension system) virtually stabi-
ratio to trend GDP. Expenditure on unemploy-
lised due to a lower growth of new retirements,
ment benefits decreased by 14.3 per cent in
in the general pension system expenditure grew
actual terms (14.1 per cent in structural terms).
by 2.4 per cent. This is partly explained by the
This was due to a significant decline in the num-
increase in the number of old-age pensioners
ber of subsidised unemployed, which was more
and the update of the lowest pensions, for the
marked than as regards the overall number of
first time since 2010. Social benefits in kind, in turn,
unemployed individuals in the economy, and to
declined by 0.7 per cent, as a consequence of
a slight increase in the average benefit. In 2016,
lower expenditure in the health sector.
2.0
Chart 4.4 •
Breakdown
of the change
in structural
primary
expenditure
| In percentage
points of trend GDP
1.0
0.0
-1.0
-2.0
-3.0
-4.0
2012
2013
2014
2015
Investment
Intermediate consumption
Wage bill
2016
Other primary expenditure (b)
Pension expenditure
Change in structural primary expenditure
Sources: INE and Banco de Portugal.
Notes: (a) Other primary expenditure includes social payments excluding pensions, general government social contributions, subsidies
and other current and capital expenditure.
30.0
Chart 4.5 •
Breakdown of the
change in public
debt-to-GDP ratio
| In percentage
points of GDP
25.0
20.0
15.0
10.0
5.0
0.0
-5.0
-10.0
2012
2013
Primary balance effect
Deficit-debt adjustments
Sources: INE and Banco de Portugal.
2014
2015
GDP growth effect
Change in debt
2016
Cumul:
2012-2016
Interest effect
The Portuguese economy in 2016
In 2016 compensation of employees and inter-
in advance in 2016 (2.4 per cent of GDP). This
mediate consumption remained unchanged as
repayment contributed to reduce the share of
a ratio to trend GDP. Compensation of employ-
debt held by non-residents, from 66.5 per cent
ees increased by 2.8 per cent, after two years of
at the end of 2015 to 58.2 per cent at the end
decline. This is explained by the growth in wages
of 2016 (Chart 4.6). This effect was reinforced by
and salaries, as a result of the full reinstatement
the increased importance of Banco de Portugal
of the wage cuts in force since 2011 and, to a
as a holder of Portuguese public debt, reflecting
lesser extent, of developments in the number
the widening of the Eurosystem’s asset purchase
of general government employees.23 Indeed,
programme to include sovereign bonds.24 More
in 2016, against a backdrop of a decrease in
the number of weekly hours worked in general government, from 40 to 35, the number
of employees recorded, for the first time since
2011, a positive quarterly average change. In
the case of intermediate consumption, the
2.9 per cent increase was strongly influenced
by the rise in expenditure related to public-private partnerships in the road sector, which was
again significant. Excluding these outlays, this
expenditure item decelerated.
specifically, between 2014 and 2016 the percentage of public debt held by the Central Bank
increased from 0.7 to 8.8 per cent, implying an
increase in the Central Bank’s exposure to sovereign risk. This occurred against the decline in the
share of non-residents and, to a lesser extent,
of other MFIs as debt holders.
In the course of 2016, the Portuguese State participated regularly in sovereign debt markets.
Against a backdrop of a widening in the differential between the cost of financing 10-year Por-
The public debt ratio increased,
reflecting a significant
accumulation of deposits
tuguese and German debt, the allotment rates
At the end of 2016, the general government
2.3 in the previous year. As regards short-term
debt stood at 130.4 per cent of GDP, 1.4 p.p.
above the figure recorded at the end of the
previous year. In a context where the primary
balance contributed to reduce the debt ratio
(by -2.2 p.p. of GDP), the change was chiefly
explained by deficit-debt adjustments (3.1 p.p.
of GDP). The fact that the implicit interest rate
on debt exceeded the nominal GDP growth rate
has also contributed to increase the ratio (by
0.5 p.p. of GDP) (Chart 4.5).
As regards deficit-debt adjustments, the figure recorded in 2016 is mostly explained by the
increase in deposits held by the general government by an amount equivalent to 2.4 p.p. of
GDP. This deposit accumulation was likely related
to the need to finance the capital injection by the
State in Caixa Geral de Depósitos, which occurred
at the beginning of 2017. Therefore, in 2016 the
debt ratio net of general government deposits
declined from 118.9 to 118.1 per cent of GDP.
of long-term securities were higher than those
recorded in 2015 (Chart 4.7). In 2016, in particular, the average rate on 10-year Treasury
Bond auctions stood at 3.1 per cent, up from
issues, the average interest rate on Treasury Bill
auctions declined slightly from 0.04 to 0.02 per
cent.
General government debt servicing expenditure declined from 4.6 to 4.2 per cent of GDP
in the year under review, which is underpinned
by a reduction in the implicit interest rate from
3.6 per cent in 2015 to 3.3 per cent in 2016.25
These developments stem, on the one hand,
from a decline in the interest burden associated with loans under the Economic and Financial
Assistance Programme and, on the other hand,
from the repayment of Treasury Bonds with higher rates than those of the new issues.
The euro area structural
primary balance declined
for the second year in a row
Similarly to developments in 2015, part of the
According to the European Commission Winter
loan granted by the IMF under the Economic
forecasts, published in February 2017, the struc-
and Financial Assistance Programme was repaid
tural primary balance in the euro area declined
47
48
BANCO DE PORTUGAL • Economic Bulletin • May 2017
100
Chart 4.6 •
Breakdown
of public
debt by holder
| Percentage
90
80
70
60
50
40
30
20
10
0
2011
2012
Non-residents
2013
Central Bank
2014
Other monetary financial institutions
2015
2016
Other financial institutions
Other residents
Source: Banco de Portugal.
20.0
Chart 4.7 •
10 year government
bond yields
| Percentage
15.0
10.0
5.0
0.0
Germany
Spain
Sep. 16
Jan. 16
May 16
Sep. 15
Jan. 15
May 15
Sep. 14
Jan. 14
May 14
Sep. 13
Jan. 13
May 13
Sep. 12
Jan. 12
May 12
Sep. 11
Jan. 11
Portugal
May 11
Sep. 10
Jan. 10
May 10
Sep. 09
Jan. 09
May 09
Sep. 08
Jan. 08
May 08
Sep. 07
Jan. 07
May 07
-5.0
Italy
Source: Thomson Reuters.
Chart 4.8 •
Fiscal policy and
cyclical position in
the euro area in
2016
| In percentage
points of potential
GDP
iscal stance (change in the structural primary balance)
6
Pro-cyclical tightening of fiscal policy
Counter-cyclical tightening of fiscal
policy
5
4
PT 12
3
2
EA 12
PT 14
1
EA 13
PT 13
0
-1
EA 14
PT 16
EA 16
Counter-cyclical loosening of fiscal policy
EA 15
PT 15
Pro-cyclical loosening of fiscal policy
-2
-3.5
-3
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
Cyclical position (change in the output gap)
Source: European Commission (Winter 2017 Forecasts).
Note: The cyclical position of the economy is assessed by the change in the output gap, which broadly corresponds to the difference between
actual GDP and potential GDP growth rates.
The Portuguese economy in 2016
slightly in 2016 (Chart 4.8). This reflects declines
Pact, 2016 was the deadline for the correction
in the structural primary balance in most Mem-
of the excessive deficit in Portugal and Greece.
ber States and particularly in Spain, Cyprus, Aus-
In both cases, the nominal deficit stood below
tria and Italy. In turn, Greece, Malta, the Neth-
3 per cent of GDP (Table 4.2). Spain and France,
erlands, Latvia and Estonia followed fiscal con-
whose correction deadlines are 2018 and 2017
solidation strategies in 2016. Regarding eco-
respectively, recorded nominal deficits above
nomic activity, the Commission estimates that
the reference value.
GDP growth in the euro area exceeded the
In 2016 public indebtedness remained at a very
change in potential GDP, although the difference between these aggregates remained negative. This implies that fiscal policy was pro-cyclical in 2016, for the second year in a row.
49
high level in most euro area countries (Chart 4.9).
In particular, the public debt ratio stood below
the 60 per cent of the GDP reference value
in only 6 of the 19 Member States. Compared
Within the European fiscal surveillance frame-
with the previous year, there were important
work, the excessive deficit procedures for Ireland,
declines in Slovenia, Ireland, Germany and the
Slovenia and Cyprus were abrogated in 2016.
Netherlands. Nevertheless, in 2016 only Mal-
As regards the euro area countries subject to
ta recorded a public debt ratio below that of
the corrective arm of the Stability and Growth
2007, before the outset of the financial crisis.
Table 4.2 • Euro area countries in excessive deficit procedure | As a percentage of GDP
Country
European Council
decision date
Current deadline for
correction
Last change in the
correction year
Fiscal balance in 2016
Portugal
02-12-2009
2016
08-08-2016
-2.0
Spain
27-04-2009
2018
08-08-2016
-4.5
Greece
27-04-2009
2016
04-12-2012
0.7
France
27-04-2009
2017
10-03-2015
-3.4
Sources: Eurostat (Excessive Deficit Procedure – April notification) and European Commission.
200
180
Chart 4.9 •
Public debt-to-GDP
ratio in the Euro
Area in 2016
160
140
120
| As a percentage
of GDP
100
80
60
40
20
Source: Eurostat.
Greece
Italy
Portugal
Cyprus
Belgium
Spain
France
Euro area
Austria
Slovenia
Ireland
Germany
Finland
Netherlands
Malta
Slovakia
Lithuania
Latvia
Luxembourg
Estonia
0
50
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Box 4.1 | Structural developments in tax revenue in 201626
In 2016 the ratio of structural revenue from taxes and social contributions to trend GDP declined
by 0.9 p.p. vis-à-vis the previous year. This stemmed from a decline in taxes on income and
wealth and, to a smaller extent, in taxes on production and imports (Chart 1). In the year under
review, structural revenue from social contributions as a ratio to trend GDP remained unchanged.
Note that the decline in structural tax revenue occurred notwithstanding the positive contribution
of the permanent component of receipts obtained within the scope of PERES (0.05 p.p. of GDP).
In 2016 structural revenue from taxes on income and wealth declined by 0.8 p.p. of trend GDP,
driven by the effect of legislative changes. In particular, revenue from taxes on households
declined by 0.6 p.p. in structural terms, largely reflecting the elimination of the Personal Income
Tax surcharge introduced in 2013 and the effect of the reform of this tax in 2015 (which impacted
2016 outturn via an increase in net refunds). The impact of the Personal Income Tax reform
exceeded the initial estimates, which negatively affects the residual component. This component
also reflects the decline in receipts from final withholding taxes on capital revenue resulting from
the decrease in interest rates on financial investments. As regards the structural revenue from
taxes on corporate income, there was a decline by 0.2 p.p. of trend GDP, reflecting measures
approved in previous years. Indeed, the collection of the Corporate Income Tax was negatively
affected by the decline in the statutory tax rate from 23 to 21 per cent (within the scope of the
reform initiated in 2014), as well as by the change, in 2015, in the regime applicable to investment
funds, which brought forward to that year revenue that would have been collected in 2016. These
effects more than offset the overall positive impact of the change in the asset assessment regime
included in the State Budget for 2016 and the permanent part of the Corporate Income Tax revenue collected under PERES (Chart 2).
Chart 1 • Breakdown of the overall change in structural fiscal burden | In percentage
points of trend GDP
3.5
3.0
2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0
-1.5
-2.0
2012
2013
2014
2015
Legislation changes
Decoupling between the macro base and GDP (in nominal trend terms)
Impact of the tax elasticity (in structural terms)
Residual
Change in structural tax burden
Sources: INE and Banco de Portugal.
2016
The Portuguese economy in 2016
In structural terms, the ratio of revenue from taxes on production and imports to trend GDP
declined by 0.1 p.p. in 2016. This was chiefly due to the fall in VAT structural revenue (-0.2 p.p.),
which more than offset the increase in the collection of other indirect taxes (0.1 p.p.). In the case
of VAT revenue, special reference should be made to the negative effect associated with the cut
in the rate applicable to some restaurant services. The residual component of this tax is negative, reflecting the unusually high growth of refunds in the first months of 2016. However, this
may have been partly offset by the possible overestimation of the effect of the above mentioned
VAT rate cut. As regards the other indirect taxes, the structural increase in the ratio of revenue
to trend GDP was largely due to several legislative changes included in the State Budget for 2016
(yielding an overall impact of 0.3 p.p. of trend GDP). Due to the magnitude of their estimated
impacts, reference should be made to the contributions of the rises in taxes on oil products and
on tobacco (0.2 p.p. and 0.1 p.p. of trend GDP respectively). A negative residual (-0.2 p.p.) possibly
indicates the overestimation of these effects.27
Finally, the ratio of structural revenue from social contributions to trend GDP remained virtually
unchanged in 2016, reflecting two symmetrical effects. On the one hand, the fact that the private
sector trend wage bill grew less than nominal GDP should have given a negative contribution
to structural revenue from social contributions. On the other hand, similarly to previous years,
the residual component net of the effect of social contributions related to the civil servants
regime was positive.28 This result may suggest gains from increased efficiency in the collection of
contributions.
Chart 2 • Breakdown of the change in structural taxes and social contributions in 2016
| In percentage points of trend GDP
0.4
0.2
0.0
-0.2
-0.4
-0.6
-0.8
Taxes on households' Taxes on firms' income
income
VAT
Other taxes on
production and
imports
Legislation changes
Decoupling between the macro base and GDP (in nominal trend terms)
Impact of the tax elasticity (in structural terms)
Residual
Total
Sources: INE and Banco de Portugal.
Social contributions
51
52
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Box 4.2 | Analysis of deviations in budget execution in 2016
In 2016 the general government deficit stood below the first official estimate for the year, presented in the State Budget for 2016, and the revised estimate included in the State Budget for
2017. However, in the course of the year, different fiscal projections released by international
institutions suggested that the official target would be missed. Against this background, an analysis of the fiscal deviations gains particular relevance, allowing not only for a better understanding
of the 2016 outcome, but also for the identification of some risks in budget execution in the current year.
The State Budget for 2016 was published in early February and entered into force only in April.
The estimate for the deficit in the National Accounts included in this document was thus based
on information for 2015 with a lower than usual degree of uncertainty and appears to have
benefitted from some information on revenue developments at the beginning of the year. The
differential between budget outturn and this estimate turned out to be very small (0.2 per cent
of GDP), but is underpinned by a significant negative deviation in total expenditure that is almost
fully offset by a similar and also negative differential on the revenue side (Chart 1).
Indeed, total general government revenue outturn fell significantly short of the amount envisaged
when the State Budget for 2016 was prepared, reflecting, in particular, an overestimation of the
collection of non-tax current revenue (including sales) and capital revenue. Growth of tax collection was also lower than expected, particularly as regards taxes on production and imports. In the
latter case the deviation may reflect some overestimation of the impact of indirect taxation hikes
included in the State Budget for 2016. Note that the decrease in tax revenue occurred in spite of
the fact that developments in economic activity and, in particular, in private consumption were
close to the projection included the State Budget for 2016. In addition, the deviation might have
been more expressive if the PERES had not been implemented.
Chart 1 • Budget execution in 2016 in national accounts: deviations vis-à-vis the State
Budget for 2016 | In million euro
Total revenue
Taxes on income and wealth
Taxes on production and imports
Social contributions
Other current revenue
Capital revenue
Total expenditure
Social payments
Compensation of employees
Intermediate consumption
Interest
Subsidies
Other current expenditure
Investment
Other capital expenditure
Overall balance
-2,500
Sources: INE and Ministry of Finance.
-2,000
-1,500
-1,000
-500
0
500
1,000
The Portuguese economy in 2016
Total expenditure outturn stood also substantially below budgeted. A determining factor contributing to this result was the fact that the State Budget for 2016 stipulated a very significant amount
of expenditure in the form of provisions, reserves and spending appropriations. This raises significantly the scope for discretion in the implementation of the budget and, given the limited nature
of information available, hinders the monitoring of budget execution in the course of the year. For
further details on the utilisation of budget control instruments in 2016, see Box 3 of the report
‘Analysis of the General Government Account 2016’, published by the Portuguese Fiscal Council.
Both intermediate consumption expenditure and other current expenditure posted in 2016
much lower figures than those considered in the State Budget for 2016. Due to its magnitude,
reference should also be made to the negative deviation in public investment, which was likely be
affected by the limited execution of revenue from structural funds received from the European
Union. Finally, similarly to previous years, expenditure on debt servicing stood below forecasts.
In fact, although the budget pointed to the near stabilisation of this expenditure item in 2016,
it actually declined vis-à-vis the previous year.
53
54
BANCO DE PORTUGAL • Economic Bulletin • May 2017
5.Supply
Moderate growth of economic
activity in 2016
path observed since early 2014. This increase
As in the past few years, in 2016 Portuguese GDP
accommodation and food services (Chart 5.3).
per capita continued the convergence towards
The dynamics observed in these sectors was
the average EU level and currently stands at
associated with very favourable developments
around 71 per cent, still lower than observed
in tourism exports (9.7 per cent in real terms, fol-
at the onset of the international economic and
lowing an 8.9 per cent increase in 2015) and the
financial crisis (Chart 5.1).
dynamics of domestic demand. The services con-
In 2016 gross value added (GVA) at basic prices
fidence indicator maintained a positive stance in
mainly reflected 3.4 per cent activity growth in the
subsectors trade and repair of motor vehicles and
increased by 0.8 per cent year-on-year, following
2016 (Chart 5.4).
1.2 per cent growth in 2015. This increase consol-
There was a 0.7 per cent year-on-year increase
idated the moderate recovery path started at the
in industrial activity, in contrast to a 2.0 per cent
end of 2013, with overall developments consist-
rise in 2015. However, this was not accompa-
ent with the coincident indicator for activity and
nied by an increase in industrial confidence over
the economic sentiment indicator (Chart 5.2).
the year.
This notwithstanding, there was a deceleration
vis-à-vis the previous year and, in this context, the
GVA level was still around 3 per cent lower than
the value recorded in 2008.
In 2016 activity in the construction sector declined by 1.8 per cent, after a virtual stabilisation in 2015. In annual average terms, activity in
this sector remained on a downward trend. This
GVA in the services sector rose by 1.1 per cent
trend reflected this sector’s structural adjust-
year-on-year in 2016, as in the previous year.
ment to a lower level of activity, following high
Growth in this sector continued on the recovery
investment in construction in previous decades.
Chart 5.2 • GVA, Coincident indicator of activity
and economic sentiment indicator
76
3
Year-on-year rate of change,
in per cent
75
74
73
72
71
70
69
68
120
1
100
-1
80
-3
67
-5
Source: European Comission (Ameco).
Note: EU15 refers to the initial 15 members of the European Union.
2016
2015
2014
2013
2012
2011
2010
2009
2008
2007
2006
2005
2004
2003
2002
2001
2000
1999
66
60
2008 2009 2010 2011 2012 2013 2014 2015 2016
Real GVA
Coincident indicator of activity
Economic sentiment indicator (r.h.s.)
Sources: European Comission, INE and Banco de Portugal.
Index 1990-2013 = 100
Chart 5.1 • GDP per capita at current prices,
PPP adjusted, Portugal as a percentage of the European
Union (EU15)
The Portuguese economy in 2016
GVA in construction at the end of 2016 account-
declined by 377 and 345 thousand individuals
ed for little over half of the value recorded
respectively (Chart 5.5).
in 2008.
The demographic trends in the Portuguese
Downward trend of the
total population and labour
force
economy cannot be easily reversed, and emigration phenomena tend to compound them.
This is a significant drag on Portuguese economic growth, given that the labour factor is
an input for production, and the population
In 2016 the resident population and the labour
dynamics itself stimulates activity in various
force continued to decline, by 0.3 per cent (Ta-
sectors. In addition, a more aged society puts
ble 5.1). These developments continued to fol-
pressure on the sustainability of social security
low the downward trend observed since 2011,
schemes.
although slightly less markedly. Since 2008 the
resident population and the labour force have
decreased by approximately 250 and 350 thousand individuals respectively. Resident popula-
Improvement in labour
market conditions
tion in the 25-34 age group declined by 2.5 per
Labour market developments in 2016 continued
cent in 2016, while the labour force fell by
to be characterised by a rise in employment and
2.7 per cent. These falls are quite relevant, but
a decline in the unemployment rate, maintaining
less sharp than those seen in the past few years.
the improvement seen from the second quarter
In cumulative terms, since 2008 the popula-
of 2013 onwards, in a context of more marked
tion and the labour force in this age group have
wage dynamics.
Table 5.1 • Labour market indicators | Year-on-year rate of change, in per cent, unless otherwise
stated
Thousands of individuals
in 2016
Population
Population 25-34 years
Years
2014
2015
Semesters
2016
S2 2015 S1 2016 S2 2016
10 306
-0,6
-0,5
-0,3
-0,5
-0,3
-0,3
1180
-3,6
-2,8
-2,5
-2,7
-2,7
-2,3
Labour force
5178
-1,1
-0,6
-0,3
-0,5
-0,7
0,1
Labour force 25-34 years
1055
-3,8
-3,1
-2,7
-3,1
-2,9
-2,4
73,2
73,4
73,7
73,5
73,4
74,0
Participation rate 15-64 years
(in % of the population)
Total employment
4605
1,6
1,1
1,2
0,9
0,6
1,8
Employees
3787
4,4
2,8
2,1
1,9
1,7
2,4
Self-employment
789
-8,2
-5,7
-3,2
-3,6
-5,0
-1,3
Total unemployment
573
-15,1
-11,0
-11,4
-9,7
-10,0
-12,8
Unemployment rate
(in % of the labour force)
13,9
12,4
11,1
12,1
11,6
10,5
Unemployment rate 25-34 years
(in % of the labour force)
15,5
13,1
12,5
12,5
13,4
11,6
Long-term unemployment
(in % of total unemployment)
65,5
63,5
62,1
62,7
61,5
62,7
5,2
5,0
4,6
5,2
4,5
4,7
Discouraged inactives
(in % of the labour force)
Source: INE.
Notes: Long-term unemployment includes the unemployed individuals that have been actively seeking employment for 12 months or more.
The discouraged inactives include the inactive individuals who were available for work but had not looked for a job during the period.
55
56
BANCO DE PORTUGAL • Economic Bulletin • May 2017
According to data released by the Ministry of
cent in 2015), maintaining the downward trend
Labour, Solidarity, and Social Security, average
started in 2013 and standing close to the level
wages declared to Social Security in 2016 grew
observed in 2010. In 2016 the decline in the
by 1.6 per cent compared to the same period
total number of unemployed year-on-year was
a year earlier, which reflects a 1.0 p.p. accelera-
particularly marked in the 35-44 age group. The
tion from the previous year. This greater wage
share of unemployed receiving unemployment
dynamics should partly reflect the increase in
benefits amounted to 25.5 per cent in 2016,
the national minimum wage. The wage dynam-
against 28.5 per cent in 2015 (Chart 5.6).
ics plays a particularly important role in a con-
In addition, the number of discouraged work-
text where collective bargaining instruments,
particularly of a sectoral nature, are still a few,
although there has been a slight increase from
the minimum level recorded in 2012. In Portugal
the wages of around 90 per cent of employees
are defined within the scope of these instruments. This greater wage dynamics seems to
have also influenced the trend of the inflation
rate in 2016 (Section 7: ‘Prices’).
ers, i.e. individuals not actively seeking a job
but who are available for work, accounted for
around 4.6 per cent of the labour force in 2016,
slightly below the value recorded in 2015 (5.0 per
cent). The measure of the unemployment rate
that also includes discouraged workers continued to evolve in parallel with that of the official
unemployment rate, after having grown more
rapidly than the latter up to 2013 (Chart 5.7).
Nevertheless, this group is formed by around
The unemployment rate
declined, but remained
at very high levels
238 thousand discouraged workers.
One of the most negative aspects of Portuguese
labour market developments in the past few
years was the very high level of long-term unem-
According to Statistics Portugal’s Labour Force
ployment, which tends to cause a sharp depreci-
Survey, in 2016 the total number of unem-
ation of human capital, having an adverse impact
ployed declined by 11.4 per cent year-on-year,
on the economy’s potential growth. In 2016 the
after a 11.0 per cent fall in 2015 (Table 5.1). The
number of unemployed seeking a job for more
unemployment rate was 11.1 per cent (12.4 per
than 12 months fell by 13.4 per cent (13.7 per
Chart 5.3 • GVA by main sectors of activity
| Index 2008 Q1 =100
120
Chart 5.4 • Confidence indicators
| Balances (quarterly mean)
20
10
0
-10
-20
-30
-40
-50
-60
-70
-80
110
100
90
80
70
60
50
2008 2009 2010 2011 2012 2013 2014 2015 2016
Total GVA
Services
Construction
Source: INE.
Agriculture, forestry, and fishing
Manufacturing
2008 2009 2010 2011 2012 2013 2014 2015 2016
Manufacturing confidence indicator
Construction confidence indicator
Services confidence indicator
Source: European Comission.
Note: Seasonally adjusted figures.
The Portuguese economy in 2016
cent fall in 2015). Nevertheless, the weight of
decline in 2015). The share of these short-term
these individuals in total unemployment has
unemployed in the labour force already stands
remained at a very high level (62.1 per cent, cor-
at levels close to those observed prior to the
responding to a total of approximately 356 thou-
international financial crisis (Chart 5.8).
sand individuals). In turn, very long-term unemployment, covering those seeking a job for more
Employment recorded
a positive trend, although
remaining at historically
low levels
group of unemployed (47.5 per cent in 2016 on
average).29 In the same vein, in 2016 the number of unemployed seeking a job for less than
12 months fell by 7.8 per cent (5.7 per cent
Chart 5.5 • Population, labour force, and employment
– Total and age group (25-34 years) | Index 2008 =100
Unemployed receiving unemployment
benefits (as a percentage of total
unemployment)
105
Chart 5.6 • Number of unemployed receiving
unemployment benefits and coverage rate
100
95
90
85
80
75
70
2009
2010
2011
2012
2013
Population
Employment
Labour force (25-34 years)
2014
2015
45
350
40
300
35
250
30
25
200
20
150
15
100
10
50
5
0
0
2008 2009 2010 2011 2012 2013 2014 2015 2016
2016
Labour force
Population (25-34 years)
Employment (25-34 years)
Coverage rate of unemployment benefits
Number of unemployed receiving unemployment benefits (r.h.s.)
Source: INE (Labour Force Survey).
Source: INE (Labour Force Survey).
22
5
20
4.5
18
4
16
3.5
14
3
12
2.5
10
2
2011
2012
Official rate
Source: INE (Labour Force Survey).
Number of unemployed (Thousands)
than 24 months, accounted for almost half of the
2008
57
2013
2014
Including discouraged workers
2015
2016
Difference (r.h.s.)
Chart 5.7 •
Official
unemployment
rate and
unemployment
rate including
discouraged
workers
| Percentage
58
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Reduction in apparent
labour productivity
According to the Labour Force Survey, total
employment increased by 1.2 per cent in 2016,
after a 1.1 per cent increase in 2015. This reflects
a higher number of employees (2.1 per cent
The current recovery phase of the Portuguese
more), given that self-employment fell markedly
economy brings together moderate activity growth
(3.2 per cent). In sectoral terms, the main con-
and relatively stronger employment growth. In this
tribution to the year-on-year rate of change in
context, apparent labour productivity based on
employment was recorded in tradable servic-
the quarterly national accounts of Statistics Por-
es, which include tourism-related activities.30
tugal declined in 2016, in the context of lower
In spite of the upward trend, employment
and atypical dynamics compared to prior phas-
remained at historically low levels, in the wake
es of economic recovery.31 However, from 2011
of an unprecedented fall observed between the
to 2013 apparent labour productivity growth
end of 2008 and early 2013, which according to
was strong, also in contrast to prior recessive
quarterly national accounts corresponded to
phases of the cycle (Chart 5.9). The explana-
around 650 thousand individuals. Approximate-
tions for the reduction in productivity are com-
ly 290 thousand jobs seem to have been recov-
plex and may relate to changes in the produc-
ered by the end of 2016. With respect to general
tive structure, in a context where the levels of
government employment, according to informa-
capital per worker remained low, after several
tion from the Directorate-General for Adminis-
years of low investment levels (Box 5.1: ‘Capital
tration and Public Employment, the number of
per worker and productivity’).
employees appears to have increased in 2016,
as in 2015, which confirms the interruption of
the downward trend observed in previous years.
Chart 5.8 • Unemployment rate by duration
of unemployment | Months
20
18
16
14
12
10
8
6
4
2
0
Chart 5.9 • Apparent labour productivity in the
last three crisis and recoveries | Biannual figures;
Minimum level of activity = semester t=100
108
106
104
102
100
98
96
94
2008 2009 2010 2011 2012 2013 2014 2015 2016
Unemployment rate
Short-term unemployment (less than 12 months)
Long-term unemployment (12 months or more)
Source: INE (Labour Force Survey).
t-4
t-3
t-2
t-1
t
t+1
1993S1
t+2
2003S1
t+3
t+4
t+5
t+6
t+7
2013S1
Sources: INE and Banco de Portugal.
Note: Labour productivity was computed dividing GVA (volume chain-linked
data, millions of euro) by total (full-time equivalent) employment, except for
the crisis in 1992, in which GDP was considered instead.
The Portuguese economy in 2016
Box 5.1 | Capital per worker and productivity
The low growth of apparent labour productivity in the Portuguese economy in the past few decades causes concern as it limits the ability to sustain higher levels of consumption without causing external imbalances. A significant catching-up of output per worker to the euro area average
is a particularly important challenge, in a context where there are still weaknesses in the Portuguese economy.
Developments in the level of output per worker relate to various structural issues whose interaction is complex. In its simpler version, the economic growth theory points to the accumulation of
inputs and the evolution of the technological framework as determinants of output growth. In this
context, the level of output per worker depends on technology and increases with the level of the
available capital stock for each worker, with lower variations expected for higher capital intensities. Economic literature has addressed this issue, identifying more elaborate formulations that
include, for example, the quality of inputs and more flexible assumptions on their substitutability. In addition, from an empirical viewpoint, the literature has shown weaknesses in the statistical
information needed for this analysis, for example as regards capital stock measurement.32 The
existing estimates for the capital stock level in economies typically result from the accumulation
of annual investment flows, less an annual depreciation rate resulting from assumptions on the
lifetime of the different capital goods.
This box presents some statistical information to place the Portuguese economy in the context
of the euro area in the 1995-2016 period in terms of the capital-labour ratio and how it relates to
productivity, measured as the level of output per worker.
Panel (a) of Chart 1 shows the cumulative trend of the capital-labour ratio since 1999 in Portugal, jointly with the cumulative trend of each component. The demographic contraction trends
experienced by the Portuguese economy, and especially the sharp cyclical developments stemming from the macroeconomic imbalances’ adjustment process, led to a strong reduction in
Chart 1 • Net capital stock per worker (2010 prices)
b) Euro area countries
300
140
250
Thousand euros
150
130
120
110
100
200
150
100
90
50
80
0
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
Index 1999=100
a) Portugal
Capital stock
Capital per worker
Source: Ameco – European Commission.
Employment
LV LT SK SI MT EE PT CY DE GR IT BE ES FI IE LU FR AU NL
2016
1999
59
BANCO DE PORTUGAL • Economic Bulletin • May 2017
employment after 2008, only partly reversed from 2013 onwards. In turn, European Commission
estimates show a contraction of the capital stock after 2011, as a result of insufficient investment
levels to offset the depreciation of installed capital. In this context, the capital-labour ratio in the
Portuguese economy started to decline after 2013, maintaining a downward path in 2016.
Panel (b) of Chart 1 shows the capital-labour ratios in euro area countries in 1999 and 2016, with
the Portuguese economy positioned in the lower segment of the distribution. Although there may
be reservations as to the accuracy of capital stock estimates, especially in the countries that have
most recently joined the European Union and the euro area, the Portuguese economy’s position
is unfavourable.
The relationship among euro area economies in terms of capital intensity and labour productivity levels for the average of the period 2007-2016 is shown in Chart 2. As was to be expected,
the chart suggests a positive relationship between the two variables, and Portugal’s position is
far behind the more advanced euro area countries. Hence, although it is important to stress that
sectoral distribution and quality of investment are important determinants of growth, the reality seems to show a need for greater aggregate investment levels to increase the productivity of
Portuguese workers.
The analysis of the quality of the capital stock existing in the economies is difficult, and the statistical problems mentioned above remain. An imperfect assessment of the capital stock quality can be made from its structure. Chart 3 is based on data from the Penn World Tables 9.033 to
show the composition of the capital stock by type of capital goods for euro area countries in the
2012-2014 period, according to four categories: (residential and non-residential) structures;
machinery (including computers, communication equipment and other machines); transport and
other equipment (including software, other intellectual property goods and cultivated assets). In
this very elementary analysis, Portugal appears as the country with the greatest proportion of
Chart 2 • Capital and GDP per worker in the euro area (average 2007-2016)
| Thousand euros
120
LU
100
IE
GDP per worker
60
80
DE
60
CY
MT
40
SK
LV
20
LT
ES
IT
BE
FI
FR
NL
AU
GR
PT
SI
EE
0
0
50
100
150
Capital per worker
Source: Ameco – European Commission.
200
250
300
The Portuguese economy in 2016
structures in the capital stock, which necessarily implies a lower share of equipment in machines
and intangible goods. These latter categories tend to be associated with higher profitability rates,
and thus there seems to be room for improving the composition of the national capital stock.
Portuguese economic growth will benefit from higher levels of investment in the firms, sectors
and types of goods with higher profitability. This investment effort should be primarily based on
higher public and private saving rates, to avoid jeopardising the current balance in the external
accounts, which is a key element of the Portuguese economy’s adjustment process of the past
few years. However, creating favourable conditions for attracting foreign direct investment, especially targeted at creating new productive capacity, is also a way to increase the levels of capital
per worker without compromising macroeconomic stability. The structural conditions needed to
attract foreign direct investment are largely also needed for a correct allocation of investment
towards more productive purposes. Efficient market functioning, a level playing field, the reduction of the so-called framework costs, and the stability of the regulatory and tax environments are
all part of this set of conditions.
Chart 3 • Structure of capital stock per type of investment (average 2012-2014)
| Percentage
100
95
90
85
80
75
70
65
60
55
50
MT
SK
IE
BE
Structures
Source: Penn World Tables.
EE
AU
LU
SI
Machinery
FI
DE
NL
IT
ES
Transport equipment
CY
LT
FR
LV
Other
GR
PT
61
62
BANCO DE PORTUGAL • Economic Bulletin • May 2017
6.Demand
Moderate GDP growth based
on the dynamics of domestic
demand and especially
of exports
of the prior recession. In this context, the GDP
In 2016 the Portuguese economic activity grew
economic agents and the need to adjust their
level at the end of 2016 was around 4 per cent
lower than at the start of 2008. The recovery of
economic activity takes place in the context of
high indebtedness levels of public and private
by 1.4 per cent in volume, following a 1.6 per
balance sheets.
cent growth in 2015 (Table 6.1). In 2016 as a
The important demographic changes that oc-
whole GDP growth in Portugal was once again
curred in Portugal in the past few years ren-
lower than in the euro area, the differential
der the analysis of per capita GDP develop-
standing at -0.3 p.p. (Chart 6.1). GDP growth in
ments particularly relevant. In 2016 this varia-
2016 showed a markedly intra-annual profile,
ble grew by 1.7 per cent (2.1 per cent in 2015),
accelerating to 1.9 per cent in the second half
which exacerbated the gap vis-à-vis GDP de-
of the year, above the value observed on aver-
velopments (Chart 6.2). As in 2014 and 2015,
age in the euro area in the same period (1.7 per
cent).
GDP growth was once again higher than that of
GVA, which in 2016 grew by 0.8 per cent (Chart
Developments in Portuguese economic activity
6.3). The discrepancy between GDP growth and
in the most recent period continued to be char-
GVA growth related to the evolution of taxes less
acterised by moderate growth, particularly tak-
subsidies, which grew by 4.8 per cent in volume
ing into account the seriousness and duration
in 2016, compared to 5.1 per cent in 2015.
Table 6.1 • GDP and main components | Year-on-year change in percentage, unless otherwise
stated
% of GDP
in 2016 2015 2016
2016
GDP
Domestic demand
2015
2016
H1
H2
Q1
Q2
Q3
Q4
100.0
1.6
1.4
1.0
1.9
1.0
0.9
1.7
2.0
98.8
2.5
1.5
1.1
1.8
1.5
0.8
1.0
2.5
Private consumption
65.8
2.6
2.3
2.1
2.5
2.5
1.6
1.9
3.0
Public consumption
18.0
0.7
0.5
0.8
0.3
1.2
0.5
0.2
0.3
Investment
14.9
4.7
-0.8
-2.1
0.6
-2.1
-2.1
-1.8
3.0
14.8
4.5
-0.1
-2.4
2.2
-2.5
-2.2
-0.1
4.5
0.0
-0.1
0.0
-0.2
0.1
0.0
-0.3
-0.2
GFCF
Change in inventories (a)
Exports
40.3
6.1
4.4
2.8
6.0
3.7
1.9
5.6
6.4
Imports
39.1
8.2
4.4
3.1
5.6
4.8
1.5
3.9
7.3
Contribution of domestic demand net of imports (a)
1.1
0.5
0.4
0.6
0.4
0.4
0.4
0.9
Contribution of exports net of imports (a)
0.5
0.9
0.6
1.2
0.7
0.6
1.3
1.1
0.5
1.3
0.2
0.2
0.9
0.7
2.5
1.6
1.1
2.0
1.4
0.8
1.3
2.7
Memo item:
GDP – change over the previous period
Domestic demand (exc. change in inventories)
98.7
Sources: INE and Banco de Portugal calculations.
Note: (a) Contributions to the real growth of GDP in percentage points. The demand aggregates net of imports are obtained by subtracting an
estimate of the imports needed to meet each component. The calculation of import content was based on data for 2005. For more information,
see the Box entitled ‘The role of domestic demand and exports in economic activity developments in Portugal’, in the June 2014 issue of the
Economic Bulletin.
The Portuguese economy in 2016
Intra-annual acceleration
path of economic activity,
reflecting a recovery of both
domestic demand and exports
of goods and services in the
second half of the year
The less buoyant economic activity in 2016
reflects less robust domestic demand growth,
largely due to a fall in investment, in parallel
with a deceleration in exports of goods and
services, in line with lower domestic demand
growth. In this context, the reduction of trade
flows with Angola had an impact on the Portuguese economy, with a particularly negative
influence on exports of goods.34 In turn, private consumption grew slightly less than in the
previous year, which was accounted for by the
less buoyant consumption of non-food current
goods and durable goods. Public consumption,
determined by the dynamics of compensation
of employees and expenses on goods and services in general government, grew by 0.5 per
cent in 2016, i.e. slightly less than in 2015. The
deceleration in some demand components
with a greater import content translated into
lower import growth in 2016 as a whole.
63
In spite of the lower dynamics over the year as a
whole, developments in economic activity in 2016
were characterised by a clearly upward intra-annual profile. Hence, following 1.0 per cent growth
in the first half of 2016, GDP increased by 1.9 per
cent in the second half. The recovery of both domestic demand and exports of goods and services contributed to this evolution.
The more buoyant domestic demand in the
second half of 2016 was particularly evident in
the last quarter, when both private consumption and investment accelerated. In a context
of rapid expansion of the consumption of dura-
Considering demand components less imports
ble goods over the whole year, the consump-
(deducting from each component an estimate
tion of current goods accelerated in the sec-
of the imports needed to meet said demand),
ond half of the year. In turn, investment grew
the contribution from domestic demand to
slightly in the second half, after a fall in the first
GDP growth is estimated to have amounted
half of the year. This development had a major
to 0.5 p.p. in 2016 and the contribution from
contribution from the significant recovery in
exports to have stood at 0.9 p.p.
investment in machinery and equipment.
Chart 6.1 • GDP growth in Portugal and in the euro
area | Year-on-year change in percentage
4
Chart 6.2 • GDP and GDP per capita in Portugal
| 2011Q1=100
102
3
100
2
1
98
0
-1
96
-2
-3
94
-4
-5
-6
2000
2002
2004
2006
Portugal
2008
2010
2012
2014
Euro area
Sources: Eurostat, INE and Banco de Portugal calculations.
2016
92
2011
2012
2013
GDP
2014
2015
2016
GDP per capita
Sources: INE and Banco de Portugal calculations.
64
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Private consumption followed
an intra-annual recovery
path over the year, with
emphasis on consumption
of current goods which
accelerated considerably
in the last quarter
In 2016 private consumption grew by 2.3 per
cent, i.e. 0.3 p.p. less than in 2015. Despite the
lower dynamics, private consumption continued to grow more than GDP. The behaviour of
consumption over the whole year was chiefly
accounted for by a deceleration in consumption of current goods and services, in particular
the non-food component, in a context where
consumption of durable goods grew once
Intra-annual developments in private consumption showed an upward profile, as a result of particularly pronounced growth in the last quarter
of the year, based on an acceleration in the consumption of current goods and services and the
maintenance of the high growth of durable goods,
particularly sharp in the last quarter of the year.
The more favourable developments in private
consumption in the second half of the year occurred in an environment of sharp improvement
in consumer confidence, which remained at historically high levels, and a higher household disposable income. The improvement in the consumer confidence indicator at the end of 2016
was particularly noticeable in the assessment of
unemployment expectations, which lowered considerably.
again considerably, though less than in 2015
In 2016 the savings rate stood at 4.4 per cent of
(Chart 6.4). In 2016 the consumption of current
disposable income, i.e. 0.1 p.p. below the rate
goods and services grew by 1.6 per cent, after
recorded in 2015 (Chart 6.6), reflecting slightly
1.8 per cent in 2015, while the consumption of
higher nominal growth of private consumption
durable goods grew by 9.5 per cent (11.9 per
than that of disposable income (3.4 and 3.2 per
cent in 2015). In particular, passenger car sales
cent respectively). The increase in household dis-
grew by 16.1 per cent (Chart 6.5).
posable income benefited from an improvement
Chart 6.3 • Recent behavior of GDP and GVA
in Portugal | 2011Q1=100
Chart 6.4 • Main contributions for the growth
of private consumption in volume | Year-on-year
change in percentage, and contributions in percentage
points; quarterly values
4
100
2
98
0
96
-2
-4
94
-6
92
2011
-8
2012
2013
2014
2015
GDP
Sources: INE and Banco de Portugal calculations.
2016
GVA
2011
2012
2013
2014
2015
2016
Durables (exc. vehicles)
Durables (vehicles)
Non-durables
Private consumption (growth in percentage)
Sources: INE and Banco de Portugal calculations.
65
The Portuguese economy in 2016
in the labour market situation, especially in terms
In 2016 the volume of investment fell by 0.8 per
of employment, in parallel with a rise in the
cent following a 4.7 per cent increase in 2015.
national minimum wage and the impact of meas-
This development involves a contraction of gross
ures included in the State Budget for 2016 relat-
fixed capital formation of 0.1 per cent, following
ing to the reversal of wage cuts and the person-
4.5 per cent growth in the previous year.
al income tax surcharge (Section 4: ‘Fiscal policy and situation’).35 Reference should also be
made to the potential effect of the reduction of
the debt service in the most recent period, stemming from the stabilisation at low levels of the
interest rates prevailing in the household sector,
After a sharp fall observed during the economic and financial crisis, the necessary recovery of
GFCF in Portugal has been sluggish. This is also
observed in other economies, namely the euro
area economy, where the investment level is still
jointly with a decline in indebtedness.36 Against
lower than seen in 2008 (Chart 6.8). Although
this background, households’ financing con-
there is no consensus on the motives for these
ditions improved, notably as regards consumer
developments, they are particularly negative for
credit (Section 3.2: ‘Monetary and financial condi-
Portugal, not only because the fall in investment
tions in Portugal’). The weight of total consumer
during the crisis period was relatively sharper,
credit in private consumption has been following
but also because capital per employee in the
an upward path, standing at values higher than
Portuguese economy remains at low levels com-
seen prior to the sovereign debt crisis (Chart 6.7).
pared with the euro area average (Box 5.1: ‘Capi-
It is an expectable development that nevertheless needs to be monitored with caution.37
tal per worker and productivity’).
In this context, the higher uncertainty, both domestic and external, in late 2015 and early 2016,
Fall in investment largely
reflecting a decline in GFCF
in construction
in parallel with the maintenance of a high level of corporate indebtedness, appears to have
contributed to restrain investment decisions,
Chart 6.5 • Sales of light passenger vehicles
| Thousands of vehicles; annual values
Chart 6.6 • Savings rate and households’ disposable
income | Percentage change and savings rate
as a percentage of households’ disposable income
12
10
8
6
4
2
0
-2
-4
-6
300
250
200
150
100
50
2001
2003
2005
2007
2009
2011
2013
Sales of light passenger vehicles
(sum of the previous 12 months)
Average 2001-2016
Sources: INE and Banco de Portugal calculations.
2015
2000
2002
2004
2006
2008
2010
Private consumption deflator
Nominal disposable income
Private consumption (nominal)
Sources: INE and Banco de Portugal calculations.
2012
2014
2016
Private consumption (real)
Savings rate
66
BANCO DE PORTUGAL • Economic Bulletin • May 2017
namely in the first half of the year (Section 2: ‘In-
the past few years, only interrupted in 2015,
ternational environment’ and Section 3: ‘Mon-
is part of a stabilisation trend of the Portu-
etary and financial conditions’). However, in-
guese housing stock, following a considerable
vestment recovered in the second half of 2016,
increase in the 1990s. This seems to have been
particularly in the last quarter, with gross capital
reinforced in the most recent period by a need
formation growing by 4.5 per cent year on year.
to reduce household indebtedness levels. By
According to the Investment Survey released by
contrast, corporate GFCF grew by 6.6 per cent
Statistics Portugal, in January 2017, 39.4 per cent
in volume in 2016, following a rate of change of
of firms claimed to have had their investment re-
2.3 per cent in the previous year and showing
strained in 2016. This was a historically low val-
a markedly upward intra-annual profile.
ue that reflected a considerable fall vis-à-vis the
The fall in GFCF in 2016 was largely accounted
share computed in July 2016 (50.2 per cent).
In addition, although the deterioration of sales
prospects continued to be the main constraint
on investment, their weight decreased substantially backed by factors such as self-financing capacity or the difficulty to obtain credit (Chart 6.9).
for by the construction component. In the wake
of the declines observed since 2002, followed
by a significant recovery in 2015, this investment component, whose weight in total GFCF
was around 50 per cent, declined once again
sharply. This may partly reflect the effect of the
By institutional sector, the reduction in gross
conclusion of a number of major public works
fixed capital formation in 2016 as a whole
at the end of 2015. GFCF in construction recov-
largely reflected a considerable decline in pub-
ered slightly at the end of the year (Chart 6.10),
lic investment, which fell by 29.2 per cent in vol-
in line with the growth of cement sales to the
ume, after having increased by 16.4 per cent in
domestic market. In addition, the number and
the previous year. Residential investment also
amount of tenders for public works grew signifi-
declined in 2016 (-1.9 per cent), with an impact
cantly in the second half of 2016 (Chart 6.11).
on activity in the construction sector (Section
After major falls at the end of 2015 and in
5: ‘Supply’). The fall in residential investment in
the first half of 2016, GFCF in machinery and
Chart 6.7 • Share of private consumption in new loans
granted by banks to consumption | New operations
Chart 6.8 • Developments in GFCF in selected euro
area countries | 2008Q1=100
in percentage
110
8
7
100
6
5
90
4
80
3
2
70
1
0
2009
2010
2011
2012
2013
2014
2015
2016
60
2008 2009 2010 2011 2012 2013 2014 2015 2016
Euro area
Italy
Sources: INE and Banco de Portugal calculations.
Germany
Spain
Sources: INE and Banco de Portugal calculations.
France
Portugal
The Portuguese economy in 2016
equipment recovered strongly in the third and
from the value recorded in 2015. The posi-
fourth quarters. This investment component
tive performance of exports is structural, rely-
grew by 2.2 per cent in the year as a whole
ing on a corporate restructuring that started
(6.0 per cent in 2015). In turn, GFCF in trans-
before the international financial crisis, and is
port equipment, although decelerating from
not accompanied by systematic declines in unit
2015, continued to grow considerably in 2016
prices (Box 6.1: ‘Developments in unit values of
(8.6 per cent).
Portuguese exports of goods’). As regards ser-
67
vices, exports of services not related to tourism
Deceleration in exports
of goods and services
reflecting a fall in energy
exports and the lower
growth of exports of services
not related to tourism
decelerated considerably. In turn, exports of
In 2016 exports of goods and services grew by
The lower growth of the international economy
4.4 per cent, compared to an increase of 6.1 per
contributed to the less favourable performance
cent in 2015. This reflects the behaviour of the
of exports in 2016 as a whole compared to the
goods component, which grew by 4.3 per cent,
previous year. In this context, exports to some
after 6.3 per cent in 2015, and of the services
emerging economies, especially Angola, contin-
component, which decelerated by 1.1 p.p. to
ued to make a negative and significant contri-
4.6 per cent (Chart 6.12). In particular, energy
bution to the change in total exports. However,
exports fell, after growing considerably in 2015.
exports of goods and services showed a clear
Excluding this component, exports of goods
upward intra-annual profile, widespread across
grew by 5.1 per cent, i.e. accelerating by 1.8 p.p.
all the main components.
tourism services continued to be highly buoyant, in line with developments in their nominal
revenue and the considerable growth of the
number of overnight stays of non-residents in
national hotels. In 2016 exports of tourism services grew by 9.7 per cent in volume (8.9 per
cent in 2015).
Chart 6.9 • Main limitation to investment
| As percentage of total firms that reported limitations
Chart 6.10 • Developments in GFCF in Portugal
by type of investment | 2008Q1=100
to investment
Other
120
Level of interest rates
100
Access to bank credit
80
Self-financing capacity
60
Investment profitability
40
Deterioration of sales prospects
0
10
20
30
40
50
60
70
20
2008 2009 2010 2011 2012 2013 2014 2015 2016
GFCF Total
GFCF – Construction
GFCF – Transport equipment
GFCF – Machinery and equipment
Share of firms (percentage)
2016
Source: INE.
2015
2014
2013
Source: INE.
68
BANCO DE PORTUGAL • Economic Bulletin • May 2017
The deflator of goods exports fell in 2016 (-3.4 per
The growth of the volume of exports of goods
cent) though less markedly when excluding the
and services in 2016 was higher than that of
fuels component (-2.4 per cent). Based on inter-
the external demand indicator normally used
national trade data, exports of goods grew by
by Banco de Portugal, calculated on the basis
1.0 per cent in nominal terms in 2016 (3.7 per
of the information used within the Eurosys-
cent in 2015). This was largely due to a fall in car
tem, leading to a further gain in the export
exports, particularly to Germany, and a decline
in exports of fuels and base metals. The reduction in car exports reflected the temporary closure of an important production unit in early
2016. By destination country, there was a sharp
fall in exports of goods to non-EU markets, specifically Angola and China, and to the Netherlands. Angola’s contribution to the growth of
nominal exports of goods in 2016 was -1.2 p.p.,
which corresponds to a drop of 28.4 per cent in
share. However, this external demand indicator does not consider the relative importance
of external trade with Angola. When taking into
account the actual weight of Angola and the
trend of its imports, external demand for Portuguese goods and services grew less in the most
recent period, which implied a slightly higher
gain in the share of exports than that obtained
with the indicator normally used (Chart 6.14).
the year as a whole. This adverse development
Considering the main intra-EU trading part-
faded away over the course of 2016 (Chart 6.13),
ners, exports of Portuguese goods and servic-
contributing to the intra-annual acceleration
es to these destinations have been experiencing
profile of exports.
overall higher nominal growth than that of total
38
Chart 6.11 • Number and amount of tenders
for public work | Year-on-year change in percentage;
half-yearly values
Chart 6.12 • Main contributions for the growth
of goods and services exports in volume
| Year-on-year change in percentage, and contributions
in percentage points; quarterly values
60
10
40
8
6
20
4
0
2
-20
0
-40
-2
H1
H2
2014
H1
H2
2015
Call for tenders - number
Source: AECOPS.
H1
H2
2016
Call for tenders - value
-4
2012
2013
2014
2015
2016
Other services
Tourism
Exports of energy goods
Exports of goods (exc. energy)
Exports of goods and services (year-on-year change, percentage)
Sources: INE and Banco de Portugal calculations.
69
The Portuguese economy in 2016
imports of goods and services from those coun-
cent, after 15.0 per cent growth in 2015) and a
tries, understood as part of external demand for
less buoyant imports of services, which grew
Portuguese goods and services (Chart 6.15). For
by 2.6 per cent in 2016, after a 6.7 per cent
example, in 2016 nominal exports of goods and
change in 2015 (Chart 6.16). Imports of goods
services from Portugal to Spain grew by 5.8 per
excluding energy grew by 6.5 per cent, decel-
cent, while total Spanish imports grew by 1.6 per
erating slightly from the value observed in 2015
cent. Although expressed in nominal terms, this
(7.1 per cent).
behaviour suggests that Portuguese exports performed well in the Spanish market.
The deceleration in imports of goods in 2016 is
consistent with the less buoyant behaviour of
The deceleration in imports
of goods and services continued,
linked to a fall in exports
of energy and services
not related to tourism
a number of demand components with a high
import content, with a particular emphasis on
a fall in exports of fuels and a significant deceleration in GFCF in machinery and equipment.
Following the path observed since early 2013,
the deflator of goods imports fell further in
2016 (-3.9 per cent). This reduction is less sig-
Imports of goods and services grew by 4.4 per
nificant when excluding fuel exports, standing
cent in 2016 (8.2 per cent in 2015). This chief-
at -2.4 per cent. Based on international trade
ly reflected a fall in energy imports (-4.0 per
data, imports of goods recorded a 1.3 per cent
Chart 6.13 • Contributions from main trading markets
for the nominal growth of goods exports excluding fuel
| Year-on-year change in percentage, and contributions
Chart 6.14 • Exports of goods and services and
external demand | Year-on-year change in percentage
in percentage points; quarterly values
12
15
9
10
5
6
0
3
-5
0
-10
-3
-15
-20
-6
2012
2013
2014
2015
2016
China
Angola
Extra-EU exc. Angola and China
Intra-EU
Nominal exports exc. energy (growth in percentage)
Source: INE (International Trade) and Banco de Portugal calculations.
2008
2010
2012
2014
2016
External demand
Portuguese exports
External demand including Angola
Sources: INE, IMF, ECB and Banco de Portugal calculations.
Note: The indicator of external demand adjusted by the importance of foreign
trade with Angola corresponds to the weighted average (weights based on exports) between the indicator of external demand computed by the ECB and
Angola’s total volume of imports of goods and services. These calculations use
the IMF’s projections (Word Economic Outlook) for the growth in volume of
Angola’s imports in 2016.
70
BANCO DE PORTUGAL • Economic Bulletin • May 2017
nominal growth in 2016. By product, in addi-
3.9 per cent, with imports originating in Angola
tion to a strong reduction in fuel purchases,
dropping by 29.1 per cent.
also imports of base metals dropped. Similar-
In 2016 imports of goods and services grew
ly to exports, the lower dynamics of imports
more than the import-weighted overall demand,
of goods was particularly evident in non-EU
which led to an increase in import penetration,
markets. In 2016 imports of goods from these
as in the three previous years.
countries recorded a nominal reduction of
Chart 6.15 • Exports of goods and services for the
main EU trading partners in 2016 and total imports
from these countries | Year-on-year change
Chart 6.16 • Main contributions for the growth
of goods and services imports in volume | Year-on-year
change in percentage, and contributions in percentage
points; quarterly values
in percentage; nominal values
Growth of Portuguese nominal
exports of goods and services by
country of destination; percentage
16
45º
16
12
12
Italy
UK
8
4
France
8
Spain
4
0
Germany
0
-4
Netherlands
-4
-8
-12
-8
-8
-4
0
4
8
12
16
Growth of total nominal imports of goods and services; percentage
Sources: Eurostat, INE and Banco de Portugal calculations.
Note: The circles represent the locus of combinations between the growth of
nominal imports of goods and services in each of the 6 countries considered (X-axis) and the growth of nominal exports of goods and services from
Portugal to each of those countries (Y-axis). The size of the circles reflect the
importance of each country in the Portuguese exporting structure.
2012
2013
2014
2015
2016
Other services
Tourism
Imports of energy goods
Imports of goods (exc. energy)
Imports of goods and services (year-on-year change, percentage)
Sources: INE and Banco de Portugal calculations.
The Portuguese economy in 2016
Box 6.1 | Developments in unit values of Portuguese goods exports
This box complements the analysis of developments in Portuguese goods exports, focusing on
changes in their unit values from a predominantly microeconomic viewpoint.
The analysis presented here is based on unit values, which approximate the behaviour of exported
goods prices. Nevertheless, prices of internationally traded goods depend on a complex set of
factors, including the quality of products and the ability to exercise market power. All calculations are based on both Comext data and detailed reporting on international trade transactions
conducted by firms, within the scope of Intrastat and Extrastat surveys. In addition, considering the volatility associated with energy prices, their unit values were excluded from the analysis
based on corporate information.
Chart 1 presents the annual rate of change in the unit values of exported goods for the 2003-15
period. This growth rate has always been positive, except for 2003, 2009 and 2013-14. The most
exported goods over the whole period and, thus, contributing the most to the total rate of change
– namely ‘vehicles’ and ‘machinery, equipment and electrical components’ –, account on average
for 17 per cent of the total value of goods exports. The growth rates of these two categories of
the Standard International Trade Classification (SITC) show that their unit values have increased
after 2010, especially in the machinery and equipment component.
Analysing the growth of exports’ unit values is more meaningful if compared to developments
in a reference region. Hence, taking into account the existing institutional and economic environment, the natural benchmark is the average rate of change in the unit values of goods
exported by the European Union (EU).39 The difference between the growth rates recorded in
Portugal and in the EU was centred around zero for the 2003-15 period, which seems to exclude
a scenario where Portuguese exporters sought to systematically compete on price (Chart 2).
Chart 1 • Unit-values growth rates of Portuguese goods exports (including energy
goods) | Percentage
8
4
0
-4
-8
-12
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
Total (incl. Energy goods)
Road vehicles (including air-cushion vehicles)
Electrical machinery, apparatus and appliances, n.e.s., and electrical parts thereof
Souce: Banco de Portugal calculations based on Comext dataset.
2013
2014
2015
71
72
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Among the products whose unit value growth rates surpassed those of the EU are, consistently,
‘articles of apparel and clothing accessories’, ‘crude animal and vegetable materials, n.e.s.’ and
‘organic chemicals’ (Chart 3). Finally, ‘footwear’ is the only category whose unit value growth rate
was always positive in the period under analysis, and in most years higher than the EU average
rate (Chart 4).
Corporate information offers an analysis of the relationship between the differences of unit values and firm characteristics, namely in terms of sector of activity and age. The share of goods
exporters by sector whose unit value growth rates are higher than the EU weighted average rate
Chart 2 • Differences in unit-values growth rates (Portugal vs EU) | Percentage points
20
15
10
5
0
-5
-10
-15
-20
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
Source: Banco de Portugal calculations based on Comext dataset.
Note: The interval delimited by ‘+‛ signs represents the sample range after eliminating outliers – following Tukey (1977)'s method.
The interval defined by the vertical line corresponds to the interquartile range and the circle to each distribution's median.
Chart 3 • Unit-values growth rate of articles of apparel and clothing accessories
| Percentage
8
6
4
2
0
-2
-4
-6
2002
2003
2004
2005
2006
2007
2008
Portugal
Source: Banco de Portugal calculations based on Comext dataset.
2009
2010
2011
EU
2012
2013
2014
2015
The Portuguese economy in 2016
is shown in Chart 5. In the period under review, on average, this share stood at 50 per cent both
for manufacturing and services firms. In 2011 and 2012 there was a decline in the proportion
of goods exports by service firms whose unit value rate was higher than the EU’s. However,
in the 2013-15 period as a whole, this percentage surpassed the previous, which may indicate
an increase in these firms’ competitiveness. Service firms that export goods may be international
trade wholesalers or may represent the commercial arm of corporate manufacturing groups.
As regards exporters’ age, there is volatility regarding prevalent changes in unit values above
the EU average. However, the share of exports associated with firms aged less than five years and
Chart 4 • Footwear unit-values growth rate | Percentage
6
5
4
3
2
1
0
-1
2003
2004
2005
2006
2007
2008
2009
2010
Portugal
2011
2012
2013
2014
2015
EU
Source: Banco de Portugal calculations based on Comext dataset.
Chart 5 • Exports' share with positive difference in unit-values growth rate
(Portugal vs EU) by sector of activity | Percentage
100
90
80
70
60
50
40
30
20
10
0
2003
2004
2005
2006
2007
Manufacture
2008
2009
2010
2011
2012
Services
Sources: Banco de Portugal calculations based on Intrastat, Extrastat surveys and Comext dataset.
2013
2014
2015
73
74
BANCO DE PORTUGAL • Economic Bulletin • May 2017
between five and ten years whose growth rate in exported goods’ unit values was above that seen
in the EU on average is considerable, often surpassing that of firms aged more than ten years, as
was the case in the 2013-15 period (Chart 6).
Chart 6 • Exports' share with positive difference in unit-values growth rate
(Portugal vs EU) by age cohort | Percentage
100
90
80
70
60
50
40
30
20
10
0
2003
2004
2005
age<=5
2006
2007
2008
2009
5<age<=10
2010
2011
2012
2013
age>10
Sources: Banco de Portugal calculations based on Intrastat, Extrastat surveys and Comext dataset.
2014
2015
The Portuguese economy in 2016
7.Prices
Slight increase in the inflation
rate, based on services
and unprocessed food prices
although to a lesser extent, to the contribution
In 2016 the inflation rate in Portugal as meas-
by 0.3 p.p. to 1.6 per cent in 2016, while the rate
ured by the rate of change in the Harmonised In-
of change in processed food prices was only
dex of Consumer Prices (HICP) was 0.6 per cent,
0.3 per cent, i.e. 0.9 p.p. below the 2015 value.
compared to 0.5 per cent in 2015 (Table 7.1 and
Chart 7.1). In the year as a whole, services and
unprocessed food prices were more buoyant.
By contrast, prices of industrial goods, in particular energy, recorded a negative change.
An analysis of contributions from the different
categories of goods to the inflation rate in 2016
(Chart 7.2) shows a positive contribution from the
change in services prices (0.7 p.p.) and to a less-
of telephone and telefax services, financial services, and also restaurants and cafés. The rate
of change in unprocessed food prices declined
Non-energy industrial goods maintained a negative rate of change (-0.3 per cent), although less
sharp than in the previous year (-0.7 per cent).
In turn, energy prices remained on the downward trend started in November 2013, declining by 1.8 per cent in 2016, after a 3.7 per cent
fall in the previous year, in a context of recovery
in oil prices throughout 2016.
er extent in unprocessed food prices (0.2 p.p.).
Energy prices contributed to
the rise in inflation, in a context
of recovery in oil prices and an
increase in the tax on oil products
In turn, energy and non-energy industrial goods
continued to make a negative contribution in
2016 (-0.1 and -0.2 p.p. respectively), while processed food made a virtually nil contribution.
In 2016 services prices grew by 1.5 per cent,
accounting for a 0.1 p.p. rise from 2015. This
The pattern of pre-tax fuel prices largely reflec-
was essentially due to accommodation services,
ted a decline in oil prices, and that of final prices
which are part of the very dynamic Portuguese
was mainly influenced by a rise in the tax on oil
tourism sector. Reference should also be made,
prices in February 2016 (Charts 7.3 and 7.4).
Table 7.1 • HICP – Main components | As a percentage
Weights Year-on-year rate of change
Annual rate of change
2016
2014
2015
2016
16 Q1
16 Q2
16 Q3
16 Q4
100.0
91.9
-0.2
0.0
0.5
0.8
0.6
0.9
0.4
0.8
0.5
0.9
0.7
1.0
0.8
0.7
81.5
0.2
0.7
0.8
0.8
0.9
0.8
0.6
57.2
24.0
10.4
13.5
33.3
25.2
8.1
42.8
-1.1
-0.7
-2.1
0.4
-1.4
-1.4
-1.5
1.1
-0.1
1.5
1.9
1.2
-1.3
-0.7
-3.7
1.4
0.0
0.8
1.6
0.3
-0.7
-0.3
-1.8
1.5
-0.2
0.4
0.1
0.5
-0.7
0.2
-3.0
1.4
-0.3
0.6
1.4
0.1
-1.1
-0.2
-3.6
1.7
0.1
1.5
3.3
0.2
-0.8
-0.2
-2.6
1.5
0.3
0.8
1.5
0.2
-0.1
-0.8
2.0
1.6
–
–
0.3
0.1
0.1
0.2
0.1
-0.1
0.1
0.2
0.1
0.3
0.2
-0.4
0.2
-0.4
Consumer Price Index (CPI)
–
-0.3
0.5
0.6
0.5
0.5
0.7
0.8
HICP – Euro area
–
0.4
0.0
0.2
0.0
-0.1
0.3
0.7
Total
Total excluding energy
Total excluding unprocessed
food and energy
Goods
Food
Unprocessed food
Processed food
Industrial
Non-energy
Energy
Services
Memo items:
Contribution of administered
prices (in p.p.)
Contribution of taxes (in p.p.)
Sources: Eurostat and Statistics Portugal.
75
76
BANCO DE PORTUGAL • Economic Bulletin • May 2017
This contrasts with developments in 2015, when
profit margins and refining costs made a positive
contribution.
Positive inflation differential
between Portugal and
the euro area continued
In 2016 the inflation rate in Portugal was higher than the euro area’s, which stood at 0.2 per
cent, maintaining the differential recorded in
2015 (Chart 7.5). Although the inflation rate in
the euro area increased markedly in the second
half of the year, only in November did it surpass
inflation in Portugal. Energy industrial goods
accounted for most of the inflation differential
between Portugal and the euro area in 2016
(Chart 7.6). In this component, the differential
moved from 3.1 p.p. in 2015 to 3.3 p.p. in 2016.
The services differential rose to 0.4 p.p. in 2016,
after 0.2 p.p. in 2015. Conversely, non-energy
industrial goods prices grew at a slower pace
than in the euro area, as seen since early 2012.
Unit labour costs made the main
contribution to the inflation rate
Chart 7.7 breaks down the year-on-year inflation rate into the contribution of the main HICP
determinants based on Banco de Portugal’s
5.0
Chart 7.1 •
Harmonized index
of consumer prices
| In per cent
4.0
3.6
3.0
2.8
2.0
1.0
1.4
0.6
0.4
-0.2
0.0
0.5
-1.0
2010
2011
2012
2013
YoY rate of change
2014
2015
2016
Annual average rate of change
Source: Statistics Portugal.
4.0
Chart 7.2 •
Contributions to
the annual growth
rate of the HICP
| In percentage
points
3.0
2.0
1.0
0.0
-1.0
2010
2011
2012
Unprocessed food
Energy
Services
Source: Statistics Portugal.
2013
2014
2015
Processed food
Non-energy industrial goods
Total
2016
The Portuguese economy in 2016
Chart 7.3 • Contributions to the year-on-year rate
of change of petrol prices | In percentage points
77
Chart 7.4 • Contributions to the year-on-year rate
of change of diesel prices | In percentage points
24.0
16.0
18.0
8.0
12.0
0.0
6.0
0.0
-8.0
-6.0
-16.0
-12.0
2011
2012
2013
2014
Oil price
Taxes
2015
2016
2011
Profit margins and refining costs
Final price
2012
2013
2014
Oil price
Taxes
Sources: ECB and Directorate General for Energy and Geology.
2015
2016
Profit margins and refining costs
Final price
Sources: ECB and Directorate General for Energy and Geology.
5.0
Chart 7.5 •
HICP inflation
in Portugal and
in the euro area
4.0
3.0
2.0
|Year-on-year rate
of change
in percentage
and percentage
points
1.0
0.0
-1.0
-2.0
Jan. 10
Jul. 10
Jan. 11
Jul. 11
Jan. 12
Jul. 12
Jan. 13
Jul. 13
Differential
Jan. 14
Jul. 14
Portugal
Jan. 15
Jul. 15
Jan. 16
Jul. 16
Euro area
Sources: Eurostat and Statistics Portugal.
1.5
Chart 7.6 •
Inflation differential
between Portugal
and the euro area
| Contributions,
1.0
0.5
0.0
-0.5
in percentage points
-1.0
-1.5
-2.0
2008
2009
2010
2011
2012
Unprocessed food
Energy
Services
Sources: Eurostat and Statistics Portugal.
2013
2014
2015
2016
2016 Q1 2016 Q2 2016 Q3 2016 Q4
Processed food
Non-energy industrial goods
Total
78
BANCO DE PORTUGAL • Economic Bulletin • May 2017
inflation analysis and forecasting model, the so-
(1.4 per cent and 9.2 per cent in the fourth quar-
called MIMO (Monthly Inflation Model).
ter respectively).
The analysis suggests that the rise in the infla-
Inflation expectations released by Consensus
tion rate over the course of 2016 essentially
Economics remained stable at around 0.6 per
reflected an increase in unit labour costs, whose
cent in 2016, in line with actual average infla-
contribution went up over the year, consistently
tion (Chart 7.8). This contrasts with the succes-
with an increase in compensations per employ-
sive upward revisions of inflation expectations
ee (Section 5: Supply). By contrast, non-ener-
observed over the course of 2015. Inflation
gy goods import prices contributed to mitigate
expectations for 2017 also remained relative-
the rise in the inflation rate throughout 2016.
ly stable throughout 2016, at around 1.2 per
Fuels, excluding taxation, also made a nega-
cent, i.e. there is an implicit expectation of price
tive but progressively lower contribution in the
inflation between 2016 and 2017 (Chart 7.9).
40
recent period, mainly reflecting the time-path of
oil prices. The Value Added Tax (VAT) applied to
restaurants declined in the second half of 2016,
generating a negative contribution from indirect taxation to inflation dynamics. In the third
quarter of 2016, the item Others – which covers
other unidentified factors contributing to infla-
GDP deflator growth higher
than HICP growth, in a context
of gradual decline in terms
of trade over the year
tion – assumed a considerable role, which may
Throughout 2016 the GDP deflator continued
be associated with restaurants and cafés not
to record a year-on-year rate of change above
fully reflecting the decline in VAT on the pric-
that of the HICP and of the domestic demand
es charged. This is reinforced by the dynamics
deflator. This pattern started in the second
of the HICP and the Harmonised Index of Con-
half of 2012. Since Portugal is a net impor-
sumer Prices at Constant Tax Rates (HICP-CT)
ter of energy goods, the recovery of oil prices
in this item. While the former index recorded a
throughout 2016 gradually reduced the positi-
year-on-year rate of change of 1.1 per cent in
ve effect of terms of trade, making it negative
the third quarter, the latter grew by 8.8 per cent
in the fourth quarter (Chart 7.10).
6.0
Chart 7.7 •
Inflation rate
disaggregation
using MIMO
| Contributions,
in percentage points
5.0
4.0
3.0
2.0
1.0
0.0
-1.0
-2.0
-3.0
-4.0
1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4
2008
2009
2010
ULC
Fuels and lubricants
Inflation
2011
2012
PMX
IT & ADM
2013
2014
2015
2016
Food commodities
Others
Sources: Eurostat and Banco de Portugal.
Notes: ULC – Unit Labour Costs; PMX – Import prices excluding energy goods; IT – Indirect taxes; ADM – Administered prices.
Obs: The effect of taxes in fuels and lubricants is included in IT & ADM.
The Portuguese economy in 2016
79
5.0
Chart 7.8 •
Inflation
expectations
and observed
inflation
4.0
3.0
2.0
| In per cent
1.0
0.0
-1.0
-2.0
2008
2009
2010
2011
2012
Current year forecast
2013
2014
2015
2016
Observed inflation (12-month average rate of change)
Sources: Consensus Economics and Statistics Portugal.
5.0
Chart 7.9 •
Inflation
expectations
for current
and next year
4.0
3.0
2.0
| In per cent
1.0
0.0
-1.0
-2.0
2008
2009
2010
2011
2012
2013
Current year forecast
2014
2015
2016
Next year forecast
Source: Consensus Economics.
5,0
4,0
Chart 7.10 •
GDP deflator,
domestic demand
deflator and terms
of trade
| In per cent
3,0
2,0
1,0
0,0
-1,0
-2,0
-3,0
-4,0
2010 Q1
2011 Q1
2012 Q1
Change in terms of trade
Source: Statistics Portugal.
2013 Q1
2014 Q1
GDP deflator
2015 Q1
2016 Q1
Domestic demand deflator
80
BANCO DE PORTUGAL • Economic Bulletin • May 2017
8.Balance of payments
In 2016 the net lending of the economy increased
to 1.5 per cent of GDP, up from 0.3 per cent in
2015. This improvement reflected a sharper reduction in investment than in savings, in a context
of relative stabilisation of net capital transfers. In
sectoral terms (Chart 8.1), households had a slight
decline in net lending, due to the reduction in capital transfers received. Current savings in this sector, as a percentage of GDP, remained at a lower level than investment, and both aggregates declined slightly in 2016. Non-financial corporations,
in turn, moved from net borrowing in 2015 to net
lending in 2016, as a result of the increase in current savings, which more than offset the slight increase in investment.
Increase in the current and capital
account surplus, based on net
exports of goods and services
In 2016 the current and capital account surplus
increased to 1.7 per cent of GDP, up from 1.2 per
cent in 2015 (Table 8.1). This increase in the balance was the result of a reduction in the deficit of
the goods and primary income accounts (both by
0.3 p.p. of GDP) and an increase in the surplus of
the services account (by 0.2 p.p. of GDP), associated with the performance of the tourism sector
(Chart 8.2). In addition, the capital account surplus
declined to 0.9 p.p. of GDP, down from 1.2 p.p. in
2015, strengthening the trend observed in recent
years, which are related to the decline in European fund transfers for investment purposes.
The buoyancy of the tourism
sector continued to contribute
decisively to the behaviour of
the goods and services account
The goods account deficit declined by 0.3 p.p.
for the second year in a row. This decline was a
result of the more pronounced growth of goods’
exports (0.8 per cent) than imports (0.3 per cent).
The services account surplus increased again, to
stand at 7.1 per cent of GDP in 2016. Exports of
services grew by 4.4 per cent, decelerating from
the previous year, when the rate of change had
stood at 7.5 per cent. Growth in 2016 was chiefly
due to the contribution of travel and tourism.
This component continued to post very significant buoyancy, reflecting a 10.7 per cent change
in 2016, up from 10.2 per cent in the previous
year. The main contribution to the deceleration
in exports of other services was made by the
behaviour of services provided by corporations,
which had a rate of change of -3.3 per cent, compared with 15.4 per cent growth in 2015.
Table 8.1 • Current and capital accounts | As a percentage of GDP
Current and capital accounts
Current account
Goods and services account
Goods
Services
2013
2014
2015
2016
3.2
1.5
1.2
1.7
1.6
0.1
0.1
0.8
1.9
1.1
1.8
2.2
-4.7
-5.5
-5.2
-4.9
6.6
6.6
6.9
7.1
of which:
Travel and tourism
3.6
4.1
4.4
4.8
Primary income account
-1.2
-2.0
-2.5
-2.2
0.9
0.9
0.8
0.8
Secondary income account
of which:
1.4
1.5
1.6
1.5
Capital account
Emigrants / immigrants remittances
1.6
1.4
1.2
0.9
Financial Account
3.3
1.6
1.2
1.7
Errors and omissions
0.1
0.1
0.0
0.0
Sources: INE and Banco de Portugal.
The Portuguese economy in 2016
Imports of services had similar developments,
with a change of 3.1 per cent, down from 5.8 per
cent in 2015. The main contributions to growth in
2016 were travel and tourism, intellectual property rights not included elsewhere and other services.
81
Income paid abroad declined
and the subsidies received from
the European Union increased
The primary income account deficit decreased by
0.3 p.p. of GDP as a result of the 0.4 p.p. decline
The trade account balance in 2016 benefitted
in income paid abroad and the maintenance of
from a positive volume effect, contrary to devel-
income received from non-resident entities. The
opments in the two previous years (Chart 8.3).
decline in income paid abroad was mainly asso-
The largest contribution, however, was made by
ciated with the decrease in the stock of exter-
the terms of trade effect, albeit less expressive
nal liabilities by resident sectors, in particular the
than in the previous year. This was associated
general government and other monetary finan-
with the development of energy prices (Section 2:
cial institutions (other MFI). The stabilisation of
‘International environment’ and Section 7: ‘Prices’).
primary income received from abroad was the
10
Chart 8.1 •
Net lending/net
borrowing of the
economy
5
0
-5
| As a percentage
of GDP
-10
-15
-20
2009
2010
2011
2012
2013
2014
Non-financial corporations
Financial corporations
General Government
Households
2015
2016
Net lending/net borrowing
Source: INE.
10
8
Chart 8.2 •
Decomposition
of current and
capital account
balance
6
4
2
0
-2
| As a percentage
of GDP
-4
-6
-8
-10
2013
2014
Capital account
Services
Goods
Sources: INE and Banco de Portugal.
2015
Secondary income account
Primary income account
Current and capital accounts
2016
82
BANCO DE PORTUGAL • Economic Bulletin • May 2017
result of mixed developments. On the one hand,
i.e., the purchase of external financial assets
there was a decline in interest received associat-
by residents was higher than the purchase of
ed with long-term debt securities held in portfo-
national assets by non-residents (Chart 8.4).
lio by Banco de Portugal, insurance corporations
and pension funds and, to a lesser extent, households. The reduction by insurance corporations
and pension funds and households was associated with a decline in the holdings of long-term
debt securities issued by non-residents, whereas
in the case of Banco de Portugal it was due to a
decline in yields on securities. On the other hand,
there was an increase in the subsidies from the
European Union.
The secondary income account components
stabilised as a percentage of GDP. Emigrants’
remittances decelerated, which was reflected in
a 0.8 per cent increase, down from 8.3 per cent
in 2015. Behaviour differed across home countries. Remittances from Switzerland, Angola and
Germany declined, but there was an increase
in transfers from France, the United States, the
United Kingdom and, to a lesser extent, Spain
and Luxembourg.
Net lending of the Portuguese
economy was reflected in a
fall in the negative balance of
the international investment
position and net external debt
These net outflows corresponded to 1.7 per cent
of GDP and contributed to the improvement in
the International Investment Position (IIP) of the
Portuguese economy. At the end of 2016, the
IIP stood at 105.1 per cent of GDP, which corresponded to an improvement of 6.9 p.p. of GDP
(Table 8.3). In addition to the effect of financial
account transactions, the total change in the IIP,
as a percentage of GDP, may be broken down
into changes in prices of the financial instruments, exchange rate changes in instruments
denominated in foreign currency, other changes, such as reclassifications, and also the denominator effect, i.e. the GDP nominal change. Therefore, in addition to the financial account effect,
the balance of the IIP as a percentage of GDP
was positively influenced by the effect of the
nominal GDP change, estimated at 3.3 p.p., and
price changes, which correspond to 2.7 p.p. of
GDP (Chart 8.5). In this latter item, both the
gains in financial assets held by residents and
the losses in national liabilities to non-residents contributed to a positive effect on the IIP
balance. In particular, rises in sovereign debt
yields are reflected in a decline in security prices with a positive impact on the IIP. Conversely, exchange-rate changes and other chang-
The current and capital account surplus recor-
es made a negative contribution, equivalent
ded in 2016 implied a financial account surplus,
to 0.7 p.p. of GDP.
Table 8.2 • Financial Account | As a percentage of GDP
2013
Financial Account
Direct investment
2014
2015
2016
3,3
1,6
1,2
1,7
-1,7
-1,5
-0,6
-2,2
Net acquisition of assets
3,0
4,2
0,5
1,9
Net incurrence in liabilities
4,7
5,7
1,1
4,1
8,0
Portfolio investment
2,6
-0,7
0,2
Net acquisition of assets
-0,1
4,2
0,6
0,9
Net incurrence in liabilities
-2,7
4,9
0,3
-7,1
0,6
1,1
0,2
0,3
Financial derivatives
Other investment
1,6
1,7
0,6
-6,9
Net acquisition of assets
-6,2
2,3
-5,4
-2,0
Net incurrence in liabilities
-7,8
0,5
-5,9
4,9
0,3
1,0
0,8
2,5
Reserve assets
Sources: INE and Banco de Portugal.
83
The Portuguese economy in 2016
Banco de Portugal and nonfinancial corporations purchased
foreign assets, whereas nonresidents disinvested from public
debt and other-MFI’s securities.
In 2016 Banco de Portugal was the main agent
purchase of debt securities, either as reserve
assets, or, in particular, as portfolio investment,
and deposits with non-resident entities. This initiative by Banco de Portugal corresponds to the
participation in the implementation of Eurosystem’s monetary policy decisions. Notwithstanding these transactions, the IIP of Banco de Portugal remained virtually unchanged, also due to a
investing in external assets, especially via the
significant increase in liabilities to non-residents,
4,000
Chart 8.3 •
Decomposition
of goods and
services account
3,000
2,000
1,000
| Million of euros
0
-1,000
-2,000
-3,000
2013
Volume effect
2014
Price effect
2015
Terms of trade effect
2016
Crossed effect
Change in balance
Sources: INE and Banco de Portugal.
Note: A positive change (negative) implies an increase (decrease) in the overall balance of the goods and services account. The change
in the overall balance of goods and services account can be decomposed in four effects: – volume effect – effect of the change in
quantities imported and exported; [Xt-1.vxt]-[Mt-1.vmt] – price effect – effect of the average price growth of external trade; [Xt-1.
pt]-[Mt-1.pt] – terms of trade effect – effect of the relative change in exports and imports prices; [Xt-1.(pxt--pt)]-[Mt-1.( pmt-pt)] –
crossed effect – effect of the interaction between the change in quantities and in prices of exports and imports; [Xt-1.vxt.pxt]-[Mt-1.
vmt.pmt] The following notation applies: Xt-1 and Mt-1 are the exports and imports in year t-1 at current prices; vxt and vmt are the
change rates in volume of exports and imports in t; pxt and pmt are the change rates of exports and imports prices in t; pt is the
average change rate of the prices of external trade in year t ((pxt+pmt)/2).
15.0
Chart 8.4 •
Net change in
financial assets and
liabilities and overall
financial balance
10.0
5.0
| As a percentage
0.0
of GDP
-5.0
-10.0
2013
2014
Overall balance
Sources: INE and Banco de Portugal.
2015
Net acquisition of assets
2016
Net incurrence of liabilities
84
BANCO DE PORTUGAL • Economic Bulletin • May 2017
particularly in the form of liabilities to the Euro-
increase in liabilities than in assets and in the de-
system. Monetary gold valuation also contribut-
terioration of the IIP balance of this sector.
ed to the increase in financial assets by 0.9 p.p.
Contrary to the increase in external assets by
of GDP (Charts 8.6 and 8.7).
Banco de Portugal and non-financial corpora-
Non-financial corporations were also engaged
tions, non-residents disinvested significantly from
in investments abroad, albeit to a lesser extent.
national public debt securities (Section 4: ‘Fiscal
These investments occurred mainly with corpo-
policy and situation’) and, to a lesser extent, from
rations with which they had established direct
securities issued by other MFI.
investment links and from which they had ob-
In the case of the general government, and in
tained funds (Box 8.1 ‘Direct investment flows
addition to transactions in debt securities issued,
into the Portuguese economy’). The balance of
reference should be made to the repayment of
these cross transactions was reflected in a higher
loans obtained under the European and Financial
112
Chart 8.5 •
Change in
International
investment
position
| As a percentage
of GDP
1.7
110
2.7
108
106
3.3
112.0
0.7
104
105.1
102
100
IIP 2015 (symmetrical)
Transactions
Changes in prices
Change in GDP
Exchange rate and IIP 2016 (symmetrical)
other adjustments
Sources: INE and Banco de Portugal.
Chart 8.6 • Net acquisiton of assets (1)
| As a percentage of GDP
Chart 8.7 • Net incurrence of liabilities(2)
| As a percentage of GDP
8
8
6
6
4
4
2
2
0
0
-2
-2
-4
-4
-6
-6
-8
Total
CB
OMFI
2015
GG
NMFI
NFPS
2016
-8
Total
CB
OMFI
2015
GG
NMFI
NFPS
2016
Sources: INE and Banco de Portugal.
Sources: INE and Banco de Portugal.
Note: (1) The net acquisiton of assets corresponds to buys less sells of foreign
assets by residents. A plus sign represents a net outflow of funds from the portuguese economy.
Note: (2) The net incurrance of liabilities corresponds to the increase less redemptions of national liabilities with non-resident entities. A plus sign corresponds to a
net inflow of funds in the portuguese economy.
The Portuguese economy in 2016
Assistance Programme. These transactions were an
also declined. The volume of these transactions
important contribution to the decline of 12.5 p.p.
was more significant than the decline in liabili-
of GDP in the stock of external liabilities in this
ties, leading to a deterioration of the negative
sector. This result also had the important contri-
balance of the IIP in this sector by 1 p.p. of GDP.
bution of both GDP and the devaluation of Treas-
Taking into account the significant transactions
ury bonds, in a context of an increase in public debt yields.
85
in debt securities issued by residents, net external debt, which is obtained by excluding capital
As regards the liabilities of other MFI, in addition
instruments and financial derivatives from the
to the decline in liabilities in the form of debt
IIP, declined by 7 p.p. of GDP, down from 101.5
securities, there was a decrease in deposits by
per cent of GDP in 2015 to 94.4 per cent of GDP
non-residents. In the same vein, investments
in 2016 (Chart 8.8).
abroad in these type of securities by other-MFI
Table 8.3 • International Investment Position | As a percentage of GDP
International Investment Position
Direct investment
Direct investment - Assets
Direct investment - Liabilities
2013
2014
2015
2016
-116.3
-117.7
-112.0
-105.1
-27.3
-31.1
-30.8
-32.0
42.1
42.7
42.6
43.5
69.4
73.9
73.4
75.5
-10.2
-10.6
-9.2
0.8
Portfolio investment - Assets
64.8
68.8
66.7
65.9
Portfolio investment - Liabilities
75.0
79.3
76.0
65.1
-84.5
-84.3
-82.0
-86.6
68.2
69.1
61.9
58.1
Other investment - Liabilities
152.7
153.4
143.9
144.7
Reserve assets and derivatives
5.7
8.3
10.0
12.8
Portfolio investment
Other investment
Other investment - Assets
Sources: INE and Banco de Portugal.
0
Chart 8.8 •
International
investment position,
net external debt
and position in
capital instruments
| As a percentage
-20
-40
-60
-80
of GDP
-100
-120
-140
2005
2006
2007
2008
Net External Debt (symmetrical)
Sources: INE and Banco de Portugal.
2009
2010
2011
Net Capital Instruments
2012
2013
2014
2015
2016
International Investment Position
86
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Box 8.1 | Direct investment flows into the Portuguese economy41
The current situation of the Portuguese economy is characterised by moderate economic growth,
relatively weak investment rate and unfavourable developments in labour productivity, against a
background of high indebtedness of the different institutional sectors. In this context, the acceleration of economic activity and the increase in productivity require an investment effort that will
be easier to implement if there is the possibility to attract external capital.
Foreign direct investment in Portugal reflects the Portuguese economy capacity to attract capital,
also resulting in positive externalities, especially in terms of sharing technology and business knowledge. Therefore, it is important to analyse in more detail the nature and composition of these flows.
Balance of payments statistics on foreign direct investment in Portugal include financial transactions
on assets and liabilities of resident corporations with their non-resident direct investors, as well as all
financial transactions between resident and non-resident corporations, held (directly or indirectly) by
the same non-resident entity.42 Diversified financial flows are thus recorded, comprising reinvested
earnings, investment in real estate, debt instruments and other capital instruments.
Reinvested earnings correspond to corporations’ results that are retained, and reflect their allocation to foreign holders of the respective capital. Their registration in the balance of payments is
therefore made under investment income and in the financial account, corresponding to an increase
in investment. Reinvested earnings are recorded in the period in which they are earned.
Investment in real estate corresponds to investment by non-residents in property and residences
in the national territory, either for personal use or for rental.
Debt instruments cover funds lent by non-resident entities – in the form of loans or debt securities – whereas capital instruments represent the property rights on the resident entity.
In an open economy as the Portuguese, foreign direct investment is an important source of financial inflows. Compared with other European countries, Portugal is among the main recipients of
direct investment as a percentage of GDP (Charts 1 and 2).
For many years, inflows into the Portuguese economy occurred mainly in the form of portfolio
investment and other investment. In the 1996-2016 period, foreign direct investment, on average, accounted for annual net inflows into the economy of 2.4 per cent of GDP. In most recent
years, however, and in a context of declining external indebtedness, direct investment flows have
been gaining relative importance in the Portuguese economy. In 2013 and 2015 they were in fact
the main source of external financing (Chart 3).
Since 2008, this investment represented, on average, 2.1 per cent of GDP. It was higher in 2011
and 2012, when it reached 4.1 per cent of GDP, and in the last two years, when it stood, on average, at 3.2 per cent of GDP.
Financing requirements of the Portuguese economy were partially met with recourse to privatisations, which translated into the purchase of assets by non-residents. Therefore, in 2011 and 2012,
the composition of foreign direct investment in Portugal primarily took the form of capital instruments,
and was associated with the non-financial corporations sector (Charts 4 and 5).
After a decline in inflows in 2013 and 2014, financial transactions intensified again in the last two
years. In terms of instrument, however, there was a decline in the amounts of investment in corporations’ capital, against an increase in transactions in debt instruments. I.e., resident corporations
that received foreign direct investment used that channel to obtain financing from the firms in the
The Portuguese economy in 2016
Chart 1 • Foreign Direct Investment – transactions – 2015 | As a percentage of GDP
5
4
3
2
1
0
-1
SK
CR
EE
EL
FI
CZ
DE
DK
ES
IT
UK
SE
AT
FR
LT
RO
LV
PO
PT
IS
BG
SL
BE
Source: Eurostat.
Chart 2 • Direct investment – stocks – 2015 | As a percentage of GDP
120
BE
Inward flows,
as a percentage of GDP
100
EE
BG
80
60
SK
40
RO
LV
IS
PT
CZ
CR
UK
ES
PO
LT
FI
SL
20
DE
IT
AT
SE
DK
NO
FR
EL
0
0
20
40
60
Outward flows, as a percentage of GDP
80
100
120
Source: Eurostat.
Chart 3 • Foreign direct investment and other inward funds in portuguese economy
– transactions | As a percentage of GDP
30
25
20
15
10
5
0
-5
-10
-15
-20
-25
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016
FDI
Sources: INE and Banco de Portugal.
Portfolio investment
Other investment
87
88
BANCO DE PORTUGAL • Economic Bulletin • May 2017
group, thus circumventing more restrictive financing conditions in Portugal. This may be related
either to loans or to issues of debt securities taken by firms in the group.
In the last three years, investment in real estate also reinforced its share as component of direct
investment.43 In 2016, flows associated with these transactions exceeded flows of investment in
other capital instruments (0.7 per cent versus 0.4 per cent of GDP). These developments reflected demand for land and property by non-residents, in line with the profile of residence permits
for investment purposes.
Chart 4 • Foreign direct investment in Portugal, by sector – transactions
| As a percentage of GDP
6
4
2
0
-2
2008
2009
2010
2011
2012
Non-financial corporations
Non-monetary financial instituitions except ICPF
General Government and Households
2013
2014
2015
2016
Other Monetary Financial Institutions
Insurance Corporations and Pension Funds (ICPF)
Sources: INE and Banco de Portugal.
Chart 5 • Foreign direct investment in Portugal, by type of transaction – transactions
| As a percentage of GDP
5
4
3
2
1
0
-1
-2
-3
2008
2009
2010
Capital
Sources: INE and Banco de Portugal.
2011
Reinvested earnings
2012
2013
Real estate
2014
2015
Debt instruments
2016
The Portuguese economy in 2016
Notes
1. Available at https://www.bportugal.pt/sites/default/files/anexos/pdf-boletim/projecoes_mar_pt_0.pdf
2. For further details, see Félix et al. (2007) ‘MIMO – A Monthly Inflation Model’, winter 2007 issue of the Economic Bulletin of Banco de Portugal.
3.Angola is not included in the set of countries used under the Eurosystem’s procedure to calculate external demand for the goods and services of each
individual country.
4.The APP is composed of four programmes: the Covered Bond Purchase Programme (CBPP), which is currently in its third programme; the Asset-backed
Securities Purchase Programme (ABSPP), the Public Sector Purchase Programme (PSPP), and the CSPP. The first two programmes started in 2014 and
the third programme in 2015.
5.In its December 2016 meeting, the Governing Council of the ECB announced its decision to extend the APP until the end of 2017. However, purchases
were reduced from €80 billion to €60 billion from March onwards.
6.The Eurosystem’s architecture only allows monetary financial institutions to have access to the ECB’s deposit facility rate.
7.In this respect, see the speech by Yves Merch, ‘Ructions in the repo market – monetary easing or regulatory squeezing?‛, 26 January 2017.
8.This fact tends to be visible in the repo market, particularly on dates close to accounting reporting dates and for rates associated with jurisdictions
perceived as safe. On the last business day of December 2016, a number of repo rates reached historical lows.
9.See Antunes, Homero and Prego (2016) ‘Firm default probabilities revisited‛, Banco de Portugal Economic Studies, Vol. 2, No. 2, Banco de Portugal.
10. Total credit to non-financial corporations includes loans by resident banks, loans by non-residents, debt securities issuance (held by residents and
non-residents), trade credits (granted by residents and non-residents), household loans and Treasury loans (in the case of loans to public undertakings).
11. This definition of indebtedness includes the sum of bank loans, debt securities and trade credits.
12. See Box “Developments in corporate indebtedness in Portugal and the euro area”, Economic Bulletin, May 2016, Banco de Portugal. The analysis considers
data from the national financial accounts and the measure of indebtedness considered is the ratio of debt (loans and debt securities) to total assets.
13. Félix, S. (2017) “Firm creation and survival in Portugal”, Banco de Portugal Economic Studies, Banco de Portugal.
14. The survey IES includes economic, financial, and accounting information of virtually all the Portuguese non-financial corporations and is available for
the period between 2005 and 2015. Nevertheless, changes in the accounting standards system in 2010 have led to a loss of data comparability.
15. A reporting gap is observed when the firm reports to the IES in year t-1 and year t+1, but is absent in year t.
16. Some of the absences that are considered exits from the market in 2015 may constitute reporting gaps. This can only be verified when the IES for 2016
becomes available. However, only about 0.2 per cent of the observations are reporting gaps in the period under analysis.
17. EBIT stands for Earnings Before Interest and Taxes.
18. These factors meet the criteria for classification as temporary measures, as defined in the Eurosystem context, which does not necessarily coincide
with the classification adopted by other institutions.
19. In August 2016, the EU Council recommended a stabilisation of the structural balance. Compliance with this recommendation shall be assessed on the
basis of the cyclical adjustment methodology and the definition of temporary measures adopted by the European Commission.
20. The impact of the resolution of BANIF in 2015 is treated as a temporary measure, according to the definition adopted by the Eurosystem. This operation had a significant impact on capital transfers, chiefly as a result of the capital increase by the State. To a lesser extent, it also affected the public
investment aggregate via the recording of real estate acquisitions by Oitante, a special-purpose entity created in the context of the resolution process.
21. Military material delivered in 2016 relates to the sale of F-16 aircrafts to Romania, which negatively affects the general government investment aggregate. In spite of their one-off impact on the balance, these operations are not treated as temporary measures in the context of the definition adopted
by the Eurosystem.
22. In 2015 a number of operations contributed to an increase in general government capital expenditure: the conversion into an equity increase of loans
granted by Wolfpart to its holding Caixa Imobiliário; the recording of requests for budget increases by Instituto de Financiamento da Agricultura e Pescas
(Agricultural and Fisheries Financing Institute); capital increases in public corporations that are not included in the general government.
23. In 2015 the general government wage cuts in force since 2011 started to be reversed, with 20 per cent of the cuts being reinstated. In 2016 the remainder
was gradually reinstated in the course of the year, by 25 per cent each quarter.
24. In early 2015, the European Central Bank announced the widening of the asset purchase programmes then in force, to also include the purchase of
bonds issued by euro area central governments, as well as by European organisations and institutions.
25. The implicit interest rate is computed as the ratio of interest expenditure to the simple average of the debt stock at the end of the current and the
previous year.
26. The analysis in this box is based on the methodology developed in the context of the Eurosystem which allows the breakdown of structural revenue from
taxes and social contributions into the following components: i) impact of legislative changes; ii) contribution of the discrepancy between the trend nominal
change in the relevant macroeconomic base and GDP; iii) effect of the structural fiscal elasticity in structural terms and iv) residual component. The residual
component corresponds to the part of the structural change that is not explained by the previous effects. For further details, see Kremer et al. (2006), A disaggregated framework for the analysis of structural developments in public finances, Working paper ECB 579, and Braz, C. (2006), ‘The calculation of cyclically
adjusted balances at Banco de Portugal: an update’, winter 2006 issue of the Economic Bulletin of Banco de Portugal.
89
90
BANCO DE PORTUGAL • Economic Bulletin • May 2017
27. In the case of the tax on oil products, the estimate for the impact of an increase in taxation reflects an ex post reassessment based on the final outturn
for the year.
28. Total revenue from social contributions includes actual and imputed social contributions referring to civil servants, which are also recorded on the
expenditure side, under compensation of employees.
29. See Box 5.2 of the October 2016 issue of the Economic Bulletin for a characterisation of very long-term unemployment in Portugal.
30. According to Eurostat's definition, tradable services include accommodation and food services, trade and repair, transportation and storage, and information and communication activities.
31. Apparent labour productivity was calculated dividing VAT (volume chain-linked series, € millions) by total employment (full-time equivalent), except
in the 1992 recession, when it was calculated on the basis of GDP.
32. For a discussion, see for example F. Caselli (2005) ‘Accounting for Cross-Country Income Differences‛, Handbook of Economic Growth, volume 1, Chapter 9.
33. Information available at https://fred.stlouisfed.org/release?rid=285.
34. Taking into account the weight of trade with Angola in Portuguese nominal exports, it is possible to estimate the mechanical impact of the fall in exports to
Angola on nominal GDP. Hence, in 2016 total exports to Angola (goods and services), excluding an estimate for import contents, appear to have made a contribution of around -0.6 p.p. to 3.0 per cent growth of nominal GDP (in 2015 the contribution was around -0.7 per cent for 3.7 per cent growth of nominal GDP).
35. In 2016, in particular, remuneration received by households grew by 3.6 per cent, after 2.4 per cent in 2015.
36. In 2016 the decline in interest expenses of households contributed 0.4 p.p. to the 3.2 per cent growth of nominal disposable income.
37. See the special issue ‘Recent developments in consumer lending: A macroprudential approach‛, Financial Stability Report, Banco de Portugal, November 2016.
38. Following 1.5 per cent year-on-year falls in the first two quarters of 2016, nominal exports of goods grew by 1.9 and 5.1 per cent respectively in the third
and fourth quarters of 2016. This was also due to the stabilisation of exports of fuels in the second half of the year, after the break observed in the first half.
39. Information on the aggregation and calculation of rates of change in European export prices in Banco de Portugal, 2016, ‘Portuguese international
traders: some facts about age, prices and markets‛, Economic Bulletin, October 2016, special issue.
40. For further details, see Félix et al. (2007) ‘MIMO – A Monthly Inflation Model ‛, Winter 2007 issue of the Economic Bulletin of Banco de Portugal.
41. This Box was based on information on Foreign Direct Investment in Portugal, according to the directional principle. I.e., all transactions with non-residents regarding resident direct investment firms are considered, irrespective of whether they are an asset or a liability of such corporations.
42. For further details on the statistical treatment of information on direct investment, see Banco de Portugal (2015), ‘Statistics of the balance of payments
and the international investment position – Methodological notes’, Supplement to the Statistical Bulletin No. 2/2015.
43. For a more detailed analysis of real-estate investment, see Box 7.1 of the May 2016 issue of the Economic Bulletin.
II
Special issue
Distribution mechanisms of monetary
policy in the Portuguese economy
Special issue
Distribution mechanisms of monetary policy
in the Portuguese economy
1. Introduction
models with heterogeneous agents, together
In the wake of the global economic and financial
with the dissemination of microeconomic data
crisis, the link between the implementation of
monetary policy and developments in inequality has re-entered the public debate. This is set
against a background where inequality has once
again been considered inseparable from economic policy. The perception that the reduction
in inequality seen throughout most of the last
century had been reversed, together with the
fact that advanced economies have failed to
return to the activity levels that would have been
reached if the trend before the crisis had continued, contributed to the greater importance
of issues related to the distribution of economic
resources.
It is well known that inequality is not, and should
not be, an objective pursued by central banks.
In the case of the euro area, the primary objective of the European Central Bank (ECB) is price
stability and, secondarily, support to the general
economic policies in the European Union. The
maintenance of price stability helps prevent the
arbitrary redistribution of income. In any case,
it is potentially important to understand the
link between the implementation of monetary
policy and inequality for two sets of reasons. On
the one hand, inequality levels prevailing in an
economy may affect monetary policy transmission to economic agents, thereby hampering its
expected results. On the other hand, monetary
on an individual basis, has helped focus academic
discussion on distribution issues. There is a new
consensus in economic literature that monetary
authorities must be aware of the redistribution
effects of monetary policy, as well as the unobserved heterogeneity in the economy. This conclusion stands even when the redistribution
impact of monetary policy is relatively small over
long periods, which are the most suitable when
analysing structural developments in inequality: over such time horizons, structural factors
become more important, including inter alia the
openness of economies, technological progress,
institutional changes in the output and labour
markets, the tax regime and the social protection
network.
The purpose of this article is to look into some
of the main monetary policy transmission channels impacting on the distribution of monetary
resources among households. Such channels
are illustrated on the basis of microeconomic
data on the Portuguese economy. However,
the analysis will necessarily compare modestly
with the title of this article. In fact, the impact
of monetary policy on the Portuguese economy
goes well beyond the channels analysed in this
article. For instance, the reduction in the ECB’s
intervention rates following the international
financial crisis made a key contribution to relieving the debt service of Portuguese households
policy may affect inequality levels prevailing in
and to reduce liquidity constraints. This channel
the euro area and individual economies and,
had a major impact on the Portuguese econo-
as such, influence a considerable part of public
my as a whole, but it will not be reviewed here.
debate and the Eurosystem’s accountability.
Studying the distribution effects of monetary
The redistribution impact of monetary policy
policy would involve describing a counterfac-
has also been increasingly reviewed in econom-
tual scenario that can only be created using a
ic literature (see e.g. Auclert, 2016, or Coibion et
proper model.
al., 2016, and respective references). The use of
The Special Issue is structured as follows. Section
increased computing power to solve economic
2 looks into the methodology and the microdata
93
94
BANCO DE PORTUGAL • Economic Bulletin • May 2017
bases used in the analysis. Section 3 describes
between its assets and its liabilities. This net
some of the main monetary policy redistribution
wealth, by construction, is predetermined and
channels, illustrating their potential impact based
generates some return, including capital gains,
on microeconomic data on the Portuguese econ-
over time. On the other hand, there is the so-
omy. Section 4 includes an exercise that analyses
called human wealth, i.e. the sum of current
the quantitative impact on the Portuguese econ-
and future income resulting from household
omy of the standard and non-standard mone-
decisions as to how to spend time in the labour
tary policy measures implemented over the past
market.
few years. Section 5 presents and discusses the
Over the lifetime of each household, human
main conclusions of this exercise.
wealth will depend on exogenous conditions
– such as the labour market environment and
2. Methodology and databases
structural changes – and idiosyncratic features
In this article, the household is the unit of analy-
of the household under review. Among house-
sis. By definition, this is the unit where inequal-
hold-specific features are the (accumulated or
ity resulting from its links with the market is of
inherent) efficiency with which household mem-
greater interest. Although in Portugal it is still
bers that participate in the labour market are
not possible to track wealth trends for a repre-
able to turn their time into income, and the start-
sentative sample of the population over time,
ing position in the labour market, which, togeth-
measures implemented over the past few years. Section 5 presents and discusses
main
conclusions
er withthe
age,
partially
determines future returns
results are now available from two rounds of the
exercise.
on that market. The transformation of initial
er the past few years. Section
5 presentsFinance
and discusses
the main conclusions
Household
and Consumption
Survey,
a Eurosystem initiative that Banco de Portugal
conditions into returns in the labour market is
and Statistics Portugal (Portuguese acronym:
anything but deterministic. As is well known, luck
INE) have implemented domestically. The inforethodology and databases
plays its part, which means that idiosyncratic
mation gathered from these surveys will make
shocks help determine, inter alia, which house-
ses
members
will resulting
be employed, the quality of
it possible
to start tracking
the impact
article, the household is the
unit of analysis.
By definition,
this isunder
the unithold
where
inequality
thepossible
match between
employees and workers
BancoAlthough
de Portugal
and Statistics
ts links with the market is ofreview
greater(see
interest.
in Portugal
it is still not
to track wealth
the unit of analysis. By definition, this is the unit where inequality resulting
(according
to two
theirrounds
skills) and
2016, Costa
andtime,
Farinha,
2012,
s for a representative samplePortugal,
of the population
over
results
are and
now available
from
of developments in
of greater interest. Although in Portugal it is still not possible to track wealth
thede
sector
of activity
where a household memCosta, 2016).
ousehold Finance and Consumption
Survey, a Eurosystem initiative that Banco
Portugal
and Statistics
ple of the population over time, results are now available from two rounds of
ber will
work. Labour
gal (Portuguese acronym: INE)
have
implemented
domestically.
The information
gathered
from income
these results from the
In this
article,
household
inequality
assessed
sumption Survey, a Eurosystem
initiative
that
Banco de
Portugalis and
Statistics
combination
of
these
opportunities.
This clearys will make it possible to start
the
impact
underofreview
(see
Banco de Portugal and Statistics
on tracking
the basis
of information
two
concepts
wealth:
wealth
NE) have implemented domestically.
The
gathered
from
these
ly shows that, in contrast to net wealth, human
gal, 2016, Costa and Farinha,in
2012, and
Costa, 2016).
sense
net wealth.
tart tracking the impact underbroad
review
(seeand
Banco
de Portugal and Statistics
wealth depends on a great deal of unavailable
sa, article,
household
inequality
is assessed
on the
basis of
concepts
of wealth: wealth in broad
2012, and
Costa, 2016).
Wealth
in broad sense
assesses,
at two
a given
moinformation. At best, information is only avail-
and net wealth.
mentofintwo
time,concepts
the rangeof
of wealth:
decisionswealth
that houseality is assessed on the basis
in broadable on probability distributions.
holds
may
make
in
the
course
of
their
h in broad sense assesses, at a given moment in time, the range oflifetime
decisions that households may make
At a given moment in time, the sum of the expect-
and,
therefore,provides
provides aagood
of the
course of their lifetime and,
therefore,
goodsummary
summary
of the resulting well-being. This
at a given moment in time, the range of decisions that households may makeed value of labour income in the course of a liferesulting
well-being.
This
aggregate
may
be
brogate may be broken down into two very different components, as illustrated by the following equation:
nd, therefore, provides a good summary of the resulting well-being. Thistime, discounted back to that moment, deterken down into two very different components,
n into two very different components, as illustrated by the following equation:mines the household’s human wealth. This aggreas illustrated by the following equation:
gate includes pension income, given its close link
to labour market participation.
ܹ݈݁ܽ‫ ݁ݏ݊݁ݏ݀ܽ݋ݎܾ݄݊݅ݐ‬ൌ ܰ݁‫݄ݐ݈ܽ݁ݓݐ‬
൅ ‫݄ݐ݈ܽ݁ݓ݊ܽ݉ݑܪ‬
(1) In practical terms,
(1)
ܹ݈݁ܽ‫ ݁ݏ݊݁ݏ݀ܽ݋ݎܾ݄݊݅ݐ‬ൌ ܰ݁‫ ݄ݐ݈ܽ݁ݓݐ‬൅ ‫݄ݐ݈ܽ݁ݓ݊ܽ݉ݑܪ‬
each household was assumed to have a wide hori-
(1)
zon when making its decisions, as would be the
On the one hand, the net wealth (or non-human
case when using an infinite-horizon agent mod-
wealth) of each household when the period
el. Assuming that labour income corresponds to
e one hand, the net wealth (or non-human wealth) of each household when the period under review
s results from the difference between its assets and its liabilities. This net wealth, by construction, is
(or non-human wealth) of each
when
the from
period
reviewa perpetuity, and taking into account a discount
underhousehold
review begins
results
theunder
difference
termined and generates some return, including capital gains, over time. On the other hand, there is
ce between its assets and its liabilities. This net wealth, by construction, is
o-called human wealth, i.e. the sum of current and future income resulting from household
ome return, including capital gains, over time. On the other hand, there is
ons as to how to spend time in the labour market.
.e. the sum of current and future income resulting from household
he in
lifetime
of each
household, human wealth will depend on exogenous conditions – such as the labour
me
the labour
market.
et environment and structural changes – and idiosyncratic features of the household under review.
old, human wealth will depend on exogenous conditions – such as the labour
Special issue
rate of 4%, this would imply that human wealth
for wealth in broad sense and net wealth com-
may be proxied by multiplying current labour
prise different households. Indeed, there is a
income by 25. This proxy is used in the exercises
relatively small overlap between households
below, although results are qualitatively robust to
across distribution quintiles, as illustrated in
the use of different discount rates. Equation (1)
Table 2. For instance, only 48 per cent of house-
can therefore be re-written as follows:
holds in the lowest quintile of wealth in broad
1
95
sense are also included in the lowest net wealth
ܹ݈݁ܽ‫ ݁ݏ݊݁ݏ݀ܽ݋ݎܾ݄݊݅ݐ‬ൌ ܰ݁‫ ݄ݐ݈ܽ݁ݓݐ‬൅ ‫݄ݐ݈ܽ݁ݓ݊ܽ݉ݑܪ‬
ൎ ܰ݁‫݄ݐ݈ܽ݁ݓݐ‬
൅ ʹͷ ൈ ‫݁݉݋ܿ݊݅ݎݑ݋ܾܽܮ‬
quintile. This
different household
composition
(2)
in both measures of wealth distribution plays a
(2)
key role when interpreting the results obtained
‫ ݁ݏ݊݁ݏ݀ܽ݋ݎܾ݄݊݅ݐ‬ൌ ܰ݁‫ ݄ݐ݈ܽ݁ݓݐ‬൅ ‫ ݄ݐ݈ܽ݁ݓ݊ܽ݉ݑܪ‬ൎ ܰ݁‫ ݄ݐ݈ܽ݁ݓݐ‬൅ ʹͷ
ൈ ‫݁݉݋ܿ݊݅ݎݑ݋ܾܽܮ‬
in Section
3.4
(2)
Knowing
the distribution
sum of these two types of wealth among households
representing a given
൅ ‫݄ݐ݈ܽ݁ݓ݊ܽ݉ݑܪ‬
ൎ ܰ݁‫݄ݐ݈ܽ݁ݓݐ‬
൅ ʹͷof
ൈ the
‫݁݉݋ܿ݊݅ݎݑ݋ܾܽܮ‬
Table 1 illustrates the high inequality levels in net
(2)
population provides a good measure of inequality among that population. The way in which this distribution
wealth and wealth in broad sense in Portugal.
distribution
the sum–ofincluding
these
reacts to Knowing
structuralthe
changes
in theofeconomy
a persistent policy change – provides a measure
In particular,
the average
istribution ofofthe
sum
of
these
two
types
of
wealth
among
households
representing
a givenvalue of net wealth in
two
types
of
wealth
among
households
repthe distribution impact of that change.
first distribution
quintile is close to zero. This
ovides
a good
measure
of inequality
among
that population.
The waythe
in which
this distribution
wo types
of wealth
among
households
representing
a given
resenting
a given
population
provides
a good
To simplify the analysis, in this article households are broken
down
into
wealth
in broad sense and net wealth
explains
the
high
inequality
tural
changes
the economy
–ofincluding
aamong
persistent
policy change – provides a measurein net wealth as measity among
thatinpopulation.
The
way
in which
this distribution
measure
inequality
that population.
2
quintiles, assuming that the average of each quintile summarises
theQ5/Q1
corresponding
population
group.
ured by the
ratio. Turning
to wealth
in The
tion impact
of that change.
ncluding
a persistent
policy
provides
a measure
The
way change
in which– this
distribution
reacts to
ratios used to gauge inequality are the ratio of the average population
to its
thedispersion
intermediate
quintile
(Average/Q3),
broad sense,
is much
lower,
given
structural
changes
in the economy
– including
a broad sense and net wealth
analysis, in this
article
households
are
broken
down
into
wealth
in
the ratio of the highest quintile to the intermediate quintile (Q5/Q3) and the ratio of the highest quintile to
that human wealth’s dispersion is relatively low.
persistent
policy
change
– and
provides
a
measure
2
ming
thethe
average
of
each quintile
summarises
the
corresponding
population
group.
The
s arethat
broken
down
into
wealth
in
broad
sense
net
lowest
quintile
(Q5/Q1).
The higher
thewealth
ratios, the more
unequal
the
distribution
is. These
Overall,
the heterogeneity
prevailing
in theratios
distri-should
of
the
distribution
impact
of
that
change.
2
auge
inequality
are
the
ratio
of
the
average
population
to
the
intermediate
quintile
(Average/Q3),
intile summarises
the corresponding
population
group.while
Thethe Average/Q3 ratio is not very sensitive to small changes in
be interpreted
with caution.
For instance,
bution of net wealth and wealth in broad sense
eaverage
highestpopulation
quintile
totothe
intermediate
quintile
(Q5/Q3)
and theifratio
of the are
highest
quintile
to
3
the
intermediate
quintile
(Average/Q3),
To
simplify
analysis,
inreacts
this
article
the distribution,
the the
Q5/Q1
ratio
very households
strongly
Q1 figures
very
suggests
that
thelow.
potential redistribution impact
ntile
higher
the
ratios,
more
unequal
the
distribution
is. These ratios should
diate(Q5/Q1).
quintileThe
(Q5/Q3)
and
the
ratio
ofinto
the
highest
quintile
to
are
broken
downthe
wealth
in broad
sense
and
of monetary
policy
on the sense
Portuguese
economy
Table 1 shows descriptive statistics on the level and distribution
of wealth
in broad
and net
wealth in the
with
For instance,
while
the Average/Q3
ratio
is not
very sensitive to small changes in
os,
thecaution.
more unequal
thewealth
distribution
is. assuming
These ratios
net
quintiles,
thatshould
the
average
cannot
be
ignored.
Portuguese economy, on the basis of the Household
Finance and Consumption Survey for 2010 (to see a
n,the
theAverage/Q3
Q5/Q1 ratioratio
reacts
very
strongly
ifsummarises
Q1 to
figures
very
low.
is each
not very
sensitive
smallare
changes
in 3
of
quintile
the
correspondcomparison between
the outcome
of the Household Finance
Consumption
Survey
2010exercisand that for
Thisand
article
includes a number
of for
stylised
3
2
ngly
if Q1 figures
are very
low.
The ratios
used to
ingthe
population
descriptive
statistics
on
level2016).
andgroup.
distribution
of wealth
in gauge
broad
sense and
wealthininbroad
the sense and net wealth
2013, see
Costa,
The various
distribution
quintiles
fornet
wealth
es discussing in what way monetary policy chang-
inequality
aresense
the ratio
the
average
popula- Survey for 2010 (to see a
onomy,
on the
basis
of
the
Household
Finance
and
Consumption
nd distribution
of
wealth
in broad
andof
net
wealth
in
the
comprise
different
households.
Indeed,
there
is a relatively
small
overlap between
households
es can be
transmitted
to household
wealth. across
tion
to
the intermediate
quintile
(Average/Q3),
tween Finance
the outcome
of
the
Household
Finance
and
Consumption
Survey
for
2010
and
that
for
ehold
and
Consumption
Survey
for
2010
(to
see
a
distribution quintiles, as illustrated in Table 2. For instance,
only 48policy
per has
centdistribution
of households
Monetary
effects in
via the
sev-lowest
the ratio
of Survey
the highest
quintile
to the
intermedista, 2016).
distribution
quintiles
in broaderal
sense
and net
wealth
hold
FinanceThe
andvarious
Consumption
for 2010
and
that
quintile
of wealth
in broad
sensefor
arewealth
alsoforincluded
in
the
lowest
wealth quintile.
This
different
channels.
Thenet
unexpected
impact on
price
ate
quintile
(Q5/Q3)
and
the
ratio
of
the
higherent
households.
Indeed,
there
is
a
relatively
small
overlap
between
households
across
tion quintiles
for
wealth
in
broad
sense
and
net
wealth
household composition in both measures of wealth distribution
a key
role when
interpreting
levels isplays
one of
the most
studied
channels inthe
lit- results
estoverlap
quintile
to
the lowest
quintile
as illustrated
in
2. 3.
For
only
48 (Q5/Q1).
per
centThe
of households
the lowest
4 instance,
eintiles,
is a relatively
small
between
households
across
obtained
in Table
Section
erature. Thisin‘Fisher
channel‛ has been analysed
higher
the
ratios,
the
more
unequal
the
distribuealth
in broad
also
in the
lowest
net wealthin quintile.
This
different
or instance,
onlysense
48 perare
cent
of included
households
in the
lowest
literature for a long time, starting with Fisher
Table 1 illustrates
the high
inequality
levels
in net with
wealth and
wealth in broad sense in Portugal. In particular,
tion is.
These
ratios
should be
interpreted
mposition
of
wealth
distribution
plays
a key role
when
interpreting
the results
included in
inboth
the measures
lowest
net
wealth
quintile.
This
different
(1933).
Briefly, an
shock leading to
the
average
value
of
net
wealth
in
the
first
distribution
quintile
is
close
to zero.unexpected
This explains
the high inequality in
4
caution.
For
instance,
while
the
Average/Q3
ratio
ction
3.
alth distribution
plays a key role when interpreting the results
an increase in the consumer price index has a
net wealthisas
measured
by the
ratio. Turning
wealth in broad sense, its dispersion is much lower, given
not
very sensitive
toQ5/Q1
small changes
in the to
disredistribution
effect
by changing the real value
tes the highthat
inequality
levels
in
net
wealth
and
wealth
in
broad
sense
in heterogeneity
Portugal. In
particular,
human wealth’s dispersion is relatively low. Overall, the
prevailing in the distribution of net
tribution, the Q5/Q1 ratio reacts very strongly if
of household
exposures.inAn increase in prices
uewealth
of net wealth
in the
first
distribution
quintile
issuggests
close
to that
zero.the
This
explains
the high inequality
et
and
wealth
in
broad
sense
in
Portugal.
In particular,
wealth
and
broad
sense
potential
redistribution
impact of monetary policy on the
3
Q1wealth
figuresinare
very low.
redistributes
from households with
measured
by the
Q5/Q1
Turning
to wealth
in broad
sense,inits dispersion
is much resources
lower, given
ution quintile
is close
to ratio.
zero.
This explains
the
inequality
Portuguese
economy
cannot
be high
ignored.
Table 1
descriptive
statistics
theprevailing
level
positive
nominal wealth
positions to households
ealth’s
dispersion
is relatively
low. Overall,
heterogeneity
in the distribution
of net
ng
to wealth
in broad
sense,
itsshows
dispersion
isthe
much
lower,on
given
This article
includes
a
number
of
stylised
exercises
discussing
in whatnominal
way monetary
policy
changes can be
and
distribution
of
wealth
in
broad
sense
and
net
with
negative
positions,
which means
alth
in broad
sense suggestsprevailing
that the in
potential
redistribution
. Overall,
the heterogeneity
the distribution
of netimpact of monetary policy on the
transmitted
to household
wealth. Monetary
policy
has distribution effects via several channels. The unexpected
wealth
in the Portuguese
economy,
on
the
onomy
cannot
be ignored.
the potential
redistribution
impact
of monetary
policy on
the basis that it generates a reduction in real wealth for
impact on
levels is one
of the
studied channels
in literature. This “Fisher channel” has been
of price
the Household
Finance
andmost
Consumption
creditors and an increase in real wealth for debtludes a number
of stylised
exercises
what way
policy Briefly,
changes
be
analysed
in literature
for a discussing
long time, in
starting
withmonetary
Fisher (1933).
ancan
unexpected
shock leading to an
Survey for 2010 (to see a comparison between
ors, if assets held at the time of the policy shock
household
wealth.
policy
has distribution
effects
The
unexpected
cises discussing
inMonetary
what
wayconsumer
monetary
policyindex
changes
beseveral channels.
increase
in
hascan
a via
redistribution
by changing
real value of household
areeffect
denominated
in cashthe
units.
thethe
outcome of theprice
Household Finance
and Conlevels
is exposures.
one effects
of the via
most
studied
channels
in
literature.
This
“Fisher
channel”
has
been
ycehas
distribution
several
channels.
The
unexpected
An increase in prices redistributes resources from households with positive nominal wealth
sumption Survey for 2010 and that for 2013, see
raturechannels
for apositions
longintime,
with
Fisherchannel”
(1933). Briefly,
an unexpected
shockmeans
leadingthat
to an
udied
literature.
This
“Fisher
has
been
tostarting
households
negative
nominal
positions, which
it generates a reduction in
Costa,
2016). The with
various
distribution
quintiles
eith
consumer
price
index
has
a
redistribution
effect
by
changing
the
real
value
of
household
Fisher (1933).
Briefly,
an
unexpected
shock
leading
to
an
real wealth for creditors and an increase in real wealth for debtors, if assets held at the time of the policy
increase ineffect
prices
redistributes
households with positive nominal wealth
distribution
by
changing
theresources
real
valuefrom
of household
shock
are
denominated
in cash
units.
ouseholds
nominal
which means
resources with
fromnegative
households
with positions,
positive nominal
wealth that it generates a reduction in
Another redistribution channel attributable to monetary policy results from developments in nominal interest
creditors
and
an
increase
in
real
wealth
for
debtors,
if assets
held at the time of the policy
al positions, which means that it generates a reduction
in
rates. Persistently increasing the average inflation rate, as well as the short-term nominal interest rate, results in
indebtors,
cash units.
alominated
wealth for
if assets held at the time of the policy
a relatively larger increase in taxes for cash-holding households – with approximately zero nominal profitability
96
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Table 1 • Distribution of wealth in broad sense and net wealth in Portugal
| Values in euros
Wealth in broad sense
Net wealth
Q1
291,814
1,282
Q2
386,690
34,010
Q3
495,483
78,982
Q4
698,424
146,449
Q5
1,375.553
529,285
649,441
157,933
Total
Q5/Q3
2.8
6.7
Q5/Q1
4.7
412.8
Average/Q3
1.3
2.0
Note: Calculations based on HFCS 2010. Average values.
Table 2 • Distribution of the population by quintiles of wealth in broad sense and of net wealth
| In percentage
Quintiles of wealth in
broad sense
Net wealth quintiles
<20
20-40
40-60
60-80
>=80
Total
<20
48
31
16
4
0
100
20-40
26
28
25
18
4
100
40-60
14
21
27
26
11
100
60-80
9
15
20
29
27
100
100
>=80
3
5
11
23
58
Total
100
100
100
100
100
Note: Calculations based on HFCS 2010. Average values.
Another redistribution channel attributable to
Consequently, the redistribution impact of mone-
monetary policy results from developments in
tary policy via both channels described above has
nominal interest rates. Persistently increasing
probably been negligible in the recent past and,
the average inflation rate, as well as the short-
therefore, is not analysed in this article. By con-
term nominal interest rate, results in a rela-
trast, the following section looks into the channels
tively larger increase in taxes for cash-holding
that may have had a significant impact on the dis-
households – with approximately zero nomi-
tribution of net wealth and human wealth among
nal profitability – compared with households
Portuguese households in the recent past. These
holding assets with positive returns. Given that
channels result from the implementation of stand-
relatively poorer agents hold a larger share of
ard monetary policy measures – more specifically,
their wealth in cash, they are negatively affected
through changes in the ECB’s intervention rates –
by the aforementioned policy change compared
as well as the adoption of non-standard monetary
with richer agents. This effect is described in
policy measures, namely through the purchase of
Erosa and Ventura (2002) and Adão and Correia
assets by the Eurosystem.
(2016).
On the one hand, as regards net wealth, the
Since the beginning of the euro area, the Por-
article analyses the transmission channel via
tuguese economy has been characterised by a
asset prices and interest rates in various mar-
monetary regime where nominal interest rates
kets, i.e. the channel through which monetary
and the inflation rate are low and not very volatile.
policy values (positively or negatively) household
Special issue
assets and liabilities. This channel is analysed
unchanged at a given moment for a specific
in detail in Sub-sections 3.1 and 3.2, through a
household. Developments in net wealth over
97
natureza tenderá diminuir as due
taxas de juro innominais partial equilibrium exercise,desta which adopts
a simtime a are
exclusively
to changes
the path (ativas e passivas) e a desta natureza tenderá a diminuir as taxas de juro nominais (ativas e passivas) e a ilar approach to that of Doepke
and Schneider
followed by prices and returns on assets and
aumentar o emprego, a atividade, os preços dos ativos e os lucros das empresas. Estes desta desta natureza natureza tenderá tenderá a diminuir a diminuir as taxas as taxas de juro de juro
nom
aumentar o emprego, a atividade, os preços dos ativos e os lucros das empresas. Estes liabilities. The question, therefore, is how, due
(2006).
são os choques estilizados a analisar nas subseções seguintes. desta natureza tenderá a diminuir as taxas de juro aumentar o emprego, a atividade, os preços dos ativos e
nominais (ativas e passivas) e a aumentar o emprego, a atividade, os preços dos ativo
to monetary
policy,
developments
in taxas thesede juro nominais (ativas e p
desta natureza tenderá a diminuir as On the other hand, são os choques estilizados a analisar nas subseções seguintes. as regards human wealth,
aumentar o emprego, a atividade, os preços dos ativos e os lucros das empresas. Estes prices or returns são os choques estilizados a analisar nas subseções seguin
change
over time and how
são os choques estilizados a analisar nas subseções se
Sub-section 3.3 describes the redistribution
aumentar o emprego, a atividade, os preços dos ativos e os lucros das emp
A estratégia de identificação adotada mantém fixa a composição da carteira de ativos this
leads
to
different
developments
wealth e a desta natureza tenderá a diminuir as taxas de juro nominais (ativas e inpassivas) são os choques estilizados a analisar nas subseções seguintes. impact of monetary
policy
via the
labour income
A estratégia de identificação adotada mantém fixa a composição da carteira de ativos são os choques estilizados a analisar nas subseções seguintes. A estratégia de identificação adotada mantém fixa a com
A estratégia de identificação adotada mantém fixa a
for each household,
under
the assumption
of A evolução do valor da num para uma determinada família. channel. This channel
arises
duedeterminado to changes in momento aumentar o emprego, a atividade, os preços dos ativos e os lucros das empresas. Estes num determinado momento para uma determinada família. A evolução do valor da unvarying portfolios.
This
exercise
is along
the para labour income,A estratégia de identificação adotada mantém fixa a composição da carteira de ativos associated with monetary polnum num determinado determinado momento momento para uma uma determinada determinada
fam
A estratégia de identificação adotada mantém fixa a composição da carte
riqueza líquida ao longo do tempo é devida unicamente a alterações nas trajetórias de são os choques estilizados a analisar nas subseções seguintes. lines as that developed
byevolução Domanski
et valor al.
icy measures, num with adeterminado varying impact
on differriqueza líquida ao longo do tempo é devida unicamente a alterações nas trajetórias de momento para same
uma determinada família. A do da riqueza líquida ao longo do tempo é devida unicamente riqueza líquida ao longo do tempo é devida unicamen
num determinado momento para uma determinada família. A evolução preços e retornos dos ativos e responsabilidades. A questão será então como, devido à (2016), taking Doepke and Schneider (2006) as
ent segments of the population.
This exercise
A estratégia de identificação adotada mantém fixa a composição da carteira de ativos riqueza líquida ao longo do tempo é devida unicamente a alterações nas trajetórias de preços e retornos dos ativos e responsabilidades. A questão será então como, devido à preços e retornos dos ativos e responsabilidades. A quest
preços e retornos dos ativos e responsabilidades. A q
riqueza líquida ao longo do tempo é devida unicamente a alterações nas t
is based on Household Finance
and Consumpa starting point.
política monetária, se altera a evolução ao longo do tempo destes preços ou retornos e num determinado momento para uma determinada política monetária, se altera a evolução ao longo do temp
família. A evolução do valor da preços e retornos dos ativos e responsabilidades. A questão será então como, devido à política monetária, se altera a evolução ao longo do tempo destes preços ou retornos e política monetária, se altera a evolução ao longo do t
tion Survey data,
together
with microeconomic
preços e retornos dos ativos e responsabilidades. A questão será então com
como daí decorre uma evolução diferente da riqueza de cada família, sob a hipótese de 3.1. Transmission channel via asset prices:
como daí decorre uma evolução diferente da riqueza de cada família, sob a hipótese de política monetária, se altera a evolução ao longo do tempo destes preços o
stock valuation
carteiras invariantes. Trata‐se de um exercício na mesma linha do desenvolvido por preços e retornos dos ativos e responsabilidades. A questão será então como, devido à como daí decorre uma evolução diferente da riqueza de cada família, sob a hipótese de carteiras carteiras invariantes. invariantes. Trata‐se Trata‐se de um de um exercício exercício na mesm
na m
como daí decorre uma evolução diferente da riqueza de cada família, sob a
carteiras invariantes. Trata‐se de the
um exercício na mesma do desenvolvido por One of
transmission
channels
oflinha mone3. Distribution
channels
of
monetary
Domanski et al. (2016), partindo de Doepke e Schneider (2006). política monetária, se altera a evolução ao longo do tempo destes preços ou retornos e carteiras invariantes. Trata‐se de um exercício na mesma linha do desenvolvido por Domanski et al. (2016), partindo de Doepke e Schneider (
Domanski et al. (2016), partindo de Doepke e Schneid
carteiras Trata‐se de um inexercício tary policyinvariantes. to the distribution
of wealth
broad na mesma linha do desen
policy in Portugal Domanski et al. (2016), partindo de Doepke e Schneider (2006). política monetária, se altera a evolução ao longo do tempo destes preços ou retornos e data from the riqueza líquida ao longo do tempo é devida unicamente a alterações nas trajetórias de Labour
Survey.
como daí decorre uma evolução diferente da riqueza de c
como daí decorre uma evolução diferente da riqueza como daí decorre uma evolução diferente da riqueza de cada família, sob a hipótese de Domanski et al. (2016), partindo de Doepke e Schneider (2006). sense stems from valuation changes in the stock
Domanski et al. (2016), partindo de Doepke e Schneider (2006). How does a monetary policy
change
affect the
3.1
Canal de transmissão pelos preços dos ativos: valorização do stock 3.1 3.1
Canal de transmissão pelos preços dos ativos: va
Canal de transmissão pelos preços dos ativo
of assets.
At first
sight,
given
the
carteiras invariantes. Trata‐se de um exercício na mesma linha do concentration
desenvolvido por 3.1in broad
Canal de transmissão pelos preços dos ativos: valorização do stock distribution of wealth
sense and net
3.1 Canal de transmissão pelos preços dos ativos: valorização do stock of
assets
in
the
top
wealth
quintile
(wealth in
Um dos canais de transmissão da política monetária à dist
Um dos canais de transmissão da política monetária à Um dos canais de transmissão da política monetária à distribuição da riqueza em sentido 3.1 Canal de transmissão pelos preços dos ativos: valorização do sto
Domanski et al. (2016), partindo de Doepke e Schneider (2006). wealth of households
in Portugal?
This section
Um dos canais de transmissão da política monetária à distribuição da riqueza em sentido broad sense and,lato decorre de alterações na valorização do stock de at
in particular,
net wealth), it
Um dos canais de transmissão da política monetária à distribuição da riqueza em sentido lato decorre de alterações na valorização do stock d
aims to answer this question.
The answer is proUm dos canais de transmissão da política monetária à distribuição da riquez
lato decorre de alterações na valorização do stock de ativos. À primeira vista, poderia could be expected that an increase in the value
lato decorre de alterações na valorização do stock de ativos. À primeira vista, poderia 3.1
Canal de transmissão pelos preços dos ativos: valorização do stock lato decorre de alterações na valorização do stock de ativos. À primeira vista, poderia vided within a partial
equilibrium framework and
parecer que, dada a concentração dos ativos no último q
parecer que, dada a concentração dos ativos no últim
lato decorre de alterações na valorização do stock de ativos. À primeira v
parecer que, dada a concentração dos ativos no último quintil da riqueza (em sentido of assets would lead to an increase in inequalunvarying composition
of
investment
portfoUm dos canais de transmissão da política monetária à distribuição da riqueza em sentido parecer que, dada a concentração dos ativos no último quintil da riqueza (em sentido parecer que, dada a concentração dos ativos no último quintil da riqueza (em sentido lato e, em particular, em termos líquidos), um aumento lato e, em particular, em termos líquidos), um aume
parecer que, dada a concentração dos ativos no último quintil da riqueza ity. However, this is not necessarily the case.
lato e, em particular, em termos líquidos), um aumento do valor dos ativos implicaria lios, thereby excluding agents’
reaction to polilato decorre de alterações na valorização do stock de ativos. À primeira vista, poderia lato e, em particular, em termos líquidos), um aumento do valor dos ativos implicaria um aumento da desigualdade. Esta conclusão não é, no e
um aumento da desigualdade. Esta conclusão não é, lato e, em particular, em termos líquidos), um aumento do valor dos ativos implicaria Indeed, a specific increase
in the value of assets
lato e, em particular, em termos líquidos), um aumento do valor dos ativ
cy changes. A stylised approach is used on the
um aumento da desigualdade. Esta conclusão não é, no entanto, correta. De facto, um parecer que, dada a concentração dos ativos no último quintil da riqueza (em sentido um aumento da desigualdade. Esta conclusão não é, no entanto, correta. De facto, um impacts on wealthdeterminado aumento do valor dos ativos impacta a distri
(in broad
sense and net) disdeterminado aumento do valor dos ativos impacta a d
redistribution impactum aumento da desigualdade. Esta conclusão não é, no entanto, correta. De facto, um of monetary policy via valum aumento da desigualdade. Esta conclusão não é, no entanto, correta. D
tribution, not only directly on the assets per se,
determinado aumento do valor dos ativos impacta a distribuição da riqueza (em sentido uation changeslato e, em particular, em termos líquidos), um aumento do valor dos ativos implicaria in the stockdeterminado aumento do valor dos ativos impacta a distribuição da riqueza (em sentido of net wealth (Sub-
lato e líquida) não só diretamente nos próprios ativos, ma
lato e líquida) não só diretamente nos próprios ativos
determinado aumento do valor dos ativos impacta a distribuição da riqueza
determinado aumento do valor dos ativos impacta a distribuição da riqueza (em sentido but also via the leverage of the various populasection 3.1), income
flowslato e líquida) não só diretamente nos próprios ativos, mas também através do grau de generated by this
um aumento da desigualdade. Esta conclusão não é, no entanto, correta. De facto, um lato e líquida) não só diretamente nos próprios ativos, mas também através do grau de alavancagem dos diferentes quintis da população. Esta re
alavancagem dos diferentes quintis da população. Es
lato e líquida) não só diretamente nos próprios ativos, mas também atravé
tion quintiles. This link is illustrated by equation
lato e líquida) não só diretamente nos próprios ativos, mas também através do grau de stock of net wealth (Sub-section 3.2) and labour
determinado aumento do valor dos ativos impacta a distribuição da riqueza (em sentido alavancagem dos diferentes quintis da população. Esta relação encontra‐se explicitada alavancagem dos diferentes quintis da população. Esta relação encontra‐se explicitada na equação (3) para o caso da riqueza líquida, definida co
na equação (3) para o caso da riqueza líquida, definida
alavancagem dos diferentes quintis da população. Esta relação encontra‐s
(3) for net wealth, which
is defined as the differincome changes (Sub-section
3.3).
alavancagem dos diferentes quintis da população. Esta relação encontra‐se explicitada lato e líquida) não só diretamente nos próprios ativos, mas também através do grau de na equação (3) para o caso da riqueza líquida, definida como a diferença entre os Ativos ence between assets
( as ) and
( ) of
giv-um (�) (�) e e Dívidas as debt
Dívidas (�) (�) de aum de dado dado quintil quintil da distribuição da distribui
na equação (3) para o caso da riqueza líquida, definida como a diferença entre os Ativos na equação (3) para o caso da riqueza líquida, definida como a diferença en
The exercise focuses on stylised monetary polna equação (3) para o caso da riqueza líquida, definida como a diferença entre os Ativos en distribution
quintile
( �).). The
equationrevela showsque o alavancagem dos diferentes quintis da população. Esta relação encontra‐se explicitada A equação (�) e as Dívidas (�) de um dado quintil da distribuição (�
aumento percentual da riqueza líquida de um determina
aumento percentual da riqueza líquida de um determ
icy changes towards greater
monetary
accome as dado Dívidas (�) de um dado quintil (�
da � ). distribuição (�revela (�) e as Dívidas (�) (�) de um quintil da distribuição A equação que o r
� ). A equação that
the percentage
increase
in net (�
wealth
of
a
(�) e as Dívidas (�) de um dado quintil da distribuição ). A equação revela que o �
�
na equação (3) para o caso da riqueza líquida, definida como a diferença entre os Ativos aumento percentual da riqueza líquida de um determinado quintil, decorrente de um �
modation, through a reduction in interest rates
�� ��� �
aumento percentual da riqueza líquida de um determinado quintil, decor
, será igual ao
, será igu
aumento percentual do valor dos ativos aumento percentual do valor dos ativos aumento percentual da riqueza líquida de um determinado quintil, decorrente de um given quintile,
resulting
from a percentage rise
� �� � ��
�
and the implementation of non-standard mon-
(�) e as Dívidas (�) de um dado quintil ��da � , será igual ao produto entre esse aumento distribuição (�� ). A equação revela que o aumento percentual da riqueza líquida de um determinado quintil, decorrente de um ����
aumento percentual do valor dos ativos �
�
����
� ��
aumento percentual do valor dos ativos , será igual ao produto entre es
etary policy measures. Although
assessing
all
�
percentual do valor dos ativos e um fator percentual do valor dos ativos e um fator que depe
que d
in the value of assets
, is obtained by�mul�
aumento percentual do valor dos ativos ��� ���
aumento percentual da riqueza líquida de um determinado quintil, decorrente de um ���� ��� , será igual ao produto entre esse aumento �� ��� �
�
aumento percentual do valor dos ativos , será igual ao produto entre esse aumento � � � �
transmission mechanisms
poses a number of
�
percentual do valor dos ativos e um fator � que depende positivamente do grau de � ��
�this
percentual do valor dos ativos e um fator �� � percentage
�
tiplying
increase
in the value �of�� que depende positivament
����
�
�
aumento percentual do valor dos ativos , será igual ao produto entre esse aumento percentual do valor dos ativos e um fator que depende positivamente do grau de challenges, one
can argue that
such a monetary
�
�� �
alavancagem desse quintil. � alavancagem desse quintil. � ��
�
���
percentual do valor dos ativos e um fator assets by� a factor
impulse will tend
to reduce
(lending and deposalavancagem desse quintil. �
��
que depende positivamente do grau de ��� that
���positively depends
�� �
alavancagem desse quintil. percentual do valor dos ativos e um fator � que depende positivamente do grau de �
� �
�
�
� �
�
� �
� ��
� �� �
�
�� � � �
� ��1
� �
it) nominal interest rates alavancagem desse quintil. and foster employ�����
������
� � ��� ��
1
�� �
on leverage
�
�
�
�
� � in that quintile.
�� �
��
�� � �� ��� ����� � �
��
�� � �� ��� ��� � �� � ��
�prices
�and
��
��
��
�
�
�
�
�
�
�
�
�
ment, activity, asset
business
profits.
alavancagem desse quintil. �� �� 1 ��
��� � � � ��
�
��
��
��
�
alavancagem desse quintil. � ���
��
��
� � � ��� �
1
1 �1 �
��
��
� �
�� analysed
� � fol� �
These stylised shocks
are
��� � ��� � ��
��� in
��the
���
���
��� � �
�� ��
� ��
�
�
�
�
�
�
�
��
����� � ��� � ���� ����1����
1
�
1 � ��
� ��
��
��
��
lowing sub-sections. ����� � � �� �
�
���
��
10 �
�� ��� ��
��� � �
��
�1
� ���� �� ���
� ��
� 10 �
����� � ��� � ����
1
�
�
�
�
�
��� (3)
1
�
��
�� identification
��
�
��
�� ��� � �strat�� �� ��
According to the
�� ��� 10 ���
� ��
��
��adopted
� �� �
� � � � � ���� �
� �
�
egy, the asset portfolio composition remains
10 10 10 10 1�
���
1�
���
஺ಲ ೔ ೔
ο஺ ೂ೔
e value of assets
ଵ (
), is obtained
by multiplying
this percentage
increase
in therelated
value to leverage in each quintile but also negativ
increase
in assets
is not only
positively
஺ ೂ೔
ೂ
ೂವ
aଵವfactor
(ଵି
) depends
that positively
on
leverage
that quintile.
೔ ೔
ο஺
ratio depends
ofinhuman
wealth
to totalinassets
in each quintile:
ೂ೔
) that
positively
on leverage
that quintile.
ೂ
assets
(
ି
ೂ೔ಲ ೔ ), is obtained by multiplying this percentage increase in the value
ಲೂ೔
஺
ೂ೔
ೂ೔
ೂ೔
ೂ೔
ೂ೔
οሺ஺ ି஽ ሻ
ο஺
஺
ο஺
t positively depends
leverage
in that
quintile.
BANCO
DE PORTUGAL
Economic
ൌೂ೔ ஺ೂ೔
ೂBulletin
ೂ• May
98 on
೔ ାுௐ
ೂ೔οሺ஺ൌ
ೂ೔ೂ೔ 2017
ି஽
ሻ ೂ೔ ο஺
( ஺ೂ೔•ି஽
஺ ೂ೔ ೔஺
ି஽
ൌ
(
οሺ஺ ೂ೔ ି஽ೂ೔ ሻ
ି஽ೂ೔ ሻ
ൌ
஺ ೂ೔ ି஽ೂ೔
ೂ
ο஺ ೂ೔
(
஺ ೂ೔
ο஺ ೂ೔ ଵ
ೂ ೂ೔ ೂൌ
ೂ
೔ ஺ ೂ೔ ೔ି஽
ି஽ ೔ ሻ ஺ೂ೔ ο஺
ವೂ೔೔
஺ ೂοሺ஺
ଵି ೂ
ಲ ೔
ೂ೔
ೂ೔
ೂ
஺ ି஽
஺ ೔
ൌ
) ஺ ೂ೔
(
ൌ
஺ ೂ೔ ାுௐ ೂ೔ ି஽ೂ೔
ο஺ ೂ೔
ଵ
ೂ೔
ೂ೔
ೂ೔
)
ವೂ೔
ି஽in
In the case of ஺
wealth
broad஺sense,
ଵିthe impact
ಲೂ೔
of changes
in assets is illustrated by equation (4).
ೂ
ೂ
஺ ೂ೔
ଵ
஺ ೂವ೔ೂ೔
) ൌ ο஺ೂ೔
஺ ೂ೔
ଵି
ೂ೔
஺ ೂ೔ ାுௐಲೂೂ೔ ି஽
೔
(3)
ଵ
ಹೈೂ೔ ವೂ೔
ଵା ೂ ି ೂ
ಲ ೔
ಲ ೔
)
In the case of net wealth, the Portuguese econo-
(3)
my is characterised by a population where grow-
೔
஺ ೔ of wealth
ο஺ ೔ ଵ in broad sense, the impact of changes in assets is illustrated by equatio
Inο஺the
case
ൌ sense,
ൌ In this
h in
changes
is illustrated
by equation
(4). corresponds
In
the wealth
aggregate
includes human
ing wealth
to
ೂ೔ broad
ೂ೔
ೂ೔ the
ೂ೔impact
ೂ೔ofcase,
) in assets
(3)a decreasing indebtವೂ೔
஽
஺
஺ ି஽
஺
ொ
ொ೔
೔ this
As
toincludes
assess
the
impact
transmission
it is necessary
case,
the wealth
aggregate
human
wealth,
( ‫ܹܪ‬
),is highly
and,
therefore,
the impact
of a per
, and,
therefore,
the impact
of a per‫ܹܪ‬
wealth,
ggregatethis
includes
human
wealth,
(ଵି
), and,
therefore,
thedistribution
impact
of a edness
percentage
ratio.via
This
significant
and channel,
implies
ಲೂ೔such,
of
wealth
in related
broadtosense,
the
impact
of
in assets
illustrated
equation
(4).
In househ
centage
increase
inquintile
assets the
isbut
notchanges
onlynegatively
positively
ot only
positively
leverage
in only
each
also
to is
the
thatand
a similar
increase
in the
value
link
between
asset
aggregate
the
amount
of by
liabilities
heldnegatively
by each
increase
in assets
is not
positively
related
to related
leverage
inpercentage
each
quintile
but
also
relate
ொ೔ negarelated to
leverage
in
each
quintile
but
also
h wealth
to total assets
in eachincludes
quintile:
of
assets
across
net
wealth
distribution
quintiles
‫ܹܪ‬
aggregate
human
wealth,
(
),
and,
therefore,
the
impact
of
a
percentage
distribution quintiles. This link is shown in Charts 1 and 2, for the distribution of wealt
ratio of human tively
wealth
to total assets in each quintile:
related to the ratio of human wealth to total
always
improves the relative
of the lowsense,
the only
impact
of changes
is illustrated
by equation
(4).
Inposition
and
net in
wealth
respectively.
Thequintile
slope
of
the lines
each
quintile
corresponds
ssets
is not
positively
related
toassets
leverage
in each
but
also for
negatively
related
to the to levera
assets ೂin each quintile:
est quintiles vis-à-vis the average quintile and the
೔
ொ೔ debt-to-assets
ο஺ ೔
஺ ೔
ଵ
the
ratio. the
The charts
show
the highest wealth quintiles (both net and
‫ܹܪ‬
cludes
human
wealth,
( ο஺
), and,
therefore,
impact
of athat
percentage
ൌ
ൌ
an
wealth
to total
assets
in
each
quintile:
)
highest quintiles. This means that higher asset
ಹೈೂ೔ ವೂ೔
ௐ ೂ೔ ି஽ೂ೔
஺ ೂ೔ ஺ ೂ೔ ାுௐ ೂ೔ ି஽ೂ೔
஺ ೂ೔
ି
ଵା
ೂ೔೔
ೂ
೔ ೔lowest
ಲೂೂ
೔
೔
have
the
οሺ஺ ೂ೔ ାுௐ
ି஽
ሻ
ο஺ ೂleverage.
஺ ೂ೔ negatively
ο஺ ೂincrease
ଵ relatively
sitively related to leverage
inಲeach
quintile
but also
related
to the
values
more the net wealth
ൌ
ൌ
)
( ஺ೂ೔ ାுௐ ೂ೔ ି஽ೂ೔
ಹೈೂ೔ ವೂ೔
஺ ೂ೔ ஺ ೂ೔ ାுௐ ೂ೔ ି஽ೂ೔
஺ ೂ೔ (4)
ି ೂ regardless
of poorerଵା
households,
of asset con- by a populatio
is characterised
ssets ೂin each
quintile: ೂ In theೂ case of netೂ wealth, the Portuguese
೔
ಲೂeconomy
ಲ ೔
οሺ஺ ೔ ାுௐ ೂ೔ ି஽ೂ೔ ሻ
ο஺ ೔
஺ ೔
ο஺ ೔
ଵ
centration. Similarly, in the case of a decrease in
corresponds
a decreasing
ൌ ஺ೂ೔ to ಹೈ
ೂ೔ ವೂ೔
) indebtedness ratio. This is highly significant and imp
( ஺ೂ೔ ାுௐ ೂ೔ ି஽ೂ೔ ൌ ஺ೂ೔ ஺ೂwealth
೔ ାுௐ ೂ೔ ି஽ ೂ೔
ுௐ ೂ೔ ି஽ೂ೔ ሻ
ೂ
ೂ
the value
of assets, more leveraged households
e distribution impact via this transmission channel,
it is ଵା
necessary
to look
into the
(4)
ೂ ି ೂ
percentage increase in ಲthe
of assets across net wealth distribution quintiles alw
೔ value
ಲ ೔
would post
a larger decline in net wealth.
set
the amount
held by each household typifying
the
ೂ೔
೔
ο஺aggregate
஺ ೂand
ο஺ ೂ೔ of liabilities
ଵ
relative position of the lowest quintiles vis-à-vis the average quintile and the
(4)highest qui
ൌThisೂ೔linkೂ೔is shown
ൌ ೂ೔1 and
ೂ೔ distribution
2,ೂ೔forವthe
of wealth in broad
sense
)
In the case
of the distribution of wealth in broad
ಹೈ
஺
஺ ାுௐ ೂ೔ ି஽inೂ೔Charts
஺
ି ೂ asset values
ଵା
that
increase
relatively
more thechannel,
net wealthit of
households,
the
distribution
impact
via this
transmission
is poorer
necessary
to lookre
೔ higher
ಲೂ
ಲ ೔ corresponds
ctively. As
The such,
slope ofto
theassess
lines for each
quintile
to leverage,
measured
by
sense across
quintiles, the link between leverage
concentration.
Similarly,
in
case
of asense)
decrease
in the
value
assets,
more leveraged h
tio. Thelink
charts
show that the
highest
wealth
quintiles
(both
net
and
in broad
(4)
and wealth
is not so clearly
monotonic
(exclud-household typif
between
the
asset
aggregate
andimpact
thethe
amount
of
liabilities
held
byofeach
As such,
to assess
the distribution
via
post a larger
decline
in nettowealth.
age.
assess
the distribution
impact
via
this
transmission
it is necessary
look
the
ing the highest
Therefore,to
in the
caseinto
of
this transmission
channel,
is necessary
look channel,
distribution quintiles.
This link
is itshown
in Charts
1 and 2,quintile).
for the
distribution
of
wealth in broa
a homogeneous
increase
in the value of
assets,
into the
linkcharacterised
between
the
asset
aggregate
and where
ealth,
Portuguese
economy
is
by adistribution
population
growing
n
thetheasset
aggregate
and
the
amount
of
liabilities
byin each
household
the
In
the
case
of slope
the
of held
wealth
broad
sense
acrosstypifying
quintiles,
the link meas
betwe
and net wealth respectively.
The
of the lines
for
each
quintile
corresponds
to leverage,
the final
impact on inequality will result from the
the
amount
of
liabilities
held
by
each
household
o
a
decreasing
indebtedness
ratio.
This
is
highly
significant
and
implies
that
a
similar
tion impact
vialink
thisistransmission
it is2,necessary
to(excluding
look into
the highest
wealth
ischannel,
not so
clearly
monotonic
quintile).
Therefore, in
quintiles.
This
shown
in Charts
1 and
for
the distribution
of the
wealth
in broad
sense
weighing
of
relative share
of human
wealthnet and in broad
ratio.
The
charts
show
highest
wealth
quintiles
(both
typifying
the distribution
quintiles.
Thisthat
link isthe
in the the
valuedebt-to-assets
of assets across
net
wealth
distribution
quintiles
always
improves
thethe
increase
in the
value
of assets,
the
impact measured
on
inequalityby
will result fro
gate
and the amount
of inhomogeneous
liabilities
held
bydistribution
each
household
typifying
lth
respectively.
The slope
of
the
each
quintile
to final
leverage,
inThis
assets
and
leverage,
asthe
illustrated by
equation
Charts
1lines
and and
2, for
forthe
the
of corresponds
e lowest
quintiles
the
average
quintile
highest
quintiles.
means
have
the vis-à-vis
lowestshown
leverage.
the
relative
share
ofwealth
human
wealth
inChart
assets
andthat
leverage,
as
illustrated
by equation (4). C
(4).
3broad
shows
the ratio
of in
human
wealthsense)
is
in Charts
1theand
2,
for
the
distribution
of
wealth
sense
wealth
in
broad
sense
and
net
respeces shown
increase
relatively
net
wealth
of
poorer
households,
regardless
ofin
asset
assets
ratio.
Themore
charts
show
that
the
highest
wealth
quintiles
(both
net
and
broad
ுௐ ೂ೔
to assets
(the
term
in )equation
the
ratio
of
human
wealth
tohouseholds
assets
(the
term
(
in equation
(4)) is lower for
the h
tively.
The
slope
the
lines
forleveraged
each quintile
Incase
the
of
net
wealth,
the
Portuguese
economy
is
characterised
by(4))ais lower
population
where
in the
of case
a lines
decrease
ineach
the
value
of of
assets,
more
would
ey,slope
of
the
for
quintile
corresponds
to
leverage,
measured
஺ ೂ೔ by
west
leverage.
for the highest quintile of wealth in broad sense
corresponds to leverage, measured by the debtn net wealth.
wealththat
corresponds
towealth
a decreasing
indebtedness
ratio.
This issense)
highly
significant
and leverage
implies inthat
wealth
in broad
sense
than for
the
other
in contrast
to the lower
this
harts show
the highest
quintiles
(both
net
and
in quintiles,
broad
than for by
the other
quintiles, in contrast
to the
to-assets
ratio.
The
charts
show
that
the
highest
of
net
wealth,
the
Portuguese
economy
is
characterised
a
population
where
growing
tribution
of wealth in broad
sense
quintiles,
linkreveals
between
leverage
and is predominant.
empirical
simulation
which
effect
Chart 4 shows
thisimpro
ratio
percentage
increase
inacross
the(both
value
ofthe
across
netleverage
wealth
distribution
always
lower
in this
quintile. Only anquintiles
empirical
wealth quintiles
net and
inassets
broad sense)
early monotonic
(excluding the
highest quintile).
Therefore,
inhighly
the case
of a
esponds
to a decreasing
indebtedness
ratio.
Thisinisthe
significant
and implies
that a similar
quintiles,
but
has
no role
simulation
reveals
which effect
is predominant.
lowest
leverage.
relative positionhave
ofthe
the
lowest
quintiles
vis-à-visanalysis.
the
average
quintile
and
the highest quintiles. Thi
se
in the value of economy
assets, the final
impact
on inequality will
result
from the weighing
of growing
Portuguese
characterised
bywealth
a population
where
increase
in the
value of isassets
across
net
distribution
quintiles
always
improves
For
that
purpose,
Table
3
(first
column)
presents
a of
stylised
simulation
of thethe
impact
of cha
that
higher
asset
values
increase
relatively
more
the
net
wealth
poorer
households,
regardless
uman wealth in assets and leverage, as illustrated by equation (4). Chart 3 shows that
asing
indebtedness
ratio.
is highly
significant
and implies
that
acase
similar
tion of
the lowest quintiles
vis-à-vis
the
average
quintile
and inthe
This means
ுௐ ೂ೔ This
ratios,
thehighest
of quintiles.
bothmore
the wealth
in broad
sense an
ealth toconcentration.
assets (the term ( Similarly,
)ofin assets
equation
(4))some
is lower
for
the highest
quintile
of
in thefor
case
of ainterquintile
decrease
in
the value
of assets,
leveraged
household
஺ ೂ೔
ue
of values
assets
across
net
wealth
distribution
always
improves
theHousehold
distributions.
Calculations
made
on the
basis of the
and Consum
asset
increase
relatively
more
the
net quintiles
wealth
of
poorer
households,
regardlessFinance
of asset
than for
the a
other
quintiles,
in contrast
the lower
leverage
inwere
this quintile.
Only
an
post
larger
decline
in nettowealth.
The
table
confirms
that
similar
increasehouseholds
in the value ofwould
assets across dis
quintiles
vis-à-vis
average
quintile
and
highest
quintiles.
This
means
which
is predominant.
Chart
4in
shows
this
ratio
for
net percentage
wealth
n.reveals
Similarly,
ineffect
the the
case
of 2010.
a decrease
thethe
value
ofaassets,
more
leveraged
Chart 1 • Relation between assets and liabilities,
Chart 2 • Relation between assets and liabilities,
In analysis.
the
case ofof
the
distribution
of wealth
inbybroad
sense
across
quintiles,
linkinequality
between
lever
role
in the
fosters
inequality
in the
distribution
of
wealth
in broad
sense and the
reduces
in net
we
by net
quintiles
wealth
in
broad
sense
net wealth
quintiles
edecline
relatively
more
the
net
wealth
of poorer
households,
regardless
of asset
in
wealth.
is
not
so
clearly
(excluding
the
quintile).
thebyca
le 3 (first
presents
stylised
of the
impact
of changes
inhouseholds
the value
Thesimulation
table
also
breaks
down
the impact
ofhighest
the valuation
of the Therefore,
various assetsinheld
ho
ase
ofwealth
acolumn)
decrease
in athe
value
of monotonic
assets,
more
leveraged
would
of the distribution
of
wealth
in
broad
sense
across
quintiles,
the link between leverage and
nterquintile
ratios,
in
the
case
of
both
the
wealth
in
broad
sense
and
the
net
wealth
homogeneous increase
in the broken
value of
assets,
theassets
final impact
on inequality
will
result
from
the we
are typically
down
into real
and financial
assets. Real
assets
chiefly
include
re
h. were
ons
on the
basis of the Household
Finance
andhighest
Consumption
Survey for
ot so made
clearly
monotonic
(excluding
the
quintile).
Therefore,
in
the
case
of
a
5
employment
businesses
, and
a large share
in net wealth
households
Portugal
an
the
relative
shareincrease
of human
wealth
in assets
andhold
leverage,
as illustrated
by of
equation
(4).inChart
3 sh
ms
that
a similar
percentage
in the value
of assetsthe
across
distribution
quintiles
of
wealth
in
broad
sense
across
quintiles,
link
between
leverage
and
ೂ
೔
us increase in the value of assets, the final impact on ுௐ
inequality
will result from the weighing of
e distribution
of wealth
broad sense
and reduces
inequality
net wealth
the ratio
of in
human
wealth
to assets
(theinterm
( distribution.
) in equation (4)) is lower for the highest qu
ೂ೔
஺ in the case of a
notonic
(excluding
thein assets
highestand
quintile).
hare of human
wealth
leverage,Therefore,
as illustrated
by equation (4). Chart 3 shows that
down wealth
the impact
of
the
valuation
of
the
various
assets
held
by
households.
Assets to the lower leverage in this quintile.
ೂother
in
broad
sense
than
for
the
quintiles,
in
contrast
೔
ுௐ
value
of
assets,
the
final
impact
on
inequality
will
result
from
thelower
weighing
of highest quintile of
human
wealth
to
assets
(the
term
(
)
in
equation
(4))
for the
own into real assets and financial assets. Real assets ೂchiefly
include real estate andisself೔
஺
empirical
which
effect is predominant.
Chart
alth
assets
andsimulation
leverage,
asreveals
illustrated
byin equation
3 shows
that 4 shows this ratio for net
ses5,in
and
hold a large
share
in net wealth
of households
Portugal and(4).
mostChart
euro area
oad sense than for the
other
quintiles,
in
contrast
to
the
lower
leverage
in
this quintile. Only an
ೂ೔ no role in the analysis.
quintiles,Note:
but
has
ுௐ
Based
on the Household
Finance and Consumption
Survey
for
Note:highest
Based on the Household
Finance
and Consumption Survey for
ssets
(the
term
(
)
in
equation
(4))
is
lower
for
the
quintile
of
೔
mulation reveals
which
effect is predominant. Chart
2010. ஺ ೂ
2010. 4 shows this ratio for net wealth
For that
purpose,
Table 3to(first
presents
a stylised
simulation
other
quintiles,
contrast
thecolumn)
lower leverage
in this
quintile.
Only anof the impact of changes in t
uthe
has
no role
in thein
analysis.
assetsisforpredominant.
some interquintile
ratios,
in the
both
thewealth
wealth in broad sense and the ne
whichof effect
Chart 4
shows
thiscase
ratiooffor
net
pose, Table 3 (first column) presents a stylised simulation of the impact of changes in the value
distributions. Calculations were made on the basis of the Household Finance and Consumption Su
analysis.
r some interquintile ratios, in the case of both the wealth in broad sense and the net wealth
2010. The table confirms that a similar percentage increase in the value of assets across distribution
Special issue
Chart 4 shows this ratio for net wealth quintiles,
made on the basis of the Household Finance
but has no role in the analysis.
and Consumption Survey for 2010. The table
For that purpose, Table 3 (first column) presents
confirms that a similar percentage increase in
a stylised simulation of the impact of changes in
the value of assets across distribution quintiles
the value of assets for some interquintile ratios,
fosters inequality in the distribution of wealth
in the case of both the wealth in broad sense and
in broad sense and reduces inequality in net
the net wealth distributions. Calculations were
wealth distribution.
99
Table 3 • Impact of changes in the value of assets on the distributions of wealth in broad sense
and net wealth | Percentage changes in initial ratios, excluding ratio values
Ratios based on quintiles of wealth in broad sense
Initial ratios
10% increase
in real estate,
tradable assets
and businesses
(1)
(2)
(3)
(4)
Q5/Q3
3.6
0.305
-0.277
0.063
0.533
Q5/Q1
11.9
0.343
-0.292
0.060
0.589
Average/Q3
1.4
0.150
-0.157
0.034
0.279
10% increase
in real estate
10% increase
in tradable assets
10% increase
in the values
of businesses
10% increase
in real estate
10% increase
in tradable assets
10% increase
in the values
of businesses
Ratios based on net wealth quintiles
Initial ratios
10% increase
in real estate,
tradable assets
and businesses
(1)
(2)
(3)
(4)
Q5/Q3
6.7
-1.411
-3.221
0.120
1.877
Q5/Q1
412.4
-48.701
-49.625
-0.015
1.686
2.0
-0.761
-1.967
0.076
1.255
Average/Q3
Note: Calculations based on HFCS 2010.
Chart 3 • Relation between human wealth and
assets, by quintiles of wealth in broad sense
Note: Calculations based on the Household Finance and Consumption Survey for 2010.
Chart 4 • Relation between human wealth and
assets, by net wealth quintiles
Note: Calculations based on the Household Finance and Consumption Survey for 2010.
100
BANCO DE PORTUGAL • Economic Bulletin • May 2017
The table also breaks down the impact of the val-
from the impact of high leverage on the first net
uation of the various assets held by households.
wealth quintile, as explained above (see Chart 2).
Assets are typically broken down into real assets
Similarly, self-employment businesses also con-
and financial assets. Real assets chiefly include
tribute, as expected, towards increased inequality
real estate and self-employment businesses5, and
hold a large share in net wealth of households in
in wealth distribution.
3.2. Transmission channel via return on net
Portugal and most euro area countries. In turn,
Por seu turno, os ativos financeiros são fundamentalmente compostos por depósitos e Por seu turno, os ativos financeiros são fundamentalmente compostos por depósitos e financial assets mainly comprise deposits and
assets: income flows
títulos transacionáveis, que incluem ações, obrigações e unidades de participação em títulos transacionáveis, que incluem ações, obrigações e unidades de participação em marketable securities, including shares, bonds
o, os ativos financeiros são fundamentalmente compostos por depósitos e fundos de investimento. fundos de investimento. and mutual fund investment units.
A monetary policy change may lead to transformations in the distribution of household wealth,
cionáveis, que incluem ações, obrigações e unidades de participação em In the case of a percentage increase in a comnot only due to changes in asset valuation (SubPara o caso de um aumento percentual de uma componente do ativo total das famílias, Para o caso de um aumento percentual de uma componente do ativo total das famílias, ponent of total household assets, e.g. housing,
vestimento. section 3.1), but also due to changes in income
��
por exemplo a habitação, designada por �
, a equação (3), relativa à distribuição da equation (3) on net wealth
represented as ��,, a equação (3), relativa à distribuição da por exemplo a habitação, designada por �
flows stemming from policy changes. This subdistribution
would
be
re-written
as
follows
(and
e um aumento percentual de uma componente do ativo total das famílias, section analyses the issue, once again by simuriqueza líquida, seria reescrita como segue (e analogamente a equação (4) para o caso riqueza líquida, seria reescrita como segue (e analogamente a equação (4) para o caso analogously for equation (4) on wealth in broad
��
a habitação, designada por �
, a equação (3), relativa à distribuição da da riqueza em sentido lato): da riqueza em sentido lato): sense):
lating stylised shocks applied to microeconomic
data available in the Household Finance and Con-
da, seria reescrita como segue (e analogamente a equação (4) para o caso Survey for 2010. More specifically, the
�� �� ��
����
��
�
������
�� �� � ��sumption
1
�� � � ��
1���
����� � ��� � �� �� � ����
m sentido lato): �
�
������������������������������������������� �
�
������������������������������������������� consists in assessing the impact
��� ��
���� �����
�
� ��� stylised
��exercise
� �(5)
��� � � ��
� �� ��� ������ �
��
�����
������ ��
1 � ��
1 � ��
on the
� distribution of net wealth and wealth in
�
��
� �� �� � ��
���
1
�� �� � ��
broad
sense of: (i) a 4 p.p. reduction in the inter� � � �
� � �
������������������������������������������� �
���
� � � � � � � ��� Deste modo, um mesmo aumento percentual nos preços da habitação nos diferentes � � � �
Deste modo, um mesmo aumento percentual nos preços da habitação nos diferentes 1 � ��
est rate on time deposits, (ii) a 4 p.p. increase in
�
the return rate on self-employment businesses,
quintis impactará a distribuição da riqueza líquida em função do aumento percentual quintis impactará a distribuição da riqueza líquida em função do aumento percentual As such, a similar percentage increase in housing
and (iii) a 4 p.p. reduction in interest rates on con um mesmo aumento percentual nos preços da habitação nos diferentes nos preços da habitação de cada quintil, do peso da habitação no total dos ativos de nos preços da habitação de cada quintil, do peso da habitação no total dos ativos de prices across quintiles will impact on net wealth
sumer loans and housing loans.
ctará a distribuição da riqueza líquida em função do aumento percentual cada quintil, e do respetivo grau de alavancagem. cada quintil, e do respetivo grau de alavancagem. distribution according to the share of housing in
This exercise is based on a number of assumpa habitação de cada quintil, do peso da habitação no total dos ativos de total assets and leverage for each quintile.
tions. First, it is once again assumed that no deciAs colunas 2, 3 e 4 do Quadro 3 caracterizam o efeito na riqueza de aumentos de 10% As colunas 2, 3 e 4 do Quadro 3 caracterizam o efeito na riqueza de aumentos de 10% e do respetivo grau de alavancagem. Columns 2, 3 and 4 of Table 3 illustrate the effect
sions have been made on wealth composition.
nos preços da habitação, no valor dos ativos financeiros transacionáveis e no valor dos nos preços da habitação, no valor dos ativos financeiros transacionáveis e no valor dos on wealth of 10% increases in housing prices, on
It only varies due to changes in interest/return
3 e 4 do Quadro 3 caracterizam o efeito na riqueza de aumentos de 10% the value of marketable financial assets and on
negócios por conta própria. Esta análise é particularmente interessante dada a perceção negócios por conta própria. Esta análise é particularmente interessante dada a perceção rates. Second, it takes into account a change in
the value of self-employment businesses. This
a habitação, no valor dos ativos financeiros transacionáveis e no valor dos de que a política monetária — em particular a política monetária não convencional — de que a política monetária — em particular a política monetária não convencional — interest rates from 5% to 1% and a change in
analysis is particularly useful given the perception
conta própria. Esta análise é particularmente interessante dada a perceção return on self-employment businesses from 1%
teve um impacto significativo sobre a valorização destes ativos. teve um impacto significativo sobre a valorização destes ativos. that monetary policy – in particular, non-standard
to 5%, although results are relatively robust to
tica monetária — em particular a política monetária não convencional — monetary policy – had a sizeable impact on the
other rate levels compatible with a similar 4 p.p.
Os resultados do exercício revelam que um aumento nos preços da habitação contribui Os resultados do exercício revelam que um aumento nos preços da habitação contribui increase in value of these assets.
acto significativo sobre a valorização destes ativos. change. Third, it is assumed that, among depospara diminuir a desigualdade da distribuição da riqueza, visto que este ativo tem uma para diminuir a desigualdade da distribuição da riqueza, visto que este ativo tem uma The exercise demonstrates that an increase in
its, only time deposits bear interest. Fourth, the
s do exercício revelam que um aumento nos preços da habitação contribui housing prices helps reduce inequality in wealth
importância relativamente homogénea nos diferentes quintis de riqueza, mas menor no importância relativamente homogénea nos diferentes quintis de riqueza, mas menor no reduction in interest rates affects all time depos-
distribution, given that housing has a relatively
r a desigualdade da distribuição da riqueza, visto que este ativo tem uma caso do quintil mais elevado de riqueza (Gráficos 5 e 6). Por seu turno, um aumento no caso do quintil mais elevado de riqueza (Gráficos 5 e 6). Por seu turno, um aumento no its and housing loans, taking into account a residhomogenous importance across wealth quin-
relativamente homogénea nos diferentes quintis de riqueza, mas menor no ual maturity
of onetipicamente year in deposits
and the fact
valor das ações, e fundos de investimento contribui para valor das ações, obrigações fundos de obrigações investimento contribui tipicamente para tiles,e but
lower
in
the
case of the
highest
wealth
that housing loans typically have a floating rate.
til mais elevado de riqueza (Gráficos 5 e 6). Por seu turno, um aumento no quintile (Charts 5 and 6). In turn, an increase in
aumentar a desigualdade, em linha com a elevada desigualdade na distribuição destes aumentar a desigualdade, em linha com a elevada desigualdade na distribuição destes Fifth, it is assumed that the reduction in inter-
theinvestimento value of shares,
bonds tipicamente and mutual funds
ões, obrigações e fundos ativos, muito concentrados e com um peso relativamente elevado no último quintil da de contribui para ativos, muito concentrados e com um peso relativamente elevado no último quintil da est rates affects 40% of consumer loans, with a
typically contributes to an increase in inequality, in
desigualdade, em linha com a elevada desigualdade na distribuição destes distribuição (Gráficos 7 e 8). A única exceção é o rácio Q5/Q1 na distribuição de riqueza distribuição (Gráficos 7 e 8). A única exceção é o rácio Q5/Q1 na distribuição de riqueza residual average maturity of 2.5 years. Finally,
line with the high inequality in the distribution of
concentrados e com um peso relativamente elevado no último quintil da the
exercise
that time
deposits
these
assets,
which
are greatly
concentrated
and alavancagem líquida, o que decorre do impacto da no elevada no primeiro quintil de mature
líquida, o que decorre do impacto da elevada alavancagem primeiro quintil de assumes
hold a relatively high share in the last distribution
over this period and that interest is reinvested.
Gráficos 7 e 8). A única exceção é o rácio Q5/Q1 na distribuição de riqueza riqueza líquida, tal como atrás explicado (ver Gráfico 2). Analogamente, o contributo dos riqueza líquida, tal como atrás explicado (ver Gráfico 2). Analogamente, o contributo dos quintile (Charts 7 and 8). The sole exception is the
e decorre do impacto da elevada alavancagem no primeiro quintil de Q5/Q1 ratio in net wealth distribution, stemming
15 15 da, tal como atrás explicado (ver Gráfico 2). Analogamente, o contributo dos Similarly, it is assumed that debt maturing over
this period is refinanced, including interest.
Special issue
Table 4 shows results for this exercise, simu-
Q1, Q3 and Q5 quintiles – is lower than in net
lating shocks that will tend to follow an expan-
wealth (Chart 9). Second, an increase in return
sionary monetary policy shock. Its main conclu-
on businesses leads to greater inequality, due
sions are as follows. First, on the asset side, a
to their greater share of the highest wealth (net
decrease in return on time deposits leads to
and in broad sense) quintiles (column 2). Third,
an increase in net wealth inequality (column 1),
a decline in loan rates tends to reduce inequali-
reflecting the combination of greater leverage
ty, given that leverage is relatively greater in the
for poorer households and a relatively similar
lowest wealth quintiles, particularly in the case
share of time deposits in total assets among
of net wealth. This conclusion stands both for
the Q1, Q3 and Q5 net wealth distribution quin-
housing loans and consumer and other loans
tiles (Chart 10). In the case of wealth in broad
(columns 3 and 4). The smaller reduction in
sense, the decrease in return on deposits helps
inequality in the case of wealth in broad sense
reduce inequality. In wealth in broad sense,
is due to the lower dispersion in leverage as
both the dispersion in the share of deposits
well as the fact that loans are more unevenly
in total assets and the dispersion in leverage
distributed across quintiles in this case.
101
across household quintiles – in particular, the
Chart 5 • Relation between real estate and total
assets, by quintiles of wealth in broad sense
Note: Calculations based on the Household Finance and Consumption
Survey for 2010.
Chart 7 • Relation between tradable assets and total
assets, by quintiles of wealth in broad sense
Note: Calculations based on the Household Finance and Consumption
Survey for 2010.
Chart 6 • Relation between real estate and total
assets, by net wealth quintiles
Note: Calculations based on the Household Finance and Consumption
Survey for 2010.
Chart 8 • Relation between tradable assets and total
assets, by net wealth quintiles
Note: Calculations based on the Household Finance and Consumption
Survey for 2010.
102
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Table 4 • Impact of changes in income flows on the distributions of wealth in broad sense
and net wealth | Percentage changes in initial ratios, excluding ratio values
Ratios based on quintiles of wealth in broad sense
Initial ratios
4 pp decrease
in the interest
rate
on deposits
4 pp increase
in the rate
of return
of businesses
4 pp decrease
in the interest
rate on housing
loans
4 pp decrease
in the interest
rate on consumer
loans
(1)
(2)
(3)
(4)
Q5/Q3
3.6
-0.004
0.158
-0.004
-0.002
Q5/Q1
11.9
-0.006
0.175
-0.004
-0.003
Average/Q3
1.4
-0.001
0.083
-0.002
-0.001
4 pp decrease
in the interest
rate
on deposits
4 pp increase
in the rate
of return
of businesses
4 pp decrease
in the interest
rate on housing
loans
4 pp decrease
in the interest
rate on consumer
loans
(1)
(2)
(3)
(4)
Ratios based on net wealth quintiles
Initial ratios
Q5/Q3
6.7
0.041
0.558
-0.042
-0.011
Q5/Q1
412.4
0.702
0.502
-2.242
-1.433
2.0
0.027
0.373
-0.024
-0.005
Average/Q3
Note: Calculations based on HFCS 2010.
3.3. Labour income channel: the redistribution impact on human wealth
economy, based on microeconomic data from
Monetary policy may also have redistribution
and Consumption Survey. The exercise first
effects due to the heterogeneous impact on
labour remuneration developments. This heterogeneity stems from the combination of two
factors: on the one hand, cyclical labour market
transitions are known to differ across demographic groups; on the other hand, the distribution of the various demographic groups is
uneven across wealth (net and in broad sense)
quintiles. Several authors have documented
that expansionary monetary policy disproportionally reduces unemployment for young people and less qualified workers. If these population segments are predominant in the lowest
wealth distribution quintiles, an expansionary
monetary policy will tend to reduce inequality
in the respective wealth distribution.
the Labour Survey and the Household Finance
shows that, comparing the 2011–13 recession
with the recovery phase in 2013–16, a considerable heterogeneity emerges in transitions between
employment and unemployment according to
age and qualification. Afterwards, evidence is presented on how individuals are unevenly distributed across quintiles of wealth in broad sense and
net wealth. Given the two time periods under
review, in this stage, the exercise used data from
the Household Finance and Consumption Survey
for 2013, which provides a better picture of the
2011–16 period. Finally, the exercise looks into
the redistribution impact of a 1% increase in net
employment growth on the Portuguese economy.
Since the first quarter of 2011, there have been
two distinct stages in the Portuguese economy.6 Up to the second quarter of 2013, an
This section analyses the potential importance
unprecedented recession in recent times took
of this channel in the case of the Portuguese
hold. Afterwards, a sustained recovery began,
Special issue
although more gradual compared with previous
more frequent for younger and more qualified
economic upturn periods. The two periods are
individuals. Comparing the two periods under
used as a reference to illustrate heterogeneity in
review, i.e. 2011 Q2-2013 Q2 and 2013 Q3-2016
flows between states in the labour market in the
Q4, the frequency of flows between employ-
Portuguese economy. More specifically, the anal-
ment and unemployment declined by approxi-
ysis focuses on the percentage change in the
mately 25% and the frequency between unem-
frequency of transitions between employment
ployment and employment increased by around
and unemployment, and vice-versa, by break-
18%, both in terms of average quarterly transi-
ing down the labour force into four demographic
groups according to two predetermined characteristics of individuals: age and qualification.
103
tions. According to microeconomic data from the
Labour Survey, the decrease in flows to unemployment was relatively broad-based across age
Unsurprisingly, transitions from employment to
groups and relatively more concentrated in less
unemployment are more frequent for younger
qualified groups, while the increase in flows to
and less qualified individuals. Moreover, transi-
employment was relatively higher in younger
tions from unemployment to employment are
and less qualified groups (Charts 11 and 12).
Chart 9 • Relation between savings deposits
and total assets, by quintiles of wealth in broad sense
Chart 10 • Relation between savings deposits and total
assets, by net wealth quintiles
Note: Calculations based on the Household Finance and Consumption
Survey for 2010.
Chart 11 • Change in the frequency of transitions
from unemployment to employment, between
2011 Q3-2013 Q2 and 2013 Q3-2016 Q4 | In percentage
40.0
35.0
30.0
25.0
20.0
15.0
10.0
5.0
0.0
-5.0
-10.0
Note: Calculations based on the Household Finance and Consumption
Survey for 2010.
Chart 12 • Change in the frequency of transitions
from employment to unemployment, betwee
2011 Q3-2013 Q2 and 2013Q3-2016 Q4
0.0
-5.0
-10.0
-15.0
-20.0
-25.0
-30.0
-35.0
Total
Age < 45 and Age < 45 and
Age >= 45
(working-age
education
education
and <65 and
persons)
below
equal to or
education
secondary
above
below
secondary
secondary
Source: Labour Survey.
Age >= 45
and <65 and
education
equal to or
above
secondary
Total
(working-age
persons)
Age < 45 and Age < 45 and
education
education
below
equal to or
secondary
above
secondary
Source: Labour Survey.
Age >= 45
and <65 and
education
below
secondary
Age >= 45
and <65 and
education
equal to or
above
secondary
104
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Charts 13 and 14 confirm that this heteroge-
data from the Labour Survey and the Household
neity affects differently the various quintiles of
Finance and Consumption Survey for 2013.
the distributions of wealth in broad sense and
The exercise assessed the distribution impact
net wealth. In terms of qualifications, the charts
show that the less qualified labour force is relatively more concentrated in the lowest quintiles
of both distributions. However, in terms of age
groups, the two distributions differ. Indeed, in the
distribution of wealth in broad sense, the share of
individuals under 45 in total labour force is relatively stable across distribution quintiles, while as
regards net wealth distribution the share of individuals under 45 is relatively higher in the lowest distribution quintiles. Furthermore, the share
of individuals aged 65 or more in total population is relatively higher in the lowest quintiles of
the distribution of wealth in broad sense, while
as regards net wealth distribution the opposite
applies, i.e. older age groups are more concentrated in the highest net wealth quintiles.
of a net increase in employment of 1%, taking
into account the circumstances characterising the economic recovery seen since 2013.
Based on the Labour Survey, a 1 p.p. growth
in employment was broken down into the four
aforementioned age groups (for the workingage population), similarly to that seen between
2011 Q2–2013 Q2 and 2013 Q3–2016 Q4. Next,
the assumption was that this growth occurred
regardless of the wealth distribution quintile of
the individuals. Under the assumption that, for
each demographic group in each wealth quintile,
job creation implies an increase in labour income
corresponding to remuneration per capita of
workers in that population segment, it is possible to estimate income changes in each segment
stemming from the 1% aggregate shock on net
The combination of these data suggests that
employment. To take into account the possible
policies contributing to a rise in employment will
existence of unemployment benefits, as well as
tend to affect the distributions of wealth in broad
the fact that income arising from a re-entry to
sense and net wealth differently. To quantify
the labour market is typically below the average,
these effects on the Portuguese economy, a styl-
it was assumed that the increase in income for
ised exercise was conducted, by cross-checking
an individual that finds a job is only 50% of the
Chart 13 • Population structure, by quintiles
of wealth in broad sense
Chart 14 • Population structure, by net wealth
quintiles
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
1.00
0.90
0.80
0.70
0.60
0.50
0.40
0.30
0.20
0.10
0.00
Quintile 1
Quintile 2
Quintile 3
Quintile 4
Quintile 5
Age >= 65
Age >= 45 and <65 and education equal to or above secondary
Age >= 45 and <65 and education below secondary
Age < 45 and education equal to or above secondary
Age < 45 and education below secondary
Note: Calculations based on the Household Finance and Consumption Survey
for 2013.
Quintile 1
Quintile 2
Quintile 3
Quintile 4
Quintile 5
Age >= 65
Age >= 45 and <65 and education equal to or above secondary
Age >= 45 and <65 and education below secondary
Age < 45 and education equal to or above secondary
Age < 45 and education below secondary
Note: Calculations based on the Household Finance and Consumption Survey
for 2013.
Special issue
income per worker calculated for each popula-
monetary policy that causes a continuous reduc-
tion segment. This factor affects the magnitude
tion in the short-term nominal interest rate for a
but not the sign of the results.
long period of time. This monetary impulse leads
Finally, it was assumed that the increase in labour
income fosters both wealth in broad sense and
net wealth. In the case of wealth in broad sense,
monetary policy is assumed to have no permanent real effects and, therefore, labour income
to changes in factors affecting wealth distribution. The ultimate goal is to compare the relative
importance of the various channels to wealth
distribution changes, while keeping in mind the
merely stylised nature of this exercise.
increases only have an impact during the peri-
Although it can be challenging to pinpoint all
od under review. In the case of net wealth, it
mechanisms at play, models can be used in order
7
was assumed that the increase in income is fully
invested in assets, with no changes to the composition of household portfolios. This choice is
not straightforward, but takes into account that
additional income expands the range of options
open to households at the end of the period.
Given the assumptions made, a 1% increase
in net employment helps reduce inequality in
net wealth and, to a lesser extent, inequality in
wealth in broad sense. In the case of net wealth,
the ratio of the top to the bottom quintile (Q5/
Q1) falls by 3.8% and the Q5/Q3 ratio by 0.1%.8
In the case of wealth in broad sense, percentage
decreases in these ratios are lower. The smaller
redistribution impact on wealth in broad sense
results, inter alia, from the fact that: (i) the bottom net wealth distribution quintile is associated
with very low wealth, (ii) a relatively larger share
of population aged over 65 – assumed, in this
exercise, to not be working – is included in the
bottom quintiles of the distribution of wealth in
broad sense, and (iii) younger groups account
for a relatively significant share in the top quintiles of the distribution of wealth in broad sense.
4. Stylised quantification of the distribution
impact of monetary policy in Portugal
after 2010
The previous sections looked into the impact of
stylised changes in prices and other relevant factors on wealth inequality indicators. This section
to rationalise such effects. For instance, in a NeoKeynesian model, an unexpected reduction in the
nominal interest rate leads to a decrease in savings – given that saving has become less appealing – and the resulting immediate increase in
consumption. This direct effect produces a temporary rise in economic activity leading to upward
pressure on labour demand by firms. The resulting increase in wages and employment is an
indirect effect of monetary policy changes, via
general equilibrium mechanisms in this type of
economy. This indirect mechanism will positively affect human wealth of households and, possibly, net wealth too, as a result of the investment
of additional resources in various assets.
A further example is the effect of economic activity on the profits of privately-owned firms. Indeed,
if, during the monetary stimulus period, GDP
grows more rapidly than without the stimulus,
corporate profits will also be greater, thus freeing up additional resources to the households
that own these firms. The accumulated additional resources correspond to an increase in
non-human wealth at the end of the period
under review.
To calculate realistic paths for the main aggregate variables associated with monetary stimulus that may help in the exercise, the PESSOA
model was estimated for the Portuguese economy (see Almeida et al., 2013 and Júlio and Maria,
2017). This stylised simulation calculates the
effect of additional monetary accommodation
aggregates these channels in their entirety as a
on any baseline scenario.
response to monetary policy changes. The start-
To make this exercise more illustrative of the Por-
ing point for this exercise is a stylised change in
tuguese experience in the recent past, monetary
105
106
BANCO DE PORTUGAL • Economic Bulletin • May 2017
accommodation is analysed from the end of
euro area level. The PESSOA model also takes
2010 onwards, immediately after the Household
into account changes in demand for Portuguese
Finance and Consumption Survey for 2010, with
goods and services by euro area countries, given
this stimulus being extended up to the end
that they are also affected by reductions in the
of 2015.
nominal interest rates.
During this period, euro area short-term nom-
A 3 p.p. monetary accommodation leads to a
inal interest rates fell markedly. From 2014
cumulative increase in GDP during the five-year
onwards, standard monetary policy was large-
period of approximately 5.2%, compared to a
ly exhausted given that ECB key interest rates,
scenario where no additional accommodation is
particularly the deposit facility interest rate, were
close to their lower bound.9 Following the introduction in 2015 of non-standard monetary policy measures based on the purchase of sovereign debt securities and other assets, interest rates impacting on households and firms
kept on decreasing. When the nominal interest
rate is close to the lower bound, the concept of
the ‘shadow‛ interest rate becomes useful. In an
economy where no lower bound is applied to
interest rates and non-standard monetary policy measures have not been implemented, this
interest rate would have accommodative monetary effects similar to those in an economy with
an interest rate set at the lower bound and nonstandard monetary policy measures in place.
The calculation of such shadow interest rates
(see, e.g. Wu and Xia, 2017, as well as the references in that study) shows that they continued
to decrease up to late 2016. This exercise gauges the impact of interest rate cuts in a benchmark case, e.g. a less aggressive reaction by the
ECB in terms of monetary accommodation.
implemented. Employment increases by around
3.9% and disposable income by approximately 8.4%. The chart also illustrates developments
in EBITDA as a percentage of assets in private
Portuguese firms. These developments result
from the impact on this corporate profitability
measure of the increase in economic activity and,
in particular, the increase in GDP during the period under review.11 The monetary stimulus leads
to an increase in EBITDA, which, cumulatively
over the period, results in a 1.9% increase in value for firms, under the assumption of a reinvestment of earnings.12 These additional resources
add, therefore, to initial non-human wealth.
Monetary policy also has an indirect impact on the
prices of a number of assets. In particular, macroeconomic developments illustrated in Chart 15
lead to changes in housing prices, which, in turn,
result in a 1.2% increase from the initial situation.13
Table 5 shows the outcome, in terms of the
distributions of wealth in broad sense and net
wealth, of the path followed by a number of variables compared to a benchmark case where
there is no monetary stimulus. The exercise
Chart 15 illustrates possible developments in a
employs the same methods and assumptions
number of aggregate variables of importance to
regarding the increase in value of stocks and
the exercise. The nominal interest rate declined
flows as described in Section 3. Given that this
throughout the five-year period and was nor-
is a five-year period, labour income increases
malised to a 3 p.p. reduction from the initial val-
are accumulated over the five years taking into
ue. This cannot be interpreted as an estimate
account the gradual rise in employment. For
of total monetary accommodation in the course
the sake of simplicity, only the Q5/Q3 and Q5/
of the period, but merely as a quantitative
Q1 ratios are presented.
10
benchmark allowing a comparison of the relative
importance of the various channels analysed in
this article. This accommodation is exogenous
to the Portuguese economy, as it is decided at
Special issue
107
Table 5 • Impact of a prolonged monetary stimulus on wealth distribution | Initial and final
ratios and changes stemming from several components of wealth in broad sense and of net wealth
Ratios based on quintiles of wealth in broad sense
3 pp decrea- 3 pp decrea1,9%
3 pp decrea- se in the
se in the
All changes
1,24%
increase in se in the interest rate interest rate
3,9%
Initial Final
except
increase in the value of interest rate on housing on consu- increase in
ratios ratios All changes employment real estate businesses on deposits
loans
mer loans employment
Q5/Q3 3.636 3.629
-0.20
-0.18
-0.068
0.10
-0.018
-0.18
-0.010
-0.025
Q5/Q1 11.86 11.83
-0.24
-0.19
-0.071
0.11
-0.027
-0.19
-0.014
-0.050
Ratios based on net wealth quintiles
3 pp decrea- 3 pp decrea1,9%
3 pp decrea- se in the
se in the
All changes
1,24%
increase in se in the interest rate interest rate
3,9%
Initial Final
except
increase in the value of interest rate on housing on consu- increase in
ratios ratios All changes employment real estate businesses on deposits
loans
mer loans employment
Q5/Q3 6.707 6.490
-3.2
-2.1
-0.83
0.36
0.18
-1.8
-0.056
-1.1
Q5/Q1 412.4 106.4
-74
-61
-20
0.32
3.1
-50
-6.7
-34
Note: Figures result from a simulated change in the nominal interest rate using the PESSOA model and satellite models estimated for the Portuguese economy. Methodological assumptions for the calculation of the impact of the various simulated changes on the distribution are detailed
in Section 3. Calculations based on the Household Finance and Consumption Survey for 2010 and 2013 and the Labour Survey.
The results indicate that an increase in mone-
from a rise in employment. The effect via hous-
tary accommodation, as illustrated by Chart 15,
ing prices – with housing in Portugal being the
under the various methodological assump-
main form of non-human wealth of households
tions adopted, leads to a sizeable reduction in
and being comparatively well distributed – also
net wealth distribution inequality, chiefly due
plays a key role in that reduction, although the
to cuts in interest rates on housing loans and
effect of monetary accommodation on housing
the increase in resources available stemming
prices is low.
Difference vis-à-vis the absence of stimulus
(percentage points)
10
Chart 15 •
Macroeconomic
developments
in a number
of variables
following monetary
stimulus
8
6
4
2
0
-2
-4
End of 2010
GDP
Nominal interest rate
Employment
EBITDA
End of 2015
Disposable income
House prices
Note: Figures result from a simulated change in the nominal interest rate using the PESSOA model and satellite models estimated
for the Portuguese economy.
108
BANCO DE PORTUGAL • Economic Bulletin • May 2017
In turn, improvements in corporate profitabi-
in asset prices, business profitability and net job
lity stemming from increased GDP growth con-
creation, and declines in interest rates on depos-
tribute to an increase in inequality, given that
its and loans.
households in the top two quintiles benefit the
Distribution effects via an increase in asset pric-
most from this channel. However, households
es depend on leverage and the asset portfolio
in the bottom quintiles also benefit indirectly
composition for each population group. In the
from improvements in corporate profitability due
case of Portugal, it is shown that households in
to a rise in employment, stemming from firms
the top wealth quintiles are, on average, less lev-
distribution inequality.
eraged than those in the bottom quintiles. This
Overall effects on wealth in broad sense are
lower than those in net wealth. Turning to the
impact of labour income, the underlying reasons are analysed in Section 3 and include the
relatively larger share of population aged over
65 in the bottom quintiles of the distribution,
together with a higher share of younger and
more qualified individuals in the top quintiles.
As such, the boost given by monetary policy to
employment is broadly shared across population groups.
The outcome of the exercise is broadly in line
with empirical and quantitative literature on the
impact of monetary policy on net wealth (see
Coibion et al., 2016, Deutsche Bundesbank, 2016
or Furceri et al., 2016). This study shows that, in
the case of Portugal and with certain caveats, the
outcome can also be applied to wealth in broad
sense, although in this case the impact is lower.
5. Conclusions
implies that increases in value of assets tend to
benefit relatively more households in the bottom quintiles, which, in turn, tends to reduce net
wealth inequality.
Leverage also decreases with quintiles of wealth
in broad sense, but less so than in the case of net
wealth. The ratio of human wealth to total assets
also decreases in parallel with wealth quintiles.
As a result, the distribution effect of increases
in the value of assets leads to an increase in inequality in wealth in broad sense.
The data suggest that increases in housing prices
will tend to reduce inequality in net wealth and
wealth in broad sense. In the case of an increase
in self-employment businesses and marketable
financial assets, it will lead to greater distribution
inequality.
Reductions in interest rates tend, in turn, to affect
the accumulation of wealth via changes in income
flows from financial investments and loan instalments. In the case of deposit rates, their reduction leads to a decline in net wealth inequality.
This Special Issue looks into the distribution
Turning to loans, they reduce inequality in the dis-
effects of monetary policy using exercises based
tribution of both types of wealth.
on empirical data on the Portuguese econo-
Net job creation potentially resulting from mon-
my. First, wealth is characterised using House-
etary policy chiefly benefits younger and less
hold Finance and Consumption Survey data, as
qualified workers. This effect tends to reduce net
well as Labour Survey microdata, and the various
wealth inequality and, to a lesser extent, inequal-
channels affecting two types of wealth – wealth
ity of wealth in broad sense, due to differences
in broad sense and net wealth – are analysed. It
in population quintile distribution for both types
is also demonstrated that households in quintiles
of wealth.
of wealth in broad sense are not necessarily those
The factors that help reduce wealth distribution
of the corresponding net wealth quintiles. Exer-
inequality are not separable from those that con-
cises conducted here assess in a stylised man-
tribute to its increase. For instance, the rise in
ner the impact of increased monetary accom-
employment (which helps reduce inequality) is
modation, which will tend to lead to increases
largely due to higher corporate profitability (which
Special issue
contributes to an increase in inequality). There-
distribution of various wealth components by
fore, it is necessary to analyse the total effect of
population groups also leads to a reduction in
the various channels on wealth distribution.
inequality, although lower.
The Special Issue also includes a stylised exercise
As a whole, the results describe monetary policy
where, using various macroeconomic and statis-
as an instrument influencing not only the gener-
tical models adapted to the Portuguese econo-
al state of the economy, but also cyclical devel-
my, all channels are simultaneously simulated.
opments in inequality, through several chan-
Results are similar to those described above,
nels. However, a number of important channels
although the relative magnitude of the various
in terms of the impact of monetary policy on the
channels can now be analysed more realisti-
economy as a whole are not included in the anal-
cally. In the case of net wealth, the total effect
ysis. More detailed research into the distribution
leads to a reduction in inequality. In the case of
impact of these mechanisms would, therefore,
wealth in broad sense, the more homogeneous
be an interesting topic for future study.
References
Adão, B. and I. H. Correia, 2016, ‘Changes in Infla-
Doepke, M. and M. Schneider, 2006, ‘Inflation and
tion Regimes‛, manuscript.
the Redistribution of Wealth‛, Journal of Political
Almeida, V., G. Castro, R. M. Félix, P. Júlio and J. R.
Economy, 114(6): 1069–97.
Maria, 2013, ‘Inside PESSOA – A detailed descrip-
Domanski, D., M. Scatigna and A. Zabai, 2016,
tion of the model‛, Working Paper No 16/2013, Ban-
‘Wealth Inequality and Monetary Policy‛, BIS Quar-
co de Portugal.
terly Review (March).
Auclert, A., 2016, ‘Monetary Policy and the Redistri-
Erosa, A. and G. Ventura, 2002, ‘On Inflation as
bution Channel‛, manuscript, Stanford University.
a Regressive Consumption Tax‛, Journal of Mon-
Banco de Portugal and Statistics Portugal, 2016,
etary Economics, 49(4): 761–795.
‘Household Finance and Consumption Survey for
Fisher, I., 1933, ‘The Debt-Deflation Theory of Great
2013‛, Press Release, 26 October.
Depressions‛, Econometrica Vol. 1, No. 4: 337-357.
Coibion, O., Y. Gorodnichenko, L. Kueng, and J.
Furceri, D., P. Loungani and A. Zdzienicka, 2016,
Silvia, 2016, ‘Innocent Bystanders? Monetary Policy
‘The Effects of Monetary Policy Shocks on Inequal-
and Inequality in the U.S.‛, University of Texas-Austin,
ity‛, IMF Working Paper WP/16/245.
manuscript.
Júlio, P and J. F. Maria, 2017, ‘Output in the Por-
Costa, S., 2016, ‘Financial situation of the house-
tuguese post-2008 period: A general equilibrium
holds in Portugal: an analysis based on the HFCS
narrative‛, Banco de Portugal Economic Studies,
2013‛, Banco de Portugal Economic Studies, 2(4),
3(2), Banco de Portugal.
pp. 15–55, Banco de Portugal.
Wu, J. C. and F. D. Xia, 2017, ‘Time Varying Lower
Costa, S. and L. Farinha, 2012, ‘Inquérito à situa-
Bound of Interest Rates in Europe‛, Working Paper.
ção financeira das famílias: metodologia e principais resultados‛, Occasional Paper No 1-2012
of Banco de Portugal.
Deutsche Bundesbank, 2016, ‘Distributional effects
of monetary policy‛, Monthly Report, September.
109
110
BANCO DE PORTUGAL • Economic Bulletin • May 2017
Notes
1. The 25x multiplier factor applied to labour income estimates the current value of that flow over a long period of time. For instance, a constant income flow
updated at a 3.5% nominal rate over 40 years has a current value of approximately 22 times the value of the flow. In practice, this factor does not influence
the outcome in the case of reasonable values in the update rate and the remaining period for the flow. As such, a 25x multiplier factor was adopted.
2. The quintiles for a given set of elements characterised by an attribute are the five groups such that each comprises a fifth of the whole, grouped in ascending order of the attribute. Therefore, the first wealth quintile for a population of households corresponds to the 20% least wealthy households, the second
quintile corresponds to the 20% of households that come immediately after in terms of wealth, and so on.
3. If households in the intermediate quintile are very differently affected from those in the bottom quintile, the Q5/Q3 and Q5/Q1 ratios may follow different
paths.
4.Furthermore, by definition, the first quintile of the distribution of wealth in broad sense largely coincides with the first income distribution quintile.
Indeed, approximately 75% of households in the first income distribution quintile are also included in the first quintile of the distribution of wealth in broad
sense. By contrast, only one-third of households in the first income distribution quintile is included in the first net wealth distribution quintile. Therefore,
the analysis of redistribution effects on the first quintile of wealth in broad sense is a better proxy for the situation of poor households, as typically defined
within the European Union (i.e. the lack of monetary income, measured in relative terms against average population income).
5. Self-employment businesses are those where a household member owns and actively participates as a worker/manager.
6. In the first quarter of 2011, the Labour Survey underwent major methodological changes.
7. That is, the increase in labour income is not multiplied by 25 in the estimate for wealth in broad sense developments at the end of the period.
8. In the Household Finance and Consumption Survey for 2013, the first net wealth quintile is negative. Calculations assume that the first quintile posts
positive net wealth, such that the Q5/Q1 ratio is close to 412, similarly to the Household Finance and Consumption Survey for 2010 (see Table 1).
9. See the Special Issue entitled ‘Life below zero: monetary policy transmission under negative interest rates‛, December 2016 Economic Bulletin, Banco de Portugal.
10. The path followed by the nominal interest rate is due to the magnitude of the reduction in shadow interest rates between the end of 2010 and the end
of 2015 as implied in Wu and Xia (2017), although other references would result in similar values.
11. The impact of an increase in GDP on EBITDA of private firms was gauged using a model where the dependent variable is EBITDA as a percentage of
assets of Portuguese small and medium-sized firms, selected among the class of autoregressive models of order zero, one and two, estimated in terms of
levels, quarterly differences and year-on-year differences, with up to two explanatory variables of a group that includes: (i) logarithmic values of current GDP,
GDP at constant prices, HICP and employment; (ii) the interest rate on the stocks of loans to non-financial corporations and the unemployment rate. The
specification was selected using the Akaike information criterion. The chosen model implies a 0.37 elasticity of quarterly changes in EBITDA as a percentage
of corporate assets vis-à-vis the year-on-year GDP growth rate.
12. This assumption implies that income flows from self-employment businesses remain constant compared with the initial value.
13. This impact is estimated on the basis of an econometric model that links housing prices in Portugal with variables such as employment, the real interest
rate and household disposable income.