FINANCIAL STABILITY REPORT May 2013
FINANCIAL STABILITY
REPORT
May 2013
EUROSYSTEM
FINANCIAL STABILITY
REPORT
MAY 2013
Lisbon, 2013
www.bportugal.pt
BANCO DE PORTUGAL
Av. Almirante Reis, 71
1150-012 Lisboa
www.bportugal.pt
Edition
Economics and Research Department
Design, printing and distribution
Administrative Services Department
Documentation, Editing and Museum Division
Editing and Publishing Unit
Lisbon, 2013
Number of copies
120
ISSN 1646-2246 (print)
ISSN 2182-0392 (online)
Legal Deposit no. 313475/10
CONTENTS
I. FINANCIAL SYSTEM STABILITY
7
11
19
23
37
41
1. OVERVIEW
2. MACROECONOMIC AND FINANCIAL RISKS
Box 2.1. Banking Union: the establishing of the Single Supervisory Mechanism
and role of the ECB
3. BANKING SYSTEM: ACTIVITY, PROFITABILITY
AND OWN FUNDS ADEQUACY
Box 3.1. Financial situation of the six major groups in the Portuguese banking
system in first quarter 2013
4. CREDIT RISK
59
Box 4.1. Restructured credit: Banco de Portugal’s action and a preliminary
analysis of results
61
Box 4.2. Overview of the house price index produced by the INE
63
Box 4.3. The loan to value ratio in the residential mortgage market in Portugal
65
5. LIQUIDITY RISK
77
6. MARKET RISK
83
Box 6.1. Evolution of residents’ Portuguese public debt portfolios
II. ARTICLES
87
Is there a risk-taking channel of monetary policy in Portugal?
Diana Bonfim, Carla Soares
105
Investment Decisions and Financial Standing of Portuguese
Firms – recent evidence
Luísa Farinha, Pedro Prego
127
Bank interest rates on new loans to non-financial
corporations– one first look at a new set of micro data
Carlos Santos
FINANCIAL SYSTEM
STABILITY
I
OVERVIEW
1
MACROECONOMIC AND FINANCIAL RISKS
2
BANKING SYSTEM: ACTIVITY, PROFITABILITY
AND OWN FUNDS ADEQUACY
3
CREDIT RISK
4
LIQUIDITY RISK
5
MARKET RISK
6
1
After two years since the beginning of the economic and financial adjustment programme, the Portu-
7
guese economy has made significant progress in closing the external imbalances and in the reduction
of its structural primary fiscal deficit, against the background of a strong downturn in economic
activity and an increase in unemployment. In this period, an adjustment of the banking system’s
balance sheet was observed, with a significant improvement of the liquidity situation, a strengthening of solvency and a reduction of leverage, although there are still major challenges regarding the
prospective evolution of profitability (Chart 1.1).
Tensions in the international financial markets have been easing since mid 2012, as a result of the
accommodative monetary policy in the main advanced economies, including the euro area (Chart
1.2). In the latter case, ECB’s non-conventional monetary policy measures had a significant impact
on the conditions in the public debt markets of countries under pressure, although fragmentation
of credit markets still persists.
In this context, in conjunction with international investors’ acknowledgement of the results of fiscal
consolidation already achieved, the perception of risk attached to Portuguese sovereign debt among
market operators improved throughout 2012. This is visible in the very substantial reduction of
yields in the Portuguese secondary public debt market. This situation culminated, already in 2013, in
issurance of public debt by the Portuguese Republic with 5 and 10 years maturity, the first since the
inception of the economic and financial adjustment programme. These developments had also been
punctuated by Portuguese banks’ issues of debt, starting from the last quarter of 2012, an indication
of improved risk perception of the sector by international investors. However, as the issues in question
involved small amounts and still carried a relatively high cost, access to the international markets,
both by the banks and the Portuguese state cannot as yet be considered normalized.
In 2012, the Portuguese economy registered a sharp drop in economic activity and a significant
increase in unemployment, in a recessionary setting in the euro area, a slowdown of global economic
activity, a contractionary fiscal policy and monetary and financial conditions, which remained restrictive, notwithstanding signs of a certain decompression at the end of the year.
In 2012, for the first time since the inception of the euro area, the non-financial private sector as a
whole achieved financing capacity, i.e., savings were higher than investment. In the case of households,
there was a significant increase in savings, in line with a downward revision of expectations on future
income flows, and a reduction of investment. From a financial viewpoint, this evolution was translated
essentially in net repayments of debt, both in the case of housing credit and for consumption and
other purposes, as the accumulated flow of financial assets was comparatively less expressive for
the year as a whole. A contraction of investment and an increase in savings were noted also in the
case of non-financial corporations, resulting in a substantial reduction of borrowing requirements.
However, debt levels remained very high in the case of households and non-financial corporations,
both from a historical viewpoint and in comparison to other countries in the euro area, particularly
non-financial corporations, whose debt continued to increase as a percentage of GDP. In this latter
sector, the high indebtedness, in conjunction with a low profitability setting, shows up in a low ratio
between “earnings before interest, taxes and depreciation” and interest payable of indebted firms,
when compared with the remaining economies of the euro area. In particular, it should be highlighted the situation of those sectors of activity which are more reliant on the domestic market such
as construction, real estate development and trade, with more difficult operating conditions and
higher loan default levels than the remaining, which make more difficult the necessary deleveraging
Overview
Overview
process. Reference should also be made to the fact that although a highly relevant proportion of
smaller companies is debt-free, those which are indebted have very high debt levels and, among
I
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
8
these, firms in the construction sector are facing highly restrictive bank financing conditions. This
situation reflects the risk associated with their business and their high levels of indebtedness. Having
these fragilities in mind, in 2012, Banco de Portugal carried on a transversal inspection of the quality
of the banks’ exposures to construction and real estate development companies. In the remaining
sectors it is possible that the less stringent perception of risk attached to the state and the banks in
the international markets, noted since mid 2012, is gradually translating into an easing of conditions
in access to credit.
The economic and financial situation that Portugal is going through raises remarkable challenges
to banks’ current and prospective profitability. The protracted recession, with a highly significant
contraction of domestic demand and a substantial increase in unemployment, is underlying higher
levels of default in the credit portfolio and the consequent need to set up or increase impairments.
This deterioration is particularly marked in the case of credit to non-financial corporations, while in the
case of households, it has been considerably mitigated by the decline in indexants of loans for house
purchase and the ensuing reduction in interest rates to historical lows. Net interest income is also
under downward pressure as a result of several factors, namely the low profitability of the residential
mortgages portfolio, contracted in the past, when put into perspective against the current funding
costs; the persistence of high funding costs, namely in the deposit base and in hybrid instruments; and
the low level of short term interest rates, which compresses the margin associated to sight deposits.
Bank profitability is also conditioned by the need to readjust commercial structures, which have to
deal with lower demand for financial services over the medium term. In such an environment, the
banks have been exploiting market opportunities to profit from non-recurring operations in the form
of repurchases of their own securities starting 2011 and, in second half 2012, the sale of Portuguese
public debt securities with highly relevant capital gains. In turn, international activity, which in past
years more than compensated the downward pressure on the profitability of domestic operations,
has been making a progressively smaller contribution to the banks’ profits.
The monitoring of developments of the delinquency in the loan portfolio is a particularly important
concern at the current juncture. In this regard, it is worth emphasizing that, although credit in default
and non-performing credit are climbing to successively higher levels, in particular in the case of loans
to non-financial corporations and to households with purposes other than for the acquisition of
housing, new episodes of default are stabilizing, when assessed in terms of flows. This stabilization
is visible both in the household and corporate sectors. The moderation in the flow of new defaults
has been occurring, in spite of the persistence of vulnerabilities in the non-financial private sector,
particularly in the case of non-financial corporations.
Also as regards defaults, reference should be made to the housing credit segment whose default
continue to be highly contained; though the current level is benefiting from the low level of interest
rates, they are much lower than those suggested by the analytic instruments available, given their
determinants, particularly the unemployment rate. This fact may indicate that the renegotiation of
debts to adjust mortgage repayment plans to the income of households with debt servicing difficulties
is becoming more frequent. In this respect, it should be noted that, notwithstanding the fact that
the debt of households in lower income brackets is high in comparison to other countries in the euro
area, the proportion of households in these strata of the population which participate in the credit
market is relatively reduced. However, credit risk management must deal with a significant volume of
housing loans with a relatively high loan-to-value ratio, namely in the default recovery process. This
is especially relevant if it is taken into account that there is evidence that house prices have fallen
by more than 10 per cent since the inception of the economic and financial assistance programme.
The unfavourable outlook for the labour market and household disposable income do not allow the
exclusion of the possibility of further reduction in house prices.
Portuguese banks have improved their liquidity indicators and maintained comfortable collateral
buffers, together with a reduction of Eurosystem financing, in the case of domestic institutions. This
resilience since the onset of the financial crisis, even taking into account that recent developments in
capital markets have been favourable to the investment of savings in alternative products. Reference
should also be made to the very pronounced reduction in interest rates on new deposits, following
the very high levels attained in the second half of 2011, which motivated prudential initiatives from
Banco de Portugal, aiming at curbing agressive practices susceptible of harming the whole system. In
turn, the continuation of the deleveraging process of the resident non-financial private sector was,
once again, visible in a reduction of bank lending while also contributing to the structural adjustment
of banks’ balance sheets to a more sustainable financing structure, which is less sensitive to changes
in international investors’ perceptions of risk. Capitalisation operations by institutions in the first half
of 2012 have also improved the system’s loss absorption capacity and reinforced its financial strength.
Against the background of increasingly sustained signs that there will be a prolongation throughout
2013 of the ongoing recession in the euro area, the renewed reduction of activity in the Portuguese
economy expected to take place in 2013, accompanied by a contraction of demand for Portuguese
companies’ goods and services and a deterioration of conditions in the labour market, suggest
that credit risk will continue to materialise and that banks will need to maintain a prudent policy to
recognise the ensuing losses. In addition, the reduction of the contribution of international activity
and the downward pressures to net interest income are particularly demanding in term of banks’
balance sheet management.
Chart 1.1
GLOBAL EVOLUTION OF RISKS IN THE PORTUGUESE BANKING SYSTEM
Solvency
Credit risk
Profitability
Liquidity risk
Dec-07
Market risk
Dec-11
Dec-12
Source: Banco de Portugal.
Note: A value away from the center signifies higher risk, having as reference the historical values for each series used in the different
risk dimensions. For further details on the methodology see “Box 1.1 Financial stability map”, Banco de Portugal, Financial Stability
Report - November 2011.
1
9
Overview
evolution has occurred in a context in which household deposits continue to display a significant
Chart 1.2
I
GLOBAL EVOLUTION OF THE MACROEONOMIC AND FINANCIAL ENVIRONMENT OF THE PORTUGUESE
BANKING SYSTEM
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
10
Households
Monetary and
financial
conditions
Non-financial
corporations
Domestic
macroeconomic
framework (risks)
Global financial
markets
Mercados
financeiros
domésticos
Dec-07
Dec-11
Global
macroeconomic
framework (risks)
Jun-12
Dec-12
Sources: Barclays Capital, Confidencial Imobiliário, European Commission, Eurostat, iBoxx, IMF, INE, Thomson Reuters and Banco
de Portugal.
Note: A value away from the center signifies higher risks or tighter monetary and financial conditions, having as reference the
historical values for each series used in the different risk dimensions. For further details on the methodology see “Box 1.1 Financial
stability map”, Banco de Portugal, Financial Stability Report - November 2011. Given the lack of availability of national accounts data
for September 2012, it was assumed that values remained unchanged vis-à-vis June.
2
The interaction between negative developments in the real economy and disturbances in financial
11
markets, most notably in what concerns the credit market, comprises the main source of risk to financial
stability in Portugal.
The contraction of the national economy, against the background of a recession at the euro area level,
negatively affect the state’s and private sector’s financial position, with an adverse impact on the profitability and quality of banks’, insurance companies’ and other financial agents’ assets. These developments
tend to lead to an increase in risks perceived by banks, with a negative impact on the supply of credit.
In parallel, the low levels of confidence exhibited by economic agents and high uncertainty over the
medium and long term developments in economic activity negatively affect consumption and investment
decisions, with a negative impact on the supply of credit.
These factors, associated with developments in macroeconomic activity, interact with risks arising from
financial markets where, despite a noteworthy decrease in tensions and in tail-risk, signs of fragility remain,
in parallel with significant heterogeneity across the euro area. In particular, financial markets remain
vulnerable to potential setbacks in the path towards the establishment of a true banking union in the
euro area and the resolution of the European sovereign debt crisis. In parallel, the working environment
of banks in Portugal and other countries under stress remains hindered by relatively high funding costs
and rigidity in balance sheet adjustment. These issues hamper the provision of financing to the economy
and translate into additional hurdles for economic recovery against a background in which important
challenges remain in the framework of the Economic and Financial Assistance Programme.
Downwards revision of economic growth prospects
According to the IMF’s April projections, world economic growth will, in 2013, be similar to 2012,
comprising a downwards revision of the preceding forecasts (Chart 2.1). Reference should also be made
to heterogeneity between the performance of the diverse economies, with which different types of
risks are associated. The US is likely to continue to post higher growth rates than most of the advanced
economies, in a context in which demand has been stimulated by monetary policy, particularly in the
housing sector and the consumption of durable goods. Japan is also likely to post relatively high growth
rates, following its recent announcement of a monetary and fiscal stimulus package, accompanied by
structural reforms. These measures will continue to contribute towards the yen’s depreciation in addition
to higher inflation in Japan, translating into relatively unappealing long term returns on its sovereign debt
securities, which could help boost demand for assets in the euro area and United States. Notwithstanding
the importance of these measures to short term economic recovery, it is also necessary to define fiscal
consolidation plans over the medium term so as to avoid uncertainty over the sustainability of the public
finances in the US and Japan. Reference should also be made to the fact that, despite the likelihood of
lower oil prices, there are risks attached to these projections, most notably due to geopolitical tensions.
Projections for the euro area point to an economic contraction in 2013 and low growth in 2014, in
a context of high and rising unemployment rates in most countries. Notwithstanding the significant
decrease in perceived redenomination risk in the euro area since the second half of 2012, financial
markets remain fragmented, with high risk premia in economies under pressure and uncertainty over
the implementation of mechanisms allowing greater financial, economic and fiscal integration in order
to mitigate the interaction between sovereign risk and financial stability. Projections for emerging market
and developing economies indicate robust growth levels, albeit lower than in the preceding forecasts.
Owing to the slowdown in external demand, developments in commodity prices comprise a source of
Macroeconomic and Financial Risks
2. Macroeconomic and financial risks
Chart 2.1
7
12
6
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
EVOLUTION OF GDP GROWTH FORECASTS FOR 2013
I
5
Jul.2012
Oct.2012
Jan. 2013
Apr. 2013
5.3
Per cent
4
3.3
3
1.9
2
1.2
1
0
World GDP
-1
Advanced United States
economies
Euro area
-0.3
Emerging
market
economies
Source: IMF.
Note: The April 2013 forecasts do not include data for Cyprus.
risk for some of these economies, which are exporters of these goods. Further, against the background
of accommodating economic policies in most emerging market and developing economies there is a
risk that the share of the slowdown of growth associated with cyclical as opposed to structural factors
is being overestimated, resulting in pro-cyclical policies. In several such economies, there is evidence that
firms are leveraging, partly through foreign currency denominated debt, which represents a vulnerability
in the case of adjustments in capital flows.
Projections for economic activity in Portugal, published in the Spring Economic Bulletin, were also revised
downwards and indicate a 2.3 per cent drop in GDP, in 2013. This evolution arises in the context of
the macroeconomic imbalances adjustment process which translates into a strong contraction of public
and private domestic demand. Export growth, in turn, is forecast to remain positive, notwithstanding
the prospects of a downturn in external demand. It should, however, be noted that these projections
are subject to risks attached to the evolution of the international environment, with a potential impact
on exports and employment. Such risks are associated with the synchronisation of the fiscal consolidation process in Europe and uncertainty over the implementation of sovereign debt crisis management
mechanisms in the euro area.
Fragmentation remains across euro area countries, notwithstanding the diminishing of
tensions in international financial markets
There was a noticeable decrease in perceived risk during the course of 2012, particularly starting from
the middle of the year, translating into a decline of tensions in the international financial markets (Charts
2.2 and 2.3). The situation, however, remains unstable, with episodes of resurfacing of tensions following
calmer periods. Important measures were announced by the ECB and European Union for the purpose
of mitigating the interaction mechanisms between the banking system and sovereign risk, following
greater pressure on Spanish and Italian sovereign debt in second quarter 2012. In particular, the announcement of the Outright Monetary Transactions programme and single supervisory mechanism enabled
a decrease in the risk of extreme events, namely redenomination risk in the euro area, translating into a
decline in risk premia associated with public and private debt. An increase in tension following the result
of the Italian elections was, however, noted and more recently, at the time of the negotiations for the
Economic and Financial Assistance Programme for Cyprus. In this context, owing to uncertainty over the
Chart 2.2
Chart 2.3
GLOBAL RISK APPETITE INDICATOR
FINANCIAL MARKET DEVELOPMENTS
2
7
6
iTraxx Financials Senior 5Y (p.b.)
Euphoria
iTraxx Europe 5Y (p.b.)
5
13
3
Macroeconomic and Financial Risks
4
10 year Govt. bonds euro (p.p.)
2
10 year Govt. bonds PT (p.p.)
1
0
-1
VIX (%)
-2
-3
Panic
-4
PSI Financials (%)
-5
PSI Geral (%)
-6
Dow Jones Euro Stoxx Banks (%)
-7
-8
Sep-08
Change between end2011 and end-2012
Change between end2012 and 10 May 2013
Dow Jones Euro Stoxx (%)
Jun-09
Mar-10
Dec-10
Sep-11
Jun-12
Source: Credit Suisse.
Mar-13
-140.0
-100.0
-60.0
-20.0
20.0
Source: Thomson Reuters.
effective implementation of the announced measures, several doubts remain over the sustainability of
the reduction of tensions in the financial markets. Reference should, herein, be made to the need for the
implementation of common resolution mechanisms and deposit guarantee schemes, in addition to the
single supervisory mechanism, moving towards a true banking union and helping to curb interactions
between sovereign and financial system risk (see “Box 2.1 Banking Union: the establishing of the single
supervisory mechanism and the role of the ECB”, of this Report).
In Portugal, this evolution was reflected in the return of several banks to the wholesale debt markets,
albeit still subject to restrictive conditions. Portugal has also resumed sovereign debt market operations
over the longer maturities having, in 2013, successfully issued debt with maturities of 5 and 10 years, with
high demand from international investors. The fact that these issues took place in a context in which the
rating on the Portuguese Republic remained below investment grade, comprised an additional difficulty
on account of the fact that the investment policy of several institutional investors, whose investments
tend to be more stable, is limited to assets with higher ratings.
There was a slight improvement in Portuguese banks’ distance-to-default, estimated on the basis of a
systemic risk indicator which includes information on banks’ balance sheets and market variables, over
the end of 2011 currently standing at levels comparable to the average noted since the onset of the
crisis1 (Chart 2.4). The evolution of this indicator, in the case of the euro area and more markedly so in
the case of the US, has been more positive, as a sign of international investors’ greater confidence in
those financial systems (Charts 2.5 and 2.6).
Notwithstanding this positive trend, the fact that financial markets remain fragmented has translated
into disturbances in the monetary policy transmission mechanism in the euro area.2 The disparities in
access to credit and in risk premia charged on issuers in different euro area member states is higher than
expectable on the basis of economic fundamentals (Chart 2.7).3 In addition to the maintenance of significant restrictions on banks’ access to the wholesale debt markets, considerable signs of segmentation in
the interbank markets also continue to be noted. This development arises in a framework in which the
1 The methodology underlying this indicator is described in Saldías (2012), “Systemic Risk Analysis Using Forward-Looking Distance-to-Default Series”, Banco de Portugal, Working Paper 16/2012.
2 For further details, see “Policy Issue Transmission of monetary policy in the euro area”, Banco de Portugal, Economic Bulletin – October 2012.
3 See “Box 2.2 Differences in monetary policy transmission between countries in the euro area”, Banco de Portugal, Annual Report 2012.
I
Chart 2.4
Chart 2.5
DISTANCE-TO-DEFAULT FOR PORTUGUESE
BANKS
DISTANCE-TO-DEFAULT FOR EURO AREA BANKS
13
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
14
14
11
Difference
Portfolio distance-to-default
Average distance-to-default
12
11
Difference
Portfolio distance-to-default
Average distance to default
10
9
10
8
9
7
8
6
7
5
6
5
4
4
3
3
2
2
1
1
0
0
-1
Sep-02
Mar-04
Sep-05
Mar-07
Sep-08
Mar-10
Sep-11
Mar-13
-1
Sep-02
Mar-04
Sep-05
Mar-07
Sep-08
Mar-10
Sep-11
Mar-13
Sources: Bankscope, Bloomberg, Thomson Reuters and Banco
de Portugal calculations.
Sources: Bankscope, Bloomberg, Thomson Reuters and Banco
de Portugal calculations.
Notes: Monthly averages based on daily observations. For a
definition of “Portfolio distance-to-default” and “average
distance-to-default” see Saldías (2012), “Systemic Risk Analysis
Using Forward-Looking Distance-to-Default Series”, Banco de
Portugal, Working Paper 16/2012.
Notes: Monthly averages based on daily observations. For a
definition of “Portfolio distance-to-default” and “average
distance-to-default” see Saldías (2012), “Systemic Risk Analysis
Using Forward-Looking Distance-to-Default Series”, Banco de
Portugal, Working Paper 16/2012.
prospects of low bank profitability, together with the maintenance of a sentiment of uncertainty and lack
of confidence, have translated into low prices in equity markets, increasing the difficulty of attracting
capital (Chart 2.8). Further, the convergence to the new regulatory framework set out in Basel III, along
with pressure from financial market participants, provides incentives towards a decline in bank leverage.
Therefore, albeit mitigated by the ECB’s provision of liquidity under favourable terms and the availability
Chart 2.6
Chart 2.7
DISTANCE-TO-DEFAULT FOR US BANKS
SPREADS VIS-À-VIS GERMANY OF EUROPEAN
GOVERNMENT BOND YIELDS (10 YEARS)
Difference
Portfolio distance-to-default
Average distance-to-default
11
10
5000
4500
9
4000
8
3500
Basis Points
7
6
5
4
3
3000
Austria
Belgium
France
Greece
Ireland
Italy
Portugal
Spain
United Kingdom
2500
2000
1500
2
1000
1
500
0
-1
Sep-02
Mar-04
Sep-05
Mar-07
Sep-08
Mar-10
Sep-11
Mar-13
Sources: Bankscope, Bloomberg, Thomson Reuters and Banco
de Portugal calculations.
Notes: Monthly averages based on daily observations. For a
definition of “Portfolio distance-to-default” and “average
distance-to-default” see Saldías (2012), “Systemic Risk Analysis
Using Forward-Looking Distance-to-Default Series”, Banco de
Portugal, Working Paper 16/2012.
0
Sep-08
Jun-09
Mar-10
Source: Thomson Reuters.
Dec-10
Sep-11
Jun-12
Mar-13
Chart 2.8
STOCK MARKET INDICES
180
Índex 100 = 31/12/2010
160
140
120
100
80
60
40
20
Sep-08
Jun-09
Mar-10
Dec-10
Sep-11
Jun-12
Mar-13
Source: Thomson Reuters.
of public capital to guarantee the resilience of the banks, these factors have contributed towards a
decline in the level of financial integration, incentivising financial institutions to focus their activity on
the national market. In addition to constraints on the banks’ access to funding, restrictions on access
to credit of other economic agents in countries under pressure, both in terms of price and quantity, are
also partly justified by expectations of a greater materialisation of credit risk, owing to weaker economic
growth prospects. These restrictions hinder economic recovery, although the contraction of economic
activity also contributes to lower demand for credit. In addition, in the case of several countries, including Portugal, the need for the gradual, orderly deleveraging of firms is also relevant, allowing for a
convergence towards a more sustainable funding structure.
Potential emergence, over the medium term, of risks deriving from a prolonged environment
of low interest rates in several economies
In economies in which accommodative monetary policy and increase in demand for financial assets with
lower perceived risk translate into historically low long term interest rates, there is some evidence of the
emergence of search for yield phenomena. These can translate into an accumulation of risks over the
medium term, to the extent that they result on an under pricing of risk, leading to greater investment
in higher risk assets and increased leverage. A prolonged environment of reduced interest rates also
exerts pressure on banks’ net interest income. In the case of insurance companies and pension funds,
low interest rates are a particularly relevant source of risk for financial insurance with a guaranteed rate
of return and for defined benefit pension plans, insofar as they exert pressure on returns on assets and
increase the current value of their liabilities. Therefore, although an accommodative monetary policy,
including non-conventional measures, plays a fundamental role in the response to the economic and
financial crisis, permitting an increase in the confidence of participants in financial markets and supporting economic activity, the emergence of risks associated with low interest rates in several countries
must be monitored in the form of an adequate macroprudential policy with a view to mitigating their
materialisation, simultaneously taking into account the need to avoid incentives for regulatory arbitrage.
Further, to the extent that these risks are associated with fragmentation in financial markets in the euro
area, their mitigation also requires coordinated policy responses between member states, in addition
to the development of a clearly designed and sustainable structural framework on a European level, in
order to strengthen convergence to a true single market.
2
15
Macroeconomic and Financial Risks
S&P 500
S&P Banks
DJ Eurostoxx
DJ Eurostoxx Banks
PSI Geral
PSI Financial
200
In the US, an increase in activity and compression of spreads in high-yield bond markets have been
noted, accompanied by signs of the re-emergence activities which were common before the crisis, such
I
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
16
as high risk loans and securities collateralised by opaque assets, suggesting an increase in investors’ risk
appetite. Volatility indices have also been declining, reducing risk valuation measures and consequently
regulatory capital requirements on several market activities. This evolution, which translates into a lower
cost of risk-taking in terms of capital, should be closely monitored by authorities so as to avoid an under
pricing of risk. It is relevant to note that, in historical terms, volatility indices tend to be low for prolonged
periods and are occasionally interspersed by sharp rises in volatility.
Continuing level of uncertainty over the quality of banks’ and insurance companies’ assets
The negative prospects for economic activity which, in several economies, arise together with significant
declines in asset values, notably in the real estate sector, will tend to contribute towards a decline in
the quality of banks’ assets, which may incentivise the restructuring of loans with a high probability of
default. This practice is desirable in cases in which it results in balanced contracts enabling debtors to
align their liabilities with their capacity to make payment. However, in cases in which a debtor’s payment
capacity remains low, it corresponds to a delay in the recognition of losses, with a potential impact on
the transparency of the quality of banks’ assets and, consequently, contributing towards maintaining
uncertainty in the markets. In Portugal, the quality of banks’ assets has been monitored by a series of
inspection programmes, contributing to increase confidence in the banking system. In particular, in the
context of the Economic and Financial Assistance Programme, Banco de Portugal has in 2011 developed a
special inspections programme on the valuation of credit portfolios, the calculation of own funds for credit
risk purposes and the assessment of the methodologies and parameters used in stress testing exercises.4
Banco de Portugal has also developed an initiative to ensure the correct identification of restructured
credit, set out in “Box 4.1 Banco de Portugal’s involvement in the domain of restructured credit”, of this
Report. Lastly an inspection programme on financial institutions’ exposure to the construction and real
estate sectors was also developed in 2012 and its main conclusions are set out in Banco de Portugal’s
press release of 3 December 2012.5
The implementation of the Economic and Financial Assistance Programme and return to the
international funding markets pose fundamental challenges for the Portuguese economy
Notwithstanding the progress achieved in the implementation of the Economic and Financial Assistance
Programme, important structural imbalances and vulnerabilities still need to be corrected in order to
ensure the sustainability of the public finances, contributing towards greater economic competitiveness
and ensuring the stability of the financial system. The implementation of the Programme has led to a
marked correction of the imbalance on the current and capital account, in the framework of strong
contraction of the economy and employment. Against this background, notwithstanding the adjustment
to the consolidation goals defined under the scope of the Economic and Financial Assistance Programme
and extending of the loan repayment period, the weakening of the macroeconomic outlook in Portugal
and in a large number of Portugal’s main export countries, represents a risk to the furtherance of the
adjustment objectives. Uncertainty also remains over the concrete measures to be taken in the future
and their impact on economic growth. There also continues to be a risk of contagion of any additional
difficulties in other euro area economies on the assessment of the risk attached to Portuguese issuers
in the international markets, which could hinder a sustained return to financing from such markets. In
4 The main results of these inspections have been set out in “Box 4.3 The special inspections programme for the
financial system (SIP)”, Banco de Portugal, Financial Stability Report – May 2012.
5 This press release is available at: http://www.bportugal.pt/en-US/OBancoeoEurosistema/ComunicadoseNotasdeInformacao/Lists/LinksLitsItemFolder/Attachments/130/combp20121203_en.pdf.
this context, the strengthening of the process of the construction of the banking union and the implementation of the Outright Monetary Transactions programme should play a relevant role in reducing
volatility, thus contributing to the sustainability of Portugal’s return to the international debt markets.
2
17
Macroeconomic and Financial Risks
fragmentation in the financial markets, increasing investor confidence and avoiding situations of excessive
BOX 2.1 | BANKING UNION: THE ESTABLISHING OF THE SINGLE
SUPERVISORY MECHANISM AND ROLE OF THE ECB
was not accompanied by the full integration of banking system supervision or the mechanisms of the
safety net to cope with possible failures in the banking sector (deposit guarantee systems, institutions’
recapitalisation and resolution mechanisms), which retained a national base.
The sovereign debt crisis in the euro area brought about investors concern with the macroeconomic
imbalances in euro area economies, the inconsistencies and fragilities of the single currency’s governance
model and the bidirectional correlation between sovereign and banking sector risk.
This situation had major implications for the euro area’s institutional and governance model. Firstly, and
following the conclusions of the European Council and the Declaration of the Euro Area Summit of June
2012, the European Commission presented, in September of 2012, a package of legislative proposals
designed to create a Single Supervisory Mechanism, accompanied by a roadmap for establishing the
Banking Union, delineating the steps to be taken to complement supervisory aspects, namely as regards
crises management, bank resolution and deposit guarantee system.
In December 2012, the Council of the European Union (EU) reached an agreement on the legislative
package submitted by the Commission and negotiations began with the European Parliament. Pursuant
to the agreement reached between the Council and the European Parliament on 19 March 2013, both
Regulations are expected to come into effect over the next few months.
The ECB will be responsible for the prudential supervision of credit institutions in the euro area, and
shall perform such tasks under the scope of a Single Supervisory Mechanism comprising the ECB and
the competent national authorities, with the objective of “contributing to the safety and strength of
credit institutions and the stability of the financial system in the EU and in each member state, having
duly considered the uniqueness and integrity of the internal market”. The Regulation provides for the
credit institutions of other member states to also be covered by the supervision of the ECB, against the
background of enhanced cooperation mechanisms to be established with the respective authorities.
The ECB has been entrusted with the following specific tasks:
- To authorise credit institutions and withdraw existing authorisations;
- To assess qualified participations;
- To ensure compliance with the prudential requirements established in EU legislation;
- To assess the adequacy of procedures, strategies and institutions’ own funds vis-à-vis the risks
incurred and to perform stress tests and on the basis of the assessment to impose additional
specific own funds requirements or other requirements provided for in EU legislation;
- To carry out supervision on a consolidated basis and participate in colleges of supervisors;
- To participate in the supplementary supervision of financial conglomerates, when applicable;
- To carry out supervisory tasks in the sphere of recovery plans and early intervention measures in
situations of non-compliance, or risks of institutions’ non-compliance, with prudential requirements.
The agreed text establishes a supervisory model in which the ECB assumes responsibility for the effective
and consistent operation of the integrated supervisory system. There will, however, be an articulated
distribution of tasks between the ECB and national supervisory authorities based on the size of the
banks and national banking systems. In this model, the ECB is responsible for the direct supervision of
the most significant credit institutions, with national supervisory authorities having responsibility for the
direct supervision of all of the others.
2
19
Macroeconomic and Financial Risks
The financial and monetary integration following the creation of the single market and single currency
The significance of institutions is based on several criteria defined in the Regulation, including quantitative
criteria (total assets of more than €30 billion, in consolidated terms, or a proportion of more than 20
percent of GDP, unless total assets are less than €5 billion).
I
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
20
As regards the most significant institutions, the ECB assumes direct responsibility for the supervisory
decisions relating to the tasks listed in the Regulation, with the national supervisory authorities being
responsible for assisting the ECB in its performance of such tasks and being closely associated with the
supervision process.
As regards the less significant institutions, the supervisory decisions are the responsibility of the national
authorities, but are subject to a common framework and general instructions to be defined by the ECB,
which may take upon itself, at any time, direct responsibility for a less significant institution when this is
considered necessary to ensure the application of rigorous, consistent supervision standards.
As regards macroprudential tasks and instruments, the national authorities have a more preponderant
role, although the ECB may apply higher requirements or more stringent measures (pursuant to the terms
of EU legislation), with the reciprocal obligation of issuing ex-ante notifications, both by the national
authorities and the ECB, regarding the measures they intend to adopt.
Reference should be made to the fact that all tasks which were not allocated to the ECB shall be the
responsibility of the national supervisors (e.g. anti-money laundering and consumer protection).
The ECB will have all the powers that Union law gives to the competent authorities in supervisory matters,
including the power to apply sanctions.
After the coming into effect of the Regulation, the ECB may request general and specific information on all
of the credit institutions covered by the Single Supervisory Mechanism and shall carry out a comprehensive
assessment, including a balance-sheet assessment, at least as regards the most significant institutions.
The supervisory function shall be independently performed by the ECB, based solely on the objectives
established in the Regulation, without prejudice to and separately from its tasks relating to monetary
policy. A Supervisory Board will be set up, comprising:
- a Chair, who shall be recognised standing and experience in the banking and finance matters
(who shall not be a member of the Governing Council of the ECB);
- a Vice Chair, to be chosen from the members of the Executive Board of the ECB;
- four ECB representatives of the ECB appointed by the Governing Council of the ECB;
- one representative of the national authority competent for the supervision of credit institutions
in each participating Member State.
The Supervisory Board shall be responsible for the planning and execution of the supervisory tasks
entrusted to the ECB and shall also prepare the draft decisions to be submitted to the Governing Council
of the ECB for approval.
The ECB may charge institutions fees (proportional to their relevance and risk profile) to finance its
expenditure on supervisory tasks.
The ECB is accountable to the European Parliament and the Council for the implementation of the
Regulation.
The ECB shall fully assume the tasks conferred on it by the Regulation one year after the entry into force of
the Regulation, unless the ECB decides to postpone the date, for reasons of the system’s operationalisation.
Work for the purposes of facilitating preparation for the coming into effect of the SSM within the
ECB has already started, particularly: (i) development of the regulatory framework; (ii) categorising of
institutions in accordance with their systemic relevance both on a level of the euro area and nationally;
(iii) establishing of information transmission mechanisms; and, (iv) scheduling of supervisory tasks and
decisions, with the objective of defining criteria to establish an efficient system for the delegation of
competence between central and national levels and adequate forms of articulation between them.
of the euro area, leveraging and increasing the efficiency of monetary policy while, at the same time,
preventing the fragmentation of the European banking sector and contamination between sovereign
and banking risk.
In turn, the application of a common set of standards governing banking activity and harmonising
supervisory procedures and practice based on demanding standards, will help to contribute towards
increasing the strength of banks in the euro area, building depositors’ confidence and promoting the
stability of the European financial system as a whole.
The Banking Union will, on the one hand, imply a significant sharing of sovereignty and on the other
the mutualising of potential future losses. However, this transfer of responsibility for supervision will
not, for the time being, be accompanied by the transfer of responsibility for financial instruments for
the safety of the system, i.e. as regards banks’ resolution and deposit guarantee, which will continue
to be national mechanisms. It is therefore necessary to create a European fund for banks’ resolution, to
allow the restructuring of institutions without affecting systemic stability and the financial situation of
the countries in which they operate, in parallel with the creation of a common deposit protection system
which will help reduce the probability of the occurrence of runs on the banks which, in a contagion situation, will rapidly condition banking system liquidity. The Banking Union should therefore encompass not
only integrated supervision, but also the sharing of bank resolution and deposit guarantee mechanisms.
Given the strong interconnection between these three domains, the preservation of financial stability in
the euro area and complete de-link between sovereign and banking risks require simultaneous progress
to be made by each. However their accomplishment will occur over variable timespans, given the different
challenges involved in their implementation.
2
21
Macroeconomic and Financial Risks
In short, the Banking Union project is an indispensable component for completing the construction
Activity in the Portuguese banking system continued to develop in an adverse environment, in 2012, both
domestically and externally, notwithstanding the diminishing of tensions in the international financial
markets. Financial markets in the euro area remained fragmented, with a deepening of the economic
recession in Portugal and consequent increase in the materialisation of credit risk. The past year, however,
was also marked by positive developments, particularly including, inter alia, a substantial improvement
in the system’s liquidity and solvency. Accordingly, the ECB’s implementation of non-conventional monetary policy measures, in conjunction with the stability of household deposits, eased the tight financing
situation in the international wholesale debt markets. On the other hand, institutions’ capitalisation
operations endowed the system with a greater loss absorption capacity, reinforcing its financial strength.
Such progress translated into an improvement of international investors’ risk perceptions, enabling two
institutions to issue bond loans in the international markets in the fourth quarter of the year, for the
first time since the inception of the economic and financial assistance programme. The same institutions
returned to the market with an additional two bond issues in January 2013. Notwithstanding the fact
that the costs associated with these issues do not, as yet, permit them to be considered as sustainable
financing sources, the fact that they were made is important to the extent that they indicated the existence of international investors’ appetite for investment in Portuguese banks’ debt issues.
The banking system’s aggregate balance sheet on a consolidated basis was, once again very sharply down
in the second half year, following several signs of stabilisation in the first half. Profitability indicators
were also down in the second half of the year, reflecting higher impairment on the loans and advances
to customers portfolio and evolution of net interest income.
According to the information available on the six major Portuguese banking groups, in the first quarter
2013, there was an improvement in profitability in comparison with the last quarter of the preceding
year (see “Box 3.1 Financial situation of the six major groups of the Portuguese banking system in the
first quarter of 2013”, of this Report). Notwithstanding, profitability remained negative for the six major
banks particularly due to a pronounced reduction in net interest income which more than offset the
significant reduction of impairment on credit. The unprecedented reduction in net interest income during
this period reflects, inter alia, the combined effect of the reduction in funding from the Eurosystem; the
bond issues on both the last of quarter 2012 and the first quarter 2013, at an average cost higher than
most of remaining liabilities; the persistence of the trend of narrowing the spread between interest rates
on loans and deposits from customers and the volume effect associated with the slight decrease in loans
to customers, coupled with the increase in customer deposits.
1 In the analysis set out in “Section 3.1 Activity” and “Section 3.2 Profitability” the aggregate defined as being
the Portuguese banking system refers to credit institutions and financial companies operating in Portugal under
the supervision of Banco de Portugal, with the exception of institutions in the Madeira offshore zone. These
include financial groups, on a consolidated basis, whose consolidation perimeter includes at least one credit
institution or an investment company and credit institutions and investment companies, on an individual basis,
which are not consolidated in Portugal (including the branches of credit institutions or investment companies).
The analysis of this “universe” is important to the extent that it is this collection of institutions to which the
Capital Requirements Directive, approved in September 2005 by the European Parliament, applies, and it is the
reference “universe” in most European countries.
3
23
Banking System: Activity, Profitability and Own Funds Adequacy
3. BANKING SYSTEM: ACTIVITY, PROFITABILITY AND OWN FUNDS
ADEQUACY1
3.1. Activity
I
There was a significant contraction of the Portuguese banking system’s balance sheet in
2012, albeit with a different level of evolution in the first and second half years
Activity in the Portuguese banking system, measured by total assets, on a consolidated basis, contracted
3.3 per cent in 2012 (Chart 3.1.1a). It should be noted that, contrary to 2011, the distribution of the
evolution of assets was uneven in the first and second halves. A virtual stabilisation of assets was noted
in the first half year, essentially justified by the increase in the financial assets portfolio (see “Chapter
6 Market Risk”, of this Report) and claims and investments in other credit institutions, as opposed to
the decline in the loans and advances to customers’ portfolio (Chart 3.1.1b). Activity in the system was
significantly down in the second half year, particularly as a reflection of the 3 per cent decline in the
loans and advances to customers portfolio.2 Loan disposal operations totalled around €4.5 billion for the
year as a whole, against around €7 billion in 2011. More than half of the operations performed in 2012,
involved loans for restructuring funds (particularly loans to companies in the construction, real estate
and tourism segments). In turn, the financial assets portfolio also made a negative contribution to the
evolution of assets in the second half of the year, reflecting the decline of financial assets at fair value
and securities held to maturity portfolios. As regards the evolution of other assets components reference
should be made to the decline of investments in credit institutions, insofar as several institutions started
to take advantage of available liquidity to reduce their exposure to the ECB.
Chart 3.1.1 a
Chart 3.1.1. b
CONTRIBUTIONS TO ANNUAL CHANGE OF
ASSETS | ON A CONSOLIDATED BASIS
CONTRIBUTIONS TO HALF-YEAR CHANGE OF
ASSETS | ON A CONSOLIDATED BASIS
12
16
10
14
8
Per cent and percentage points
Per cent and percentage points
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
24
6
4
2
0
-2
-4
12
10
8
6
4
2
0
-2
-4
-6
-6
-8
-8
Jun-08 Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-12
Jun-08 Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-12
Other assets
Claims and investments in other credit institutions
Securities, derivatives and investments
Net credit to customers - adjusted for securitisation operations
Tangible and intangible assets
Claims and investments in central banks
Other credits and amounts receivable (securitized)
Rate of change of assets
Source: Banco de Portugal.
Source: Banco de Portugal
Notes: Securities, derivatives and investments include financial assets at fair value through profit or loss, available for sale
financial assets, investments held to maturity, investments in
subsidiaries and hedge derivatives. Net credit to customers adjusted for securitisation operations excludes the other credit
and amounts receivable (securitised) component, classified in
the credit portfolio.
Notes: Securities, derivatives and investments include financial assets at fair value through profit or loss, available for sale
financial assets, investments held to maturity, investments in
subsidiaries and hedge derivatives. Net credit to customers adjusted for securitisation operations excludes the other credit
and amounts receivable (securitised) component, classified in
the credit portfolio.
2 Including securitised and non-derecognised loans, adjusted for credit disposal operations.
Customer resources account for an ever higher proportion of financing in the Portuguese
banking system, particularly in the case of the domestic banks
preceding year’s. An increase in the importance of customer resources (whose respective proportion of
the assets total was up to around 51 per cent) and central banks was noted, offset by a reduction of
the importance of other credit institutions’ resources and debt securities. Reference should, however, be
made to the expectable slowdown in the growth of households’ deposits, particularly in the second half
of the year, in comparison to the last two years (see “Chapter 5 Liquidity Risk”, of this Report). It should
also be noted that customer resources have become increasingly more relevant to domestic institutions
(accounting for 56 per cent of total assets) than to resident non-domestic banks in Portugal (with 32 per
cent of total assets). Reference should also be made to the fact that the increase in financing obtained
from the Eurosystem in the first half year3 did not continue through the second half, in a context in
which several domestic institutions began to reduce their exposure to the ECB. As regards other financing
sources, reference should be made to the fact that two institutions returned to the international wholesale
debt market with two bond issues in the fourth quarter of the year, notwithstanding the trend towards
a decline in other credit institutions’ resources and debt securities reflecting, inter alia, the persistence
of refinancing difficulties in the debt market in addition to the own bonds repurchasing operations of
the biggest banking groups, as noted in 2011.
The capitalisation operations of banking institutions, in 2012, were fundamental in providing them with
a greater loss absorption capacity, reinforcing their financial strength, in an adverse macroeconomic and
financial framework. Significant growth of own funds (Core Tier I) in the banking system, was consequently
noted, comprising a substantial increase in subordinated liabilities (of around €5 billion), as a result of
the issue of hybrid instruments subscribed for by the Portuguese state, in addition to institutions’ capital
increases. Although the impact of these operations was more evident in the first half year, reference
should also be made to the fact that during the course of the second half year, two institutions made
fresh capital increases (see “Chapter 3.3 Own Funds Adequacy”, of this Report).
Continuation of a declining trend on the external assets of the domestic banking system in
2012
In 2012, the external assets of the domestic banking system, on a consolidated basis, were down 11 per
cent, to approximately 23 per cent of domestic institutions’ assets (Table 3.1.1).4 The sharpest fall occurred
in the second half of the year. As opposed to December 2009, the proportion of external assets, down
by more than 6 percentage points, kept pace with the domestic activity deleveraging rate, in a context
marked by a retraction of the international financial integration process. On a geographical counterparty
level reference should be made to the reduction of the proportion of the developed countries (particularly euro area economies) and an increase in the proportion of developing countries in Europe, a trend
which has been in force since December 2010. As regards institutional counterparties, reference should
be made to increased exposure to the public sector, particularly in Spain and Italy and a reduction of the
proportion of the non-banking private sector which, however, remains dominant.
3 As referred to in the last issue of the Financial Stability Report, this increase was much more marked in the first
quarter of 2012 and reflected the use of the ECB’s 3 year LTRO in February.
4 International exposure is analysed in accordance with the methodological guidelines of the Bank for International Settlements for the reporting and publication of the “Consolidated banking statistics”. In this analysis only
the sub-set of domestic institutions, on a consolidated basis is considered, as non-domestic institutions are part
of the consolidation perimeter of the banking systems of the countries of their respective head offices. International claims comprise claims on residents outside the country in which the bank is headquartered excluding claims of branches and subsidiaries abroad on residents in the countries in which those offices are located if those
claims are denominated in local currencies. Local assets in local currency, in turn, represent assets over residents
inside the country in which the foreign office booking the claim is located, denominated in local currency.
3
25
Banking System: Activity, Profitability and Own Funds Adequacy
The evolution of the banking system’s financing structure, in 2012, was, to a large extent, similar to the
Table 3.1.1
I
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
26
CONSOLIDATED FOREIGN CLAIMS OF THE DOMESTIC BANKING SYSTEM ON AN IMMEDIATE RISK
BASIS - STRUCTURE | PER CENT
Dec.2009 Jun.2010 Dec.2010 Jun.2011 Dec.2011 Jun.2012 Dec.2012
Total (106 €)
117 727
120 299
114 903
103 497
100 732
98 393
89 446
As a percentage of total assets
29.3
29.2
27.8
25.4
25.3
25.0
23.3
International claims
71.2
70.7
69.2
67.5
67.7
65.0
62.7
19.0
18.0
17.4
19.1
20.2
24.2
19.6
4.7
5.1
4.4
4.3
3.4
3.2
3.6
39.1
41.1
41.3
37.6
38.1
30.6
31.4
8.4
6.5
6.1
6.6
6.0
6.9
8.1
18.3
15.3
14.3
12.4
12.4
13.5
12.3
4.5
5.3
4.9
4.4
4.3
4.4
6.3
48.1
49.7
49.7
50.4
50.8
46.8
43.8
0.2
0.5
0.2
0.2
0.2
0.3
0.3
Maturity
Up to 1 year
From 1 up to 2 years
More than 2 years
Other
Institutional Borrower
Banks
Public sector
Non-banking private sector
Other
Geographical Borrower
Developed countries
51.6
48.4
48.8
48.4
46.4
42.7
41.5
Offshore centres
5.8
6.0
4.9
4.8
7.7
7.7
4.9
Developing countries in Europe
5.2
5.5
5.6
5.9
6.0
6.1
6.5
Other
8.5
10.8
9.8
8.4
7.6
8.5
9.8
28.8
29.3
30.8
32.5
32.3
35.0
37.3
Local assets in local currency
Geographical Borrower
Developed countries
20.0
20.2
20.0
20.1
19.7
20.2
21.0
Offshore centres
0.5
0.5
0.6
0.4
0.5
0.9
1.1
Developing countries in Europe
5.1
4.8
5.1
5.9
5.3
6.5
8.3
Other
3.1
3.8
5.2
6.1
6.7
7.4
6.9
Local assets in local currency (106 €)
33 899
35 204
35 440
33 608
32 519
34 479
33 333
Local liabilities in local currency (106 €)
24 819
22 237
25 291
22 802
25 389
26 419
29 499
Memo:
Source: Banco de Portugal.
3.2. Profitability
Slight recovery in the profitability of the Portuguese banking system in 2012, although values
are still negative and intra-annual evolution very marked
The profitability of the Portuguese banking system currently represents the sector’s main challenge.
Adverse macroeconomic conditions continue to contribute towards higher impairment, particularly on
credit, accompanied by strong pressure on net interest income, the main component of operating income
from banking operations. Net interest income remains compressed, in a context of very low interbank
market interest rates and reduced flows of new operations. Reference should also be made to the negative impact of liabilities management operations, i.e. the conversion of equity into debt instruments by
several institutions5, in addition to the high financing costs associated with the deposits base and hybrid
instruments issued by banks in their recapitalisation operations using public capital. An additional factor
is the slowdown of international activity.
5 Interest on equity instruments takes the form of dividends which are not considered as a cost in net interest
income. The conversion of these instruments into debt securities means that the cost of this financing is accounted for in net interest income.
In 2012 the evolution of profitability of the Portuguese banking system, on a consolidated basis, was
different between the first and second half years (Chart 3.2.1). The evolution of income before tax and
contribution of earnings from financial operations (owing to capital gains on Portuguese public debt
securities following the decline in yields as well as own bonds repurchases) and the operating costs and
other provisions and impairment components (Chart 3.2.2a)6. Notwithstanding, income before tax and
non-controlling interests once again dipped into the red in the second half year, particularly translating
the negative contribution made by the other provisions and impairment component, increase in provisions
and impairment associated with loans and advances to customers and decline in net interest income
(Chart 3.2.2b). As regards the increase in provisions and impairment on credit, a relevant role was played
by the On-site Inspections Programme (OIP) on the eight biggest banking groups in the second half year,
which concentrated on specific credit portfolio sectors such as construction, real estate development
and tourism. The programme estimated the need for an additional €861 million in impairment, for the
second half year.
Globally, the last two years have been marked by negative rates of return, albeit with different intra-annual patterns. The empirical distribution curve on returns on assets suggests, on the other hand, that
the level of dispersion between institutions has remained relatively unchanged since 2011 (Chart 3.2.3).
Chart 3.2.1
RETURN ON ASSETS (ROA) AND RETURN ON EQUITY (ROE)
ROE
25.0
ROA (r.h.s.)
1.6
20.0
1.2
15.0
Per cent
5.0
0.4
0.0
Per cent
0.8
10.0
0.0
-5.0
-10.0
-0.4
-15.0
-0.8
-20.0
-25.0
1999
2002
2005
-1.2
2008H1 2009H2 2011H1 2012H2
Source: Banco de Portugal.
Notes: It is not possible to provide data prior to 2007 for the aggregate under consideration as the adoption of the International
Accounting Standards (IAS) was not transversal to all institutions with different accounting systems coexisting in 2005 and 2006. The
data presented in this chapter are therefore based on different aggregates of institutions. In particular, up to 2004 the list of institutions refers to banks and savings banks, with the exception of banks headquartered or operating exclusively in the Madeira offshore
zone and/or operating mainly with non-residents. Branches of credit institutions headquartered in another European Union member
state - excluding those not classified as monetary financial institutions (MFIs) - in addition to the branches of credit institutions headquartered in third countries were classified as banks. From December 2004 to 2009, two sets of institutions are considered. A first
set for the period December 2004 to December 2007, made up of thirteen banking groups which adopted the Adjusted Accounting
Standards in the preparation of their respective financial statements in 2005 (representing, in December 2004, around 87 per cent
of the total assets of the set of institutions analysed up to the said date). The second set was for the period March 2007 to 2012.
The period of superimposition of the different sets of institutions enables a consistent analysis of the changes to be achieved. The
half-year data have been annualised.
6 The improvement noted should, however, be considered in context owing to the fact that results for second half
2011 were heavily penalised by non-recurring events, including the impact of the special inspections programme
(SIP), the partial transfer of pension funds to the Social Security System and recognition of impairment on Greek
public debt.
3
27
Banking System: Activity, Profitability and Own Funds Adequacy
non-controlling interests in the first half of the year was favourable, in essentially reflecting the positive
I
Chart 3.2.2 a
Chart 3.2.2. b
CHANGE OVER THE PREVIOUS HALF-YEAR – ROA
| BREAKDOWN OF COMPONENTS
CHANGE OVER THE PREVIOUS HALF-YEAR – ROA
| BREAKDOWN OF COMPONENTS
1.5
1.5
1.0
1.0
Percentage points
Percentage points
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
28
0.5
0.0
0.0
-0.5
-0.5
-1.0
0.5
-1.0
2012H1/2011H2
2012H2/2012H1
Net interest income
Net comissions
Income from equity instruments
Income from financial operations
Other operating income
Operational costs
Provisions and impairment on credit to customers
Other provisions and impairment
Appropriation of income from associated companies and goodwill
Change in ROA
Source: Banco de Portugal.
Note: Return on assets calculated on income before tax and minority interests.
Net interest income significantly down in 2012, reflecting a context of reduced money market
interest rates, decline of the loans-to-deposits ratio and high financing costs
The unfavourable context affecting the activity of Portuguese banks, in 2012, conditioned the evolution
of net interest income. The main component of operating income from Portuguese banking operations
was therefore very sharply down year-on-year (17.5 per cent), particularly reflecting the decline related
to operations with customers.
A breakdown of net interest income shows that, except for money market operations, all components
evolved negatively in the last year. As in the first half year, the evolution of the margin associated with
money market operations continued to be explained by the drop in interbank interest rates. The significant increase in central banks’ resources obtained in the first half of the year also started to reduce
in the second half year, in a context in which several domestic banks began to reduce their exposure
to the ECB. As regards operations on financial instruments and notwithstanding the marked decline in
debt securities during the course of 2012, the reduction of the spread between interest on lending and
borrowing (negative price effect) on this type of instrument had a constraining effect on the evolution of
net interest income (Table 3.2.1). Reference should, herein, be made to the negative impact associated
with several institutions’ needs to recapitalise on the basis of hybrid instruments at a high cost.
As in the first half year, a significant reduction of the margin on operations with customers, mainly
responsible for the drop in net interest income, was also noted in the second half year. This evolution,
on the one hand, reflected fewer loans granted to the non-financial private sector and a slight increase
in deposit-taking (negative volume effect) and, on the other, a compression of the spread between
implicit interest rates on credit and deposits. The narrowing of this spread – noted in the case of domestic
activity (Chart 3.2.4) – is explained by diverse factors. Firstly and as referred to in the preceding issue of
Chart 3.2.3
RETURN ON ASSETS (PER CENT) | EMPIRICAL DISTRIBUTION
3
Dec-11
Dec-12
-5.0 -4.5 -4.0 -3.5 -3.0 -2.5 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0
Source: Banco de Portugal
Note: Empirical distribution obtained by the use of a gaussian kernel in which institutions are weighted by assets; indicator calculated
on income before taxes and minority interests.
this report, the rigidity which tends to characterise the transmission of the money market interest rate
to interest rates on deposits, in Portugal, is particularly relevant due to the observed significant drop
in interbank interest rates. Additionally, in the current context of a historically low level of interbank
interest rates, the margin associated with sight deposits, on which practically no interest was earned,
is particularly low. Reference should also be made to the higher interest on deposits during the course
of 2011, largely as a result of greater competition in taking-in customer resources over relatively longer
Table 3.2.1
IMPLICIT AVERAGE INTEREST RATES OF THE MAIN BALANCE SHEET ITEMS (a) | PER CENT
2008
Interbank assets
2009
2010
H1
H2
H1
H2
4.24
4.01
2.10
1.16
2011
2012
H1
H2
H1
H2
1.19
1.50
1.80
2.00
H1
1.60
H2
1.34
Claims on central banks
2.76
2.70
1.13
0.64
0.63
0.74
0.92
1.13
0.49
0.24
Claims on other credit institutions
2.07
1.81
0.94
0.36
0.52
0.62
0.66
0.55
0.56
0.43
Investments in credit institutions
4.73
4.50
2.45
1.38
1.40
1.79
2.16
2.42
1.92
1.63
3.88
Non-interbank assets
6.17
6.49
4.83
3.49
3.33
3.56
4.03
4.68
4.52
Other financial assets
Credit
5.98
6.07
4.57
3.26
3.38
3.41
3.63
4.24
4.24
4.16
Interest-bearing assets
5.92
6.15
4.51
3.20
3.11
3.34
3.74
4.32
4.15
3.66
Interbank liabilities
1.57
4.67
4.65
2.43
1.56
1.26
1.30
1.83
2.27
1.85
Resources from central banks
3.45
4.45
1.15
0.83
1.16
0.80
1.20
1.53
1.14
0.88
Resources from other credit institutions
4.71
4.67
2.53
1.64
1.27
1.44
2.01
2.51
2.14
1.89
Deposits
2.90
3.17
2.39
1.61
1.38
1.60
2.06
2.67
2.76
2.51
Responsabilidades representadas por
títulos sem carácter subordinado
4.63
4.99
3.16
2.38
2.52
2.96
3.19
3.55
3.84
3.94
Subordinated liabilities
5.56
5.50
4.51
3.50
3.34
3.15
3.41
3.91
3.84
6.45
lnterest-bearing liabilities
3.80
4.01
2.65
1.86
1.70
1.84
2.23
2.72
2.68
2.54
Interest bearing assets - Interest bearing
liabilities
2.12
2.15
1.86
1.34
1.42
1.50
1.51
1.60
1.47
1.12
Credit - deposits
3.27
3.32
2.44
1.87
1.95
1.96
1.97
2.01
1.76
1.38
Non-interbank liabilities
Spreads (percentage points):
Source: Banco de Portugal.
Note: (a) Implicit average interest rates are calculated as the ratio between interest flows in the period under consideration and the
average stock of the corresponding balance sheet item.
Banking System: Activity, Profitability and Own Funds Adequacy
29
maturities, in a context of the persistence of difficulties in access to financing in the wholesale debt
market. Notwithstanding, the prudential measure implemented by Banco de Portugal, in November
and 3.2.5b), with a reduction of the average total cost of customer deposits in second half 2012 having
already been noted. Reference should also be made to the fact that the level of dispersion of interest
rates on new operations of deposits with an agreed maturity has been declining since the start of 2012,
in both segments, with such dispersion, in the case of non-financial corporations, being already lower
than noted in January 2010. The banks have also continued to charge higher spreads on most of their
new loans, especially in the non-financial corporations segment, simultaneously reflecting the objective
of stabilising net interest income but also the deteriorating prospects regarding credit risk on the resident
private non-financial sector. This evolution reflects banks’ added difficulties in managing the average
spread on their housing loan portfolios. Most of these loans, characterized by long maturities, are associated with low, fixed spreads, taking into account the banks’ current financing costs.
Slight improvement in operating efficiency in 2012, in a context of significant endeavours to
reduce operating costs
A reduction in operating costs continued to be noted in 2012, although less intensely than in the preceding year. Particular reference should be made to the sharp drop of 9.1 per cent in employee costs,
partly on account of the reduction in the size of the staff complement. Reference should, however, be
made to the fact that the evolution of employee costs has been affected not only by the part transfer
of banks’ pension funds to the Social Security System in 20118, but also by the impact of the decline of
liabilities for death grants in 20129, for which the amounts of the two years are not directly comparable.
Chart 3.2.4
INTEREST RATE SPREADS IN OPERATIONS WITH CUSTOMERS
6
Spread on loans
Spread on time deposits (r.h.s., reversed)
Total spread
6-month moving average of 6-month Euribor
Spread on total deposits (r.h.s., reversed)
-6
5
-5
4
-4
3
-3
2
-2
1
-1
0
0
-1
1
-2
2
Percentage points
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
30
20117, made it possible to ease the upwards pressure on the cost of customer deposits (Charts 3.2.5a
Per cent and percentage points
I
-3
3
Jan03Jan04Jan05Jan06Jan07Jan08Jan09Jan10 Jan11 Jan12Jan13
Source: Banco de Portugal.
Note: The spread on lending operations was calculated as the difference between the interest rates on outstanding amount of loans
(supplied in the Monetary and Financial Statistics) and the 6 month moving average of 6 months Euribor, whereas the spread on
borrowing operations is the difference between the 6 months moving average of months Euribor, and interest rates on outstanding
amounts of deposits. The total spread comprises the difference between the interest rate on loans and deposits. Last observation:
January 2012.
7 In April 2012, Banco de Portugal introduced a change to this measure, with the aim of increasing the penalty on
short term deposits. For further details see the consolidated version of Banco de Portugal “Instruction 28/2011”
which includes the changes made by “Instruction 15/2012” at: http://www.bportugal.pt/sibap/application/
app1/instman.asp?PVer=P&PNum=28/2011.
8 For further details, see: “Box 4.2 Accounting and prudential impact of the partial transfer of banking sector
pension funds to the Social Security System”, Banco de Portugal, Financial Stability Report-May 2012.
9 Impact of the publication of Decree Law 133/2012, which reduces the attributing of death grants. This change
in the law generated a decline in the banks’ pension funds’ liabilities, with a positive impact on their income
statements.
INTEREST RATES APPLIED ON NEW OPERATIONS
OF DEPOSITS WITH AN AGREED MATURITY TO
NON-FINANCIAL CORPORATIONS BY THE EIGHT
MAJOR RESIDENT BANKING GROUPS
INTEREST RATES APPLIED ON NEW OPERATIONS
OF DEPOSITS WITH AN AGREED MATURITY TO
HOUSEHOLDS BY THE EIGHT MAJOR RESIDENT
BANKING GROUPS
7
6
6
5
5
4
4
Per cent
7
3
2
2
max
min
average
1
0
Jan-10
3
Jul-10
Jan-11
Jul-11
Jan-12
Jul-12
max
min
average
1
Jan-13
0
Jan-10
Jul-10
Jan-11
Jul-11
Jan-12
Jul-12
Jan-13
Source: Banco de Portugal.
Note: Last observation: February 2013.
As regards operating efficiency, there was a slight improvement in the cost-to-income ratio10 insofar as the
reduction in operating costs was higher than that of net operating income from banking operations. The
ratio was down 2.5 p.p. over 2011 to 59 per cent in 2012. This evolution is in contrast to the preceding
year, which was a period of deterioration of this indicator as the consequence of a sharper decline in
operating income from banking operations. Reference should also be made to greater heterogeneity, in
2012, translating different levels of operating efficiency in the Portuguese banking system (Chart 3.2.6).
Chart 3.2.6
COST TO INCOME RATIO (PER CENT) | EMPIRICAL DISTRIBUTION
Dec-11
30
40
50
60
70
80
Dec-12
90 100 110 120 130 140 150 160 170
Source: Banco de Portugal
Notes: Empirical distribution obtained by the use of a Gaussian kernel, in which institutions are weighted by total assets; indicator
calculated as the ratio between operating costs (defined as the sum of staff costs, general administrative costs and depreciation and
amortisations) and gross income.
10 The cost-to-income indicator is defined as the ratio between operating costs (comprising the sum of general
administrative expenditure, staff costs and depreciation) and gross income.
3
31
Banking System: Activity, Profitability and Own Funds Adequacy
Chart 3.2.5b
Per cent
Chart 3.2.5a
Unfavourable evolution of international activity in 2012
The results of the international subsidiaries and branches of Portuguese banking groups have made a
highly favourable contribution to their respective consolidated results, over the last few years. In parti-
32
two years, have made it possible to partly offset domestic activity weakness and, accordingly reduce the
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
I
cular, income before tax and non-controlling interests deriving from international activity over the last
impact on groups’ profitability indicators. Notwithstanding, there was a slowdown of international activity, in 2012. This took the form of a decline in the results of the international subsidiaries and branches
of Portuguese banking groups (Table 3.2.2). This decline is essentially explained by the unfavourable
evolution of net interest income, significant increase in staff costs and higher impairment, reflecting
the increased materialisation of credit risk in international activity. Notwithstanding differences in the
underlying evolutions between institutions and geographies, there is an across-the-board trend towards
a slowdown in international activity for the biggest banking institutions. Poorer results in Spain are
particularly associated with a reduction of activity, particularly the deterioration of credit quality. In turn,
lower profitability, in Angola, is related with the reduction of net interest income, largely on account of
the reduction of yields on domestic public debt securities and central bank securities which make up a
significant part of the balance sheets of the subsidiaries of Portuguese banks.
3.3. Own funds adequacy11
Over the last two years, Portuguese banking institutions have achieved an ambitious reinforcement of
their solvency levels, in line with Banco de Portugal and European Banking Authority requirements.12 It
should also be remembered that this endeavour enabled the prudential impacts of the special inspections
programme (SIP) to be accommodated and the partial transfer of pension liabilities to the Social Security
System, recognised in June 2012, in which institutions benefited from a grace period (Banco de Portugal,
Table 3.2.2
RELEVANCE OF INTERNACIONAL ACTIVITY FOR THE INCOME OF THE EIGHT MAJOR RESIDENT
BANKING GROUPS | PER CENT
Relative weight of foreign
subsidiaries
International activity
y.o.y. rate of change
Domestic activity
y.o.y. rate of change
2010 2010 2011 2011 2012 2012 2011 2011 2012 2012 2011 2011 2012 2012
Jun Dec Jun Dec Jun Dec Jun Dec Jun Dec Jun Dec Jun Dec
Net interest income
26.4 28.7 28.7 28.9
28.8 30.7 16.1
3.8
-9.7 -13.1
3.4
Commissions
19.8 20.3 21.2 20.7
26.5 25.2
0.6 27.1 23.4
-1.2
8.0
Gross income
22.4 24.2 25.6 28.9
25.4 26.6 11.5 10.3
6.6
-5.6
Administrative costs
22.6 23.3 24.7 24.7
28.2 28.1 10.2
6.7
7.8
0.0
-1.8
6.7
7.4 10.5
0.7
2.8 -10.1 -20.2
-2.2
-6.4 -13.5
-1.2
-5.1
-4.2
7.8
6.1
-9.8
-9.3
21.1 21.4 22.7 22.5
27.2 26.9 10.6
Impairment
14.9 14.8
10.3 11.9 -19.8 17.8 83.2 47.3 44.0 130.0 54.5
Income before tax and minority
interests
28.7 34.8 77.0
- 243.1
- 32.7 13.8 -21.9 -53.7 -84.0
-
-
-
Net income
13.8 16.9 38.4
-
- 52.0 34.6 -23.9
-
-
-
of which: staff costs
8.9
8.2
-
- -61.1
0.1 -15.5 -12.8
-2.6
Source: Banco de Portugal.
11 The set of institutions analysed in this section is not the same as in “Section 3.1. Activity” and “Section 3.2 Profitability” owing to the exclusion of the branches of financial groups headquartered in European Union member
countries.
12 For further details, see: “Box 4.3 New capital adequacy requirements: recent developments and outlook for
2012”, Banco de Portugal, Financial Stability Report-November 2011.
Official Notice 1/2012). In December 2012, the Portuguese banking system had an average Core Tier 113
ratio of 11.5 per cent, up 2.8 percentage points over December 2011. The highly significant increase,
half year. More specifically, reference should be made to the importance of the issue of hybrid financial
instruments eligible as core own funds, subscribed for by the Portuguese state, for three of the biggest
banking groups, for an amount of around €5 billion. Two institutions made fresh capital increases in
the second half of the year. Accordingly, at the end of 2012, the Core Tier 1 ratio of seven of the eight
biggest Portuguese banks was higher than the objective of 10 per cent defined in the economic and
financial assistance programme (Banco de Portugal, Official Notice 3/2011).14
Significant growth of basis own funds, particularly deriving from the capitalisation
operations of banking institutions in the first half year
The highly significant increase in original own funds, in 2012, reflected the capitalisation operations of
Portuguese banking groups for the purpose of improving their solvency levels (Chart 3.3.1). This evolution
is essentially related with banking institutions’ operations in the first half year. An increase in eligible capital
deriving from the issue of equity-like instruments, the capital increase of one of the biggest banking
institutions and the increase in share issue premiums associated with the capital increase of another
banking group were noted in the period. There was also a very sharp increase in reserves, particularly
based on the incorporation of retained earnings.15 Two institutions made fresh capital increases in the
second half of the year, justifying the increase noted in eligible capital. It should also be noted that as
regards recapitalisation operations as a whole, up to the end of 2012, the Portuguese state invested
€4.5 billion out of the €12 billion comprising the global appropriation of the banking solvency support
mechanism16 as set out in the economic and financial assistance programme.
In turn, own funds requirements were down 4 per cent, contributing no more than 14 per cent to the
improvement in the system’s Core Tier 1 ratio. It should be noted, in this respect, that the proportion
of risk-weighted to total assets has been declining over the last few last years in the case of most
institutions (Chart 3.3.2). The banking system’s average assets weighting factor has been reduced by
around 2 percentage points, over the last two years, owing to the various initiatives of several banking
groups to optimise their risk-weighted assets, in particular, the association of guarantees and collateral
to exposures held.
13 The Core Tier I ratio establishes a minimum level of capital that the institutions must assure based on own funds
requirements deriving from the risks associated with their activity. The ratio, as such, is assessed on the quotient
between “core” own funds and risk-weighted positions. “Core” own funds include an institution’s highest
quality capital, in terms of its stability and capacity to absorb losses, less any losses and certain elements with
no autonomous realisation value, based on the principle of an institution as a “going concern”. Risk-weighted
positions represent a measure of the risks deriving from financial activity, namely credit, market (including
minimum own funds requirements related to foreign exchange and trading portfolio) and operational risks. In
Portugal, the Core Tier 1 measure is based on the Basel III rules applicable in 2013 for the definition of Common
Equity Tier 1, i.e. prior to the application of the transitory regime for certain deductions. In particular, it does not
include the deduction relative to investments in non-consolidated financial institutions, nor deferred tax assets
deductions.The calculation of the Core Tier 1 ratio is defined in Banco de Portugal’s Official Notice 1/2011.
14 Only BANIF, Banco Internacional do Funchal, SA, failed to comply with this requirement. However, in January
2013 the respective recapitalisation plan providing for an increase in capital of €1.4 billion was approved at the
respective shareholders’ meeting of BANIF. Therefore, based on an initial investment of €1.1 billion by the state,
BANIF has now complied with the minimum solvency requirement imposed by Banco de Portugal in the context
of the economic and financial assistance programme.
15 This development is in line with the recommendation of Banco de Portugal to retain earnings or, in the case of
their distribution, to immediately reinvest them in capital in order to strengthen own funds.
16 One of the institutions included in these operations has already paid off €300 million in financing obtained
through hybrid finanical instruments.
3
33
Banking System: Activity, Profitability and Own Funds Adequacy
in 2012, particularly reflected the capitalisation operations of the biggest banking groups in the first
Chart 3.3.1
Chart 3.3.2
BREAKDOWN OF ORIGINAL OWN FUNDS
EVOLUTION OF THE PORTUGUESE BANKS’ RISKWEIGHTED ASSETS TO ASSETS RATIO
I
40 000
35 000
Elegible capital
Minority interests
Other positive elements
Negative elements
Original own funds (total)
Non-core elements
Original own funds (core)
90
80
70
Dec 2012
45 000
EUR millions
30 000
25 000
60
20 000
50
15 000
10 000
40
5 000
30
0
30
40
50
-5 000
60
70
80
90
Dec 2010
-10 000
Dec-08Jun-09Dec-09Jun-10Dec-10Jun-11Dec-11Jun-12Dec-12
Source: Banco de Portugal.
Eight major banking groups
Other banking groups (aggregate)
Banking system
average
Source: Banco de Portugal.
Note: For banks that make use of IRB methods in the computation of capital requirements, risk-weighted assets were adjusted
to assure proper comparability with banks that rely on standard
methods.
Across-the-board increase of Core Tier 1 ratio
As in 2011, it was also possible, in 2012, to note the importance the banks attach to what are considered
to be core elements, which were up by approximately 26 per cent. The increase noted in original own
funds was fully justified by an increase in better quality own funds, in terms of their stability and loss-absorption capacity, as non-core elements were down by around 8 per cent (Chart 3.3.3).
Chart 3.3.3
OWN FUNDS ADEQUACY OF THE PORTUGUESE BANKING SYSTEM
13.0
Core Tier-I ratio
Original own funds adequacy ratio, Tier 1
Overall own funds adequacy ratio
12.0
11.0
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
34
10.0
9.0
8.0
7.0
6.0
Jun Dec Jun Dec Jun Dec Jun Dec Jun Dec
2008 2008 2009 2009 2010 2010 2011 2011 2012 2012
Source: Banco de Portugal.
Notes: The series presented exclude BPN and BPP. It should be noted that BPP was liquidated in April 2010, after which it ceased to
be included in the universe of banking institutions.
The Portuguese banking system’s Core Tier 1 ratio was therefore significantly up, in 2012, to an end of
year 11.5 per cent, against the 10 per cent objective defined in the economic and financial assistance
programme. The improvement of this indicator was transversal to most of the institutions analysed (Chart
3.3.4). Reference should also be made to the increase in heterogeneity between institutions noted in
comparison to 2011.
From an accounting viewpoint, there was a significant increase in the shareholders’ equity to total assets
ratio, particularly in the first half of the year (Chart 3.3.5). This evolution is explained by several factors.
Firstly, equity benefited from the improvement in international investors’ perceptions of the risk attached
to the Portuguese state, as reflected in the decline of the negative value of reserves resulting from the
fair value appreciation of debt securities. Secondly, the capital increases made over the course the year,
in addition to the increase in share issue premiums deriving from the capital increase made by another
bank, made a positive contribution to the evolution of equity. Lastly, the contraction of activity in the
Portuguese banking system, noted in 2012, also made a favourable, albeit less marked contribution, to
the evolution of this ratio.
Chart 3.3.4
Chart 3.3.5
CORE TIER-I RATIO (PER CENT) | (ORIGINAL OWN
CAPITAL TO ASSETS RATIO | ADJUSTED FOR NON-
FUNDS - NON-CORE ELEMENTS)/ (CAPITAL REQUIREMENTS *
RECURRENT EVENTS OBSERVED IN THE SECOND SEMESTER OF
12.5)
2011
Dec-11
8.0
Dec-12
7.5
Capital/Assets
Tangible capital/Tangible assets
Adjusted capital/assets
Adjusted Tangible capital/Tangible assets
Per cent
7.0
6.5
6.0
5.5
5.0
4.5
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
4.0
Jun Dec Jun Dec Jun Dec Jun Dec Jun Dec
2008 2008 2009 2009 2010 2010 2011 2011 2012 2012
Source: Banco de Portugal.
Source: Banco de Portugal.
Notes: Empirical distribution obtained by the use of a gaussian
kernel in which institutions are weighted by assets. The series
presented exclude BPN. It should be noted that BPP was liquidated in April 2010, after which it ceased to be included in the
universe of banking institutions.
Notes: It should be noted that BPP was liquidated in April
2010, after which it ceased to be included in the universe of
banking institutions. Non-recurrent events comprise the Special Inspections Programme (SIP), the partial transfer of banks’
pension funds to the Portuguese Social Security System and
impairment charges related to Greek public debt.
Banking System: Activity, Profitability and Own Funds Adequacy
35
Significant improvement of the accounting capital to assets ratio
1
3
box 3.1 | Financial situation of the six major groups in the
Portuguese banking system in first quarter 20131
According to the information available on the six major Portuguese banking groups, in first quarter 2013,
3
basis (Table 1) in comparison to the end of the preceding year. This evolution is essentially justified by the
37
decline in the net loans and advances to customers’ portfolio (including securitised, non-derecognised
Banking System: Activity, Profitability and Own Funds Adequacy
there was a slight contraction of activity in the banking system, measured by total assets on a consolidated
assets) and the significant drop in claims and investments in central banks and other credit institutions.
There was also an increase in available for sale financial assets and financial assets at fair value portfolios,
particularly associated with debt instruments issued by non-resident entities.
First quarter saw a continuation of the trend towards the recomposition of banks’ financing structures,
based on an increase in the proportion of customer resources and decline in the proportion of debt
securities. Reference should, however, be made to the fact that as in the fourth quarter of the preceding
Table 1
BALANCE SHEET OF THE SIX MAJOR BANKING GROUPS | ON A CONSOLIDATED BASIS
Structure (as a
Year-on-year rates of change
percentage of total
(per cent)
assets)
2010 2011 2012 2013
2012
2013
Quarterly rates of change
(per cent)
2012
2013
Dec. Dec. Dec. Mar. Mar. Jun. Sep. Dec. Mar. Mar. Jun. Sep. Dec. Mar.
Cash and claims on central banks
2.1
2.7
3.7
2.8
12.5 45.1 16.1 33.9 26.9 -18.9 33.0 -21.8 58.7 -23.1
Claims and investments in other
credit institutions
3.2
4.2
3.2
3.0
-7.0
-2.2
-5.3 -25.7 -5.6 -26.1 8.7
-1.2
-6.4
-6.0
Securities, derivatives and investments 19.7 18.0 19.0 20.2
3.7
0.9
-0.1
1.9
-0.2
8.1
-4.1
-0.4
-1.3
5.9
Net credit to customers
60.0 58.1 61.1 61.0
-3.7
-0.2
-0.5
1.7
-0.4
1.3
3.5
-0.8
-2.3
-0.7
Securitised non-derecognised assets
9.6
-10.1 -36.8 -42.3 -42.8 -40.1 -6.6 -29.9 -10.0 -2.9
-2.1
10.4
6.1
6.0
Tangible and intangible assets
1.0
1.0
1.1
1.1
-6.6
-2.1
-0.3
10.9
7.1
-1.9
1.7
3.1
7.8
-5.2
Other assets
4.4
5.7
5.9
5.8
20.0 11.1
0.8
-0.3
0.7
-2.9
2.3
-2.1
2.5
-1.9
-2.5
-4.5
-3.2
-3.6
-0.2
-0.5
-2.0
-0.5
-0.6
9.1
0.8 -19.3 12.6
6.5
-8.9
-7.7
-9.9
-36.0 -25.3 -30.0 -24.6 -9.3 -14.4 -8.1
-5.4
1.3
2.9
Total assets
100.0 100.0 100.0 100.0 -1.8
Resources from central banks
9.9
10.9 11.4 10.3 29.6 24.7
Resources from other credit
institutions
7.3
5.6
Resources from customers and other
loans
46.4 52.4 55.6 56.9 10.3
-1.9
0.3
1.8
1.7
Liabilities represented by securities
20.1 17.2 13.5 12.9 -22.0 -22.8 -22.6 -24.0 -18.2 -11.6 -5.8
-6.0
-3.1
-4.8
Subordinated liabilities
2.1
1.3
2.7
2.6
-35.5 52.7 81.3 95.6 98.2
-2.2 104.2 -0.1
-1.9
-0.9
Other liabilities
7.5
7.5
5.9
6.0
-4.3 -17.1 -24.3 -23.8 -18.5 -5.7 -12.6 -4.0
-3.6
0.8
6.8
5.1
6.6
6.7
-22.6 -8.9
Capital
4.4
4.5
3.1
1.4
2.6
1.9
2.4
1.9
27.2 18.4
7.7
9.6
6.0
1.6
0.3
100.0 100.0 100.0 100.0 -1.8
-2.5
-4.5
-3.2
-3.6
-0.2
-0.5
-2.0
-0.5
-0.6
71.4 70.9 70.5 70.5
-3.1
-4.1
-5.2
-3.7
-5.0
0.6
-0.9
-1.5
-2.0
-0.7
Credit to customers including
non-derecognised securitisation
operations (adjusted for loan disposal
operations)
71.6 72.8 73.6 73.6
-1.8
-3.2
-4.1
-2.2
-3.7
0.9
-0.8
-1.0
-1.3
-0.6
Total liabilities and capital
Memo:
Credit to customers including
non-derecognised securitisation
operations
Source: Banco de Portugal.
1 The total assets of the six banking groups analysed in this Box (Caixa Geral de Depósitos, Espírito Santo Financial Group, Banco Comercial Português, Banco BPI, Santander Totta and Caixa Económica Montepio Geral)
accounted for around 78 per cent of the Portuguese banking system’s assets in December 2012. To neutralise
the impact of the integration of Finibanco in Caixa Económica Montepio Geral, data prior to 2011 were revised
to include the said institution.
year, two institutions resumed their financing operations in the international markets with two bond
issues, as a result of an improvement in the risk perception of international investors. On the other hand
and as noted in second half 2012, there continued to be a decline in central banks’ resources. This was
liquidity situation of the main banking institutions.
First quarter 2013 saw an improvement in the profitability of the six major banking groups in comparison to the last quarter of 2012, notwithstanding the fact that income before tax and non-controlling
interests continued to be negative (Chart 1). The net interest income and income from services and
commissions contributed negatively to this evolution, in the context of a reduction of activity in the
Portuguese banking system (Chart 2, Table 2). The evolution of net interest income is also conditioned by
other factors, particularly including the historically low level of interbank interest rates and high financing
costs, particularly associated with the deposits base. Similarly, the reduction of exposure to the ECB,
as already noted, also conditioned the evolution of net interest income. The recognition of impairment
on the credit portfolio continued to have a negative effect on banks’ profitability levels, although there
was a substantial decline in comparison to the preceding quarter’s particularly high amounts, being the
main reason for the improved results. There was also a slight decline in income from financial operations
and operating costs.
The Core Tier 1 ratio of the six main banking groups improved slightly to 11.7 per cent at the end of
September 2012 (Table 3). This evolution reflects the increase in basis own funds, through an increase
in reserves and eligible, as own funds requirements remained practically unchanged.
Chart 1
Chart 2
RETURN ON ASSETS (ROA) AND RETURN ON
EQUITY (ROE) OF THE SIX MAJOR BANKING
GROUPS
PROFIT AND LOSS ACCOUNT OF THE SIX MAJOR
BANKING GROUPS | QUARTERLY FLOWS
ROE
Adjusted ROE
ROA (r.h.s.)
ROA ajustado (r.h.s.)
1.0
15
10
2,000
0.5
5
Net interest income
Income (net) from services and commissions
Income from financial operations and associated impairment
Operating costs
Provisions and impairment on credit to customers
Other provisions and impairment
Income before tax and minority interests
1,500
0
1,000
-5
-0.5
-10
-15
-1.0
-20
EUR millions
0.0
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
38
particularly marked in the case of one of the groups analysed and confirms the improvement of the
Per cent
I
500
0
-500
-1,000
-1,500
-25
-1.5
-30
-2,000
-2,500
-35
-2.0
Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 Q1
2008 2008 2009 2009 2010 2010 2011 2011 2012 2012 2013
Source: Banco de Portugal.
Notes: Indicators calculated on net income. Quarterly data
have been annualised. The adjusted figures exclude a set of
non-recurrent events observed in the fourth quarter of 2011:
the Special Inspections Programme (SIP), the partial transfer of
banks’ pension funds to the Portuguese Social Security System
and impairment charges related to Greek public debt.
-3,000
Q1
2011
Q2
2011
Q3
2011
Q4
2011
Source: Banco de Portugal.
Q1
2012
Q2
2012
Q3
2012
Q4
2012
Q1
2013
Table 2
PROFIT AND LOSS ACCOUNT OF THE SIX MAJOR BANKING GROUPS | ON A CONSOLIDATED BASIS, AS A
PERCENTAGE OF AVERAGE ASSETS
3
2012
Cumulative income (year to date)
2013
2012
2013
1Q
2Q
3Q
4Q
1Q
Mar.
Jun.
Sep.
Dec.
Mar.
Net interest income
1.33
1.27
1.10
1.14
0.86
1.33
1.30
1.23
1.21
0.86
Income (net) from services and commissions
0.69
0.78
0.70
0.74
0.69
0.69
0.74
0.72
0.73
0.69
Income from financial operations
0.47
0.54
0.35
0.50
0.36
0.47
0.50
0.45
0.47
0.36
Other income
0.08
0.25
0.05
0.00
-0.02
0.08
0.16
0.13
0.10
-0.02
Gross income
2.56
2.84
2.20
2.39
1.89
2.56
2.70
2.54
2.50
1.89
Operating costs
1.35
1.34
1.42
1.53
1.35
1.35
1.35
1.37
1.41
1.35
Provisions and impairment
0.94
1.79
1.25
1.94
0.92
0.94
1.36
1.33
1.48
0.92
of which: associated with credit to
costumers
0.77
1.51
1.04
1.26
0.70
0.77
1.14
1.11
1.14
0.70
Consolidation differences and appropriation
of net income
-0.05
-0.09
-0.32
-0.10
-0.18
-0.05
-0.07
-0.15
-0.14
-0.18
Income before tax and minority
interests
0.32
-0.20
-0.15
-0.98
-0.21
0.32
0.06
-0.01
-0.25
-0.21
Income tax profit
0.12
0.05
-0.09
-0.45
-0.04
0.12
0.08
0.03
-0.09
-0.04
Income before minority interests
0.20
-0.25
-0.06
-0.52
-0.17
0.20
-0.02
-0.04
-0.16
-0.17
Minority interests
0.09
0.04
0.10
0.04
0.00
0.09
0.06
0.07
0.07
0.00
Net income
0.11
-0.29
-0.16
-0.56
-0.17
0.11
-0.09
-0.11
-0.22
-0.17
Source: Banco de Portugal.
Note: Quarterly and cumulative income have been annualised for the calculation of the respective percentages over average assets.
Table 3
OWN FUNDS ADEQUACY OF THE SIX MAJOR BANKING GROUPS | ON A CONSOLIDATED BASIS, EUR MILLIONS
2012
2013
Mar.
Jun.
Sep.
Dec.
Mar.
(A)
24913
29944
29786
29347
29657
(B)
1218
1179
1173
1156
1196
2. Capital requirements
(C)
20165
20100
19831
19458
19450
3. Core Tier - I ratio (%)
(A-B)/(C x 12.5)
9.4
11.4
11.5
11.6
11.7
1. Own funds
Original own funds
of which: non-core elements
Source: Banco de Portugal.
39
Banking System: Activity, Profitability and Own Funds Adequacy
Quarterly income (flow)
4
The materialisation of credit risk remained at very high levels over the course of 2012, with successive rises
41
in the non-performing loans ratio and particularly so in the case of loans to households for consumption
and other purposes and loans to non-financial corporations (Chart 4.1). The most recent data indicate
a certain stabilisation or even reduction of the annual flow of new overdue and other doubtful loans in
the non-financial private sector (Chart 4.2). This favourable evolution is visible in the banks’ main credit
segments. Nevertheless, in a framework in which the adjustments needed to correct the imbalances in
the Portuguese economy are in progress, there are still vulnerabilities particularly in several non-financial
corporations sectors.
In 2012, adjustments of private consumption to values more compatible with a lower level of permanent
income, in conjunction with a continuing trend towards a reduction of investment, translated into a
significant increase in the households’ savings rate and net lending. In the case of non-financial corporations, the contraction of investment and higher savings were reflected in a substantial reduction of their
net borrowing. These developments contributed to a situation in which the non-financial private sector
registered a financing capacity situation for the first time since the inception of the euro area (Chart 4.3).
Debt levels in the non-financial private sector are still, however, very high, even slightly increasing in the
case of non-financial corporations. In several sectors, particularly those more geared to the domestic
Chart 4.1
Chart 4.2
NON-PERFORMING LOANS RATIO
CREDIT DEFAULT INDICATORS ON BANK LOANS
TO THE NON-FINANCIAL PRIVATE SECTOR
Non-financial private sector
Non-financial corporations
Housing
Consumption and other purposes
12.0
18.0
16.0
10.0
2.5
10.0
8.0
1.5
6.0
8.0
Per cent
2.0
12.0
Per cent
Per cent
14.0
Default ratio
Non-performing loans ratio
Annual flow of new overdue and other doubtful bank loans (rhs)
3.0
1.0
4.0
6.0
0.5
4.0
2.0
2.0
0.0
Dec-08
Dec-09
Dec-10
Dec-11
Dec-12
Source: Banco de Portugal.
Notes: See non-performing loans definition, footnote 1.
Last observation: December 2012.
0.0
0.0
-0.5
Jan-99 Jan-01 Jan-03 Jan-05 Jan-07 Jan-09 Jan-11 Jan-13
Source: Banco de Portugal.
Notes: See credit risk indicators definitions, footnote 1.
Last observation: March 2013, except non-performing loans
ratio (December 2012).
1 Three credit risk indicators are preferentially used in this chapter. The default ratio is defined as total loans
overdue for more than 30 days and other doubtful loans expressed as a percentage of the outstanding amount
of loans adjusted for securitisation. The annual flow of new overdue and other doubtful loans is expressed as a
percentage of the loans, adjusted for securitisation, asset write-downs/write-offs, reclassifications and starting
December 2005, credit disposals. The default ratio and the annual flow of new overdue and other doubtful
loans are based on Monetary Financial Statistics. Lastly, non-performing loans correspond to a broader concept
of credit risk made up of three elements: the amount owed on credit with overdue instalments of principal or
interest for a period of more than 90 days; the amount outstanding of restructured credits not covered by the
preceding item and which fulfil certain characteristics; and lastly the amount of credit comprising installments of
principal or interest overdue for less than 90 days, but upon which there is evidence to justify their classification
as non-performing credit, namely a debtor’s bankruptcy or liquidation.
Credit Risk
4. Credit risk1
market and those mainly populated by smaller companies, low levels of profitability increase the difficulty
of the necessary deleveraging process. In the case of small companies, although a significant number
particularly for those in the construction sector, the level of restrictiveness in terms of lending from resident banks is very high. This situation reflects the high level of corporations credit risk, largely deriving
from the excessive growth in debt recorded in the years preceding the current recession. In the case
of other sectors, it is possible that the improvement in the perception of sovereign and banking sector
risks observed in 2012, is beginning to trickle down into a reduction of the level of restrictiveness in
lending criteria.
In a context in which the macroeconomic outlook points towards further contraction of activity in 2013,
with negative implications on demand for goods and services and on the labour market situation, it is
unlikely that the materialisation of credit risk is significantly reduced in the short term. Moreover, there
is a fraction of the banks’ portfolio of mortgage loans with relatively high loan-to-value ratios, which is
relevant if we take into account that housing prices fell more than 10 per cent since the beginning of
the Economic and Financial Assistance Programme (see “Box 4.2 Characterization general price index of
housing produced by INE”, of this Report). However, given banking system capitalisation operations over
the course of 2012, the risks attached to financial stability deriving from the materialisation of credit risk
are currently lower than in the past. It is, in any event, important for the banks to continue to increase
the amount of impairment provisions on credit portfolio losses, in order to increase their capacity to
absorb additional losses (Chart 4.4).
Households
A significant increase in savings contributed towards the increase in households’ net lending
in 2012
The correction of imbalances in the households’ balance sheet continues to proceed gradually and is one
element of the Portuguese economy’s adjustment process. There was a significant increase in households’
savings, in 2012, which are currently at a peak since the inception of the euro area, both in absolute
Chart 4.3
Chart 4.4
NET LENDING/NET BORROWING OF NONFINANCIAL PRIVATE SECTOR
NON-PERFORMING LOANS COVERAGE RATIO
Non-financial private sector
Non-financial corporations
Housing
Consumption and other purposes
Households
Non-financial corporations
Non-financial private sector
10.0
120.0
5.0
100.0
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
42
of companies is debt-free, the debt ratios of the indebted ones are very high. For these companies, and
Percentage of GDP
I
0.0
80.0
60.0
-5.0
40.0
-10.0
20.0
-15.0
1999 2001 2003 2005 2007 2009 2011
Source: INE.
2011 2012
H2
H2
0.0
Dec-08
Dec-09
Dec-10
Dec-11
Dec-12
Source: Banco de Portugal.
Notes: The non-performing loans coverage ratio is calculated
dividing accumulated provisions/impairments by non-performing loans (see definition, footnote 1).
levels and as a percentage of disposable income. This evolution, accompanied by the declining trend of
the investment rate, translated into a significant increase of households’ net lending (Chart 4.5).
adapt consumption to permanent income prospects as well as a rise in precautionary savings. Indeed, in
2012 there was a sharp contraction of private consumption, in the context of a significant reduction of
household disposable income, marked deterioration of conditions in the labour market, tight financing
conditions and high levels of uncertainty over the macroeconomic outlook. Another possible cause for
the increase in the savings rate may have been a composition effect on disposable income. In fact, in
2012 the sharp reduction of compensation of employees was accompanied by a substantial increase
in property income which, according to the available evidence, presents a higher propensity to savings
than labour income.2
In 2012, net interest received by households, as a percentage of disposable income, which, in 2011,
was already at positive levels, increased significantly, reflecting a continuation of the increase in interest
received and a reduction of interest paid (Chart 4.6). The reduction of the money market interest rates
since the end of 2011, has translated into a significant decline of interest rates on loans to households. For
this situation contributes the mechanical effect on the interest rates on outstanding amounts of housing
loans, most of which are indexed to Euribor with fixed spreads. Additionally, the spreads on new loans,
both for housing and consumption and other purposes have, since the start of 2012, interrupted the
upwards trend registered in preceding years, although still remaining at a historically high level (Chart
4.7). In turn, a contributory factor to the increase in interest received, in 2012, was the maintenance of
the interest rates of bank deposits at high levels, notwithstanding the reduction of interest rates on new
deposits registered over the course of the year. As expected, given greater inertia to a downwards path,
the persistence of interest rates at high levels in 2012 was particularly evident in the case of interest rates
on the outstanding amount of deposits with longer agreed maturities. The share of these deposits on
total deposits increased significantly in 2011, at a time of strong competition from the banks in terms
of deposit-taking (Chart 4.8) (see “Chapter 5 Liquidity risk”, of this Report).
Chart 4.5
Chart 4.6
HOUSEHOLDS’ NET LENDING, SAVING AND
INVESTMENT
INTEREST RECEIVED AND PAID BY HOUSEHOLDS
Net interest income (a-b)
Interest received (a)
Interest paid (b)
15
12
12
10
Percentage of disposable income
Percentage of disposable income (a)
Net lending / borrowing
Saving
Net capital transfers
Acquisitions less disposals of real assets (b)
9
6
3
0
-3
-6
-9
8
6
4
2
0
-2
-12
-15
-4
1999 2001 2003 2005 2007 2009 2011
2011 2012
H2
H2
Source: INE.
Notes: (a) Disposable income adjusted for the change in net
equity of households on pension funds. (b) Corresponds to the
sum of GFCF, changes in inventories, acquisitions less disposals
of valuables and acquisitions less disposals of non-produced
non-financial assets.
1999 2001 2003 2005 2007 2009
2011
2011
H2
2012
H2
Sources: INE and Banco de Portugal.
Notes: (a) Difference between interest income received by
households included in the income account and the respective
financial intermediation services indirectly measured (FISIM). (b)
Corresponds to the sum of interest payable by households included in the income account with the respective FISIM.
2 See “Box 5.1 The increase in the household savings rate in 2012: an explanation based on macro and microeconomic evidence”, Banco de Portugal, Annual Report - 2012.
4
43
Credit Risk
Higher savings rates, in 2012, are likely to have reflected a change in household behaviour in order to
For 2012, as a whole, net transactions of financial assets by households remained at a
reduced level …
Notwithstanding the increase in households’ financing capacity, net acquisitions of financial assets remained
44
opposed to the first half of the year, in which net transactions were negative, in the second half year,
at a reduced level, in 2012 (Chart 4.9). The evolution, in intra-annual terms, however, was irregular. As
at a time when households’ financing capacity attained its most expressive increase, net acquisitions of
financial assets were significant. Some recomposition of household portfolios was noted in both halves.
There was a significant increase over the course of 2012, in net investment by households in bonds issued
by non-financial corporations and, to a lesser degree, in debt securities issued by the banks. Additionally,
in the second half year there was a not negligible increase in investments in unquoted shares and other
equity. These movements were mainly offset, in the first half year, by a reduction of investments in life
insurance and pension funds, as was the case in 2011, and by a decline of investments in deposits in
the second half year.
Various factors were responsible for this recomposition of household portfolios. Firstly there was a sharp
appreciation in the value of long term Portuguese debt securities in 2012, in the context of a significant
decline in the perception of sovereign risk and improvement in banking system liquidity and solvency.
Secondly, taking advantage of more favourable market conditions, several non-financial corporations
issued bonds that were placed by banks on their retail customers. Lastly, there was a downwards
correction of interest rates on new deposits since the start of the year, largely determined by the ECB’s
implementation of a series of non-conventional monetary policy measures enabling banks to improve
their liquidity conditions (Chart 4.8).3 Notwithstanding, the share of deposits in household portfolios
reversed only a part of the increase registered in 2012, remaining at a very high level.
Chart 4.7
Chart 4.8
INTEREST RATES ON BANK LOANS TO
HOUSEHOLDS FOR HOUSE PURCHASES AND
FOR CONSUMPTION
INTEREST RATES ON DEPOSITS FROM
HOUSEHOLDS
14.0
Spread on new operations – housing (rhs)
Spread on new operations – consumption (rhs)
Interest rate on new operations – housing;
Interest rate on new operations – consumption
Interest rate of outstanding amounts – housing
Spread on outstanding amounts – housing (rhs)
New operations: up to 2 years
New operations: more than to 2 years
Outstanding amounts: up to 2 years
Outstanding amounts: more than 2 years
5.0
14.0
4.5
4.0
10.0
10.0
3.5
8.0
8.0
6.0
6.0
4.0
4.0
2.0
2.0
0.0
0.0
Per cent
12.0
Percentage points
12.0
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
I
3.0
2.5
2.0
1.5
1.0
Jan-99
Jan-01
Jan-03
Jan-05
Jan-07
Jan-09
Jan-11
Jan-13
Source: Banco de Portugal.
Notes: The interest rate spread on new loans to households for
house purchases is calculated using 6 months Euribor. The Interest rate spread on new loans to households for consumption is
calculated using, respectively, 6-month Euribor, 1-year Euribor
and the 5-year euro interest rate swap rate, in cases in which
the initial rate fixation period is up to 1 year, between 1 and 5
years and more than 5 years. Last observation: March 2013.
0.5
0.0
Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-12
Source: Banco de Portugal.
Note: Last observation: March 2013.
3 The Banco de Portugal’s imposition of a deduction on the own funds of the banks offering deposits with excessive interest rates also contributed to the reduction of interest rates on deposits at the end of 2011/start of
2012. This regime was reinforced in second quarter 2012 (see “Chapter 5 Liquidity Risk”, of this Report).
Chart 4.9
FINANCIAL ASSETS OF HOUSEHOLDS | TRANSACTIONS
Other debits and credits (a)
Life insurance and pension funds
Quoted shares
Securities other than shares
Total transactions of financial assets
Loans, trade advances and credits
Non-quoted shares and other equity
Mutual fund shares
Currency and deposits
45
Credit Risk
Percentage of disposable income
25
4
20
15
10
5
0
-5
-10
-15
1999 2001 2003 2005 2007 2009
2011
2011
H2
2012
H2
Sources: INE and Banco de Portugal.
Notes: Consolidated figures. (a) Includes other technical insurance reserves and other receivables.
… but there was a net repayment of debt for the second consecutive year
In financial terms, the main counterpart to households’ financing capacity was, as in 2011, net debt
repayments. Household debt as a percentage of disposable income has continued to decline gradually
since 2010. Nevertheless, the fact that it still remains at a very high level suggests that the adjustment
process on household balance sheets will have to proceed over the next few years (Chart 4.10).
The evolution of the household indebtedness ratio is underpinned by a significant drop in total credit
to households, which recorded successively greater declines between mid 2010 and mid 2012 and a
certain stabilisation in the rate of reduction in the most recent period (Chart 4.11). The reduction of total
credit reflected a particularly sharp fall in the case of bank loans for consumption and other purposes
and also significant reductions in the case of bank loans for housing, which have a higher persistence
given their longer maturities. Since the middle of 2012, in the context of an improvement in banking
system liquidity and solvency, the trend towards a significant contraction of loans to households appears
to be increasingly fuelled by demand side factors. The results of the bank lending survey suggest that
there were no additional increases in the degree of restrictiveness of the credit standards applied to
loans to households during this period. On the contrary, according to the surveyed banks, demand for
loans from households continued to fall. Nevertheless, in first the quarter 2013 there were signs of a less
marked contraction in demand, especially in the case of consumer credit and other lending. The decline
in consumer confidence and negative prospects for the housing market continued to fuel the reduction
in households’ demand for loans.
Materialisation of credit risk remains high in the case of credit for consumption and other
purposes and at a more reduced level in the case of housing loans …
The reduction of household disposable income, the deterioration of the situation in the labour market
and, in particular, the increase in the unemployment rate, continued to translate into a significant materialisation of credit risk of bank loans to households for consumption and other purposes, whose default
ratio has continued to display a markedly upwards trend since 2008 (Chart 4.12). The default ratio on
housing loans has also continued to display an upwards trend, but shows a more moderate dynamic and
relatively low levels (Chart 4.13). Factors of different nature contributed to this distinct situation. On the
one hand, housing loans tend to be concentrated in households with a lower probability of default (Chart
Chart 4.11
FINANCIAL DEBT OF HOUSEHOLDS | END OF PERIOD
CREDIT GRANTED TO HOUSEHOLDS |
POSITIONS
CONTRIBUTIONS TO THE ANNUAL RATE OF CHANGE
160
140
Bank loans - house purchase
Bank loans - consumption
Bank loans - other purposes
Other loans
3.0
Loans granted by resident banks
Loans granted by other resident financial institutions
Other loans granted by residents and trade credit
Loans granted by non-residents
Total credit (rhs)
Loans granted by resident banks for house purchases (rhs)
8.0
120
100
80
60
2.0
6.0
1.0
4.0
0.0
2.0
-1.0
0.0
-2.0
-2.0
-3.0
-4.0
-4.0
-6.0
Per cent
Percentage of disposable income
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
46
Percentage points
I
Chart 4.10
40
20
0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
Sources: INE and Banco de Portugal.
Notes: Consolidated figures.
-5.0
-8.0
2010 Q1 2010 Q3 2011 Q1 2011 Q3 2012 Q1 2012 Q3 2013 Q1
Source: Banco de Portugal.
Notes: See credit aggregates definitions in Monthly Economic
Indicators. Last observation: March 2013.
4.14).4 On the other hand, average monthly instalments have displayed a downwards trend since the
beginning of 2012 as a reflection of the significant reduction in money market interest rates (Chart 4.15).5
A slight reduction in the level of households’ annual flow of new overdue and other doubtful
loans has been noted since the first quarter of 2012
As has been the case since the onset of the sovereign debt crisis, the annual flow of new overdue and
other doubtful loans for consumption and other purposes stays at levels much higher than its average
since the inception of the euro area. In contrast, the annual flow of new overdue and other doubtful
loans in the case of housing have remained more contained during the crisis and close to the average
noted since the inception of the euro area. Both credit segments witnessed a slight dip in the flows of
new loans in default since the first quarter 2012. This evolution may have benefited from the fact that
the negative evolution of the labour market have coexisted with a reduction in interest rates. The interruption in the increase in default flows was not, however, captured by models in which such flows are
determined by GDP growth, the unemployment rate and interest rates on loans (Charts 4.12 and 4.13)6.
In a context in which money market interest rates are at very low and the unemployment rate at very high
levels, the existence of non-linear relationships not captured by the models may provide an explanation
for the divergence between the forecasted and the actual flows. Another possible explanation may lie
in the higher number of credit restructuring situations (see “Box 4.1 Banco de Portugal’s involvement
in the domain of restructured credit and preliminary analysis of recent developments”, of this Report).
4 See S. Costa, (2012), “Households’ default probability: an analysis based on the results of the HFCS”, Banco de
Portugal, Financial Stability Report – May 2012.
5 See “Box 4.1 The impact of money market interest rates on Portuguese households’ disposable income”, Banco
de Portugal, Economic Bulletin – October 2012.
6 See: N. Alves and N. Ribeiro, (2011), “Modelling the evolution of households’ defaults”, Banco de Portugal,
Financial Stability Report – November 2011.
Chart 4.12
Chart 4.13
OVERDUE AND OTHER DOUBTFUL BANK LOANS
TO HOUSEHOLDS FOR CONSUMPTION AND
OTHER PURPOSES
OVERDUE AND OTHER DOUBTFUL BANK LOANS
TO HOUSEHOLDS FOR HOUSE PURCHASES
Annual flow of new overdue and other doubtful bank loans
Model for the annual flow (anchored in 2009 Q1)
Default ratio (rhs)
12.0
2.5
10.0
2.0
8.0
Per cent
Per cent
2.5
2.0
0.3
1.5
0.2
1.5
6.0
1.0
4.0
0.1
0.5
2.0
0.5
0.0
Jan-99 Jan-01 Jan-03 Jan-05 Jan-07 Jan-09 Jan-11 Jan-13
1.0
0.0
0.0
Jan-99 Jan-01 Jan-03 Jan-05 Jan-07 Jan-09 Jan-11 Jan-13
Source: Banco de Portugal.
Notes: See credit risk indicators definitions, footnote 1. See
model in Alves and Ribeiro, (2011), “Modelling the evolution
of households’ defaults”, Banco de Portugal, Financial Stability
Report - November. Last observation: March 2013.
Credit Risk
3.0
0.5
-0.1
Per cent
14.0
0.4
Per cent
47
Annual flow of new overdue and other doubtful bank loans
Model for the annual flow (anchored in 2009 Q1)
Default ratio (rhs)
3.5
4
0.0
Source: Banco de Portugal.
Notes: See credit risk indicators definitions, footnote 1. See
model in Alves and Ribeiro, (2011), “Modelling the evolution
of households’ defaults”, Banco de Portugal, Financial Stability
Report - November. Last observation: March 2013.
Although there are still risks of new increases in the materialisation of credit risk…
The divergent behaviour of default flows observed and predicted by their usual determinants suggests
that the sustainability of the recent developments should be viewed with caution. The outlook for the
Portuguese economy, published in the Spring Economic Bulletin, embody the persistence of adverse
conditions in the labour market, a further reduction of disposable income in 2013 and contemplate
downside risks on households’ financial situation deriving in particular from the implementation of
additional fiscal adjustment measures. In such circumstances the possibility of an increase in credit risk
should not be excluded, even if interest rates on loans to households continue to fall.
…there are several mitigating factors which may contain the dimension of this
materialisation
The materialisation of credit risk is particularly significant in the case of households with the highest
debt ratios. It should, herein, be noted that, according to the results of the Household Finance and
Consumption Survey (HFCS) for euro area countries, recently published by the ECB, the indebted Portuguese households with the lowest income levels had a much higher typical level of indebtedness than
households in the euro area with the same characteristics, in 2010 (Chart 4.16 (a), (b) and (c))7. However,
one factor mitigating the implications of this situation for the stability of the Portuguese financial system
consists of the fact that Portuguese households with the lowest income levels account for a relatively
reduced share of the credit market (Chart 4.16 (d)). Other important factors which attenuate the risk
of an increase in defaults in loans to households in Portugal are the fact that loans for housing, which
represent around 80 per cent of loans to households and approximately half of Portuguese banks’ credit
7 The HFCS is a harmonised survey on the financial situation of households which is the responsibility of the
Eurosystem. In April 2013, the ECB released the results at the euro area level for the first wave of the survey, in
which the reference period for most countries is 2010. The Portuguese version of this survey corresponds to the
Inquérito à Situação Financeira das Famílias (ISFF), developed by Banco de Portugal and INE. The results of the
ISFF 2010 were published in May 2012.
DISTRIBUTION OF THE OUTSTANDING AMOUNTS
ON HOUSEHOLDS’ LOANS BY PROBABILITY OF
DEFAULT PERCENTILE | DATA FOR THE SECOND QUARTER
AVERAGE MORTGAGE INSTALMENT
OF 2010
Total interest
Repayment of capital
100
450
400
75
350
300
50
EUR
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
48
Chart 4.15
Per cent
I
Chart 4.14
250
200
25
150
100
0
Total
Mortgages
50
Non-mortgage loans
Type of loans
<25
25-50
50-75
75-90
0
Jan-09 Jul-09 Jan-10 Jul-10 Jan-11 Jul-11 Jan-12 Jul-12 Jan-13
>=90
- dados do segundo trimestre de 2010-
Source: INE and Banco de Portugal (Household Finance and
Consumption Survey).
Note: See S. Costa, (2012), “Households´ default probability:
an analysis based on the results of the HFCS”, Banco de Portugal, Financial Stability Report – November.
Source: INE.
Note: Last observation: March 2013.
portfolios relating to the non-financial private sector, are reasonably well collateralised (see Box 4.3 “The
loan-to-value ratio in the residential mortgage market in Portugal”, of this Report) and concentrated in
households with a lower credit risk. Reference should, lastly, be made to the fact that credit institutions’
application of the new legislation entering into force at the end of 2012/start of 2013 on the prevention
and management of default situations might help to attenuate the materialisation of credit risk and its
implications for the financial situation of the more vulnerable households.
Non-financial corporations
In 2012, the profitability of non-financial corporations remained low and indebtedness high
The risks associated with the fragility of the financial situation of non-financial corporations remained
high over the course of 2012. Corporate activity once again contracted significantly and profitability
remained at low levels. This situation resulted from a marked contraction of domestic demand owing
to a series of fiscal adjustment measures and a slowdown in external demand, given the sluggishness
of the activity of trading partners, particularly in euro area countries. The risks related with the financial
situation of non-financial corporations are also associated with their continued high indebtedness levels
(Chart 4.17). Notwithstanding the fact that the trend on interest rates on loans to non-financial corporations is downwards, they remained high. The fact that the value of interest payments are significant
in comparison to corporate income levels conditions financial autonomy of corporations as well as their
access to financing from external sources. The permanent reduction of the level of corporate debt is a
sine qua non for the consolidation of the adjustment process although this is a slow process with high
short term costs. In order to encourage corporate deleveraging, the deductibility of interest paid by
companies for tax purposes has been under discussion. In this scope, the State Budget for 2013 foresees
a change to the IRC code that limits the amount of tax deductible interest costs.
Chart 4.16
HOUSEHOLDS’ INDEBTEDNESS AND PARTICIPATION IN DEBT
(a) Median debt-income ratio
80
350
60
250
49
Credit Risk
400
70
Per cent
Per cent
450
300
4
(b) Median debt service-income ratio
90
Euro area
Portugal
500
50
40
200
30
150
20
100
10
50
0
0
Total
<20
20-40
40-60
60-80
80-90
>90
Total
<20
Income percentile
(c) Median debt-wealth ratio
50
20-40
40-60
60-80
Income percentile
80-90
>90
(d) Percentage of households holding debt
In per cent of the number of households in each class
70
45
40
Per cent
35
30
25
20
15
10
5
0
60
50
40
30
20
10
0
Total
<20
20-40
40-60
60-80
Income percentile
80-90
>90
Total
<20
20-40
40-60
60-80
Income percentile
80-90
>90
Source: ECB (Household Finance and Consumption Survey).
Investment down and savings up, translating into a reduction of companies’ borrowing
requirements
The adjustment of the financial situation of companies is particularly evident in an analysis of financing
flows and their underlying factors. In 2012, the borrowing requirements of non-financial corporations
were, once again, significantly down from 4.5 per cent to 2.1 per cent of GDP (Chart 4.18). This evolution
resulted from a sharp drop in investment, particularly GFCF, reflecting expectations of a highly depressed
evolution of domestic demand, low use of productive capacity and, to a lesser extent, the maintenance
of restrictive financing terms. There was also a recovery in the savings rate of non-financial corporations,
largely as a result of the increase in the gross operating surplus, for which the reduction of compensation
of employees contributed positively (Chart 4.19). As stated above, the trend on interest rates on loans
to non-financial corporations over the course of the year was downwards but not enough to offset the
marked rise noted up to the end of 2011, so that interest paid minus interest received was once again
up over the preceding year (Chart 4.20 and Chart 4.21). The savings of non-financial corporations are
at levels close to those noted in the period immediately preceding the onset of the international financial crisis. They remain, however, clearly below the levels registered at the time of the inception of the
euro area (Chart 4.21). Notwithstanding the fact that the contributions to savings made by the gross
operating surplus and interest, as a percentage of GDP, are at levels comparable to those at the time
of the inception of the euro area, the contribution to saving made by corporate dividends is currently
much more negative.
Chart 4.17
Chart 4.18
DEBT OF NON-FINANCIAL CORPORATIONS | END
NET BORROWING, SAVING AND INVESTMENT OF
NON-FINANCIAL CORPORATIONS
OF PERIOD FIGURES
I
160
Total debt(a)
Financial debt (b)
Net lending / borrowing
Gross saving
Net capital transfers
Acquisitions less disposals of real assets (a)
15
140
10
Percentage of GDP
Percentage of GDP
120
100
80
60
5
0
-5
40
-10
20
-15
0
-20
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
1999
2002
2005
2008
2011
2011-H1 2012-H2
Sources: INE and Banco de Portugal.
Notes: Consolidated figures. (a) Total debt = financial debt
+ trade credits and advances received from other sectors. (b)
Financial debt = loans received + debt securities issued by non-financial corporations.
Source: INE.
Notes: (a) Corresponds to the sum of gross fixed capital formation, changes in inventories and acquisitions less disposals
of valuables and acquisitions less disposals of non-produced
non-financial assets.
Chart 4.19
Chart 4.20
INTERNAL FINANCING OF NON-FINANCIAL
CORPORATIONS | 4 QUARTERS MOVING AVERAGE
INTEREST RATE ON BANK LOANS TO NONFINANCIAL CORPORATIONS
8.0
18.5
9
18
8
17.5
7
17
6
16.5
5
16
4
2003-Q1
2006-Q1
2009-Q1
2012-Q1
7.0
4.5
Interest rate on
outstading
bank loan amounts
Spreas (rhs)
5.0
4.0
3.5
6.0
Per cent
10
3
2000-Q1
9.0
19
Gross operating surplus (rhs)
3.0
6 month
Euribor
2.5
4.0
2.0
3.0
1.5
2.0
1.0
15.5
1.0
0.5
15
0.0
Jan-99
Jan-01
Jan-03
Jan-05
Jan-07
Jan-09
Jan-11
Percentage points
Gross saving
Percentage of GDP
11
Percentage of GDP
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
50
0.0
Jan-13
médias móveis de 4 trimestres
Source: INE.
Source: Banco de Portugal.
Notes: Rates and spread refer to end of period outstanding
amounts. Up to December 2002, the rates on the outstanding
amounts are estimated. The spread is calculated as the difference between the rate on the outstanding amounts and the
6-month moving average of 6-month Euribor. Last observation:
March 2013.
Chart 4.21
CONTRIBUTIONS TO THE GROSS SAVING OF NON FINANCIAL CORPORATIONS
4
10
9
20
51
8
7
10
6
5
5
4
0
Percentagem do PIB
Percentagem do PIB
15
Gross operating surplus
Property income (net)
Distributed income of corporations (net)
Interest (net)
Other property income (net)
Taxes on income and wealth
Other current transfers (net)
Gross saving (rhs)
3
-5
2
-10
1
-15
1999
2002
2005
2008
2011
0
2011-S1 2012-S2
Source: INE.
Notes: The term net refers to the difference between the values registered in resources and the values registered in uses. In quarterly
national accounts the information about the components of property income (distributed income of corporations, interest and other
property income) is not available.
Corporate performance weak but stable
There was a high level of heterogeneity in the evolution of corporate profitability in accordance with
companies’ sector of activity, size or export propensity. Microeconomic accounting information for 2012
is only available for a corporate sample, which is biased towards the largest companies8 (which may be
particularly important in the case of companies in the trade sector). The rate of change in gross margin
continued to be negative for the various corporate subsets and relatively stable over the course of the
year. In the fourth quarter, whose profitability indicators reflect corporate results for the year as a whole,
operating profitability was similar to the same period of 2011 in the various groups of companies analysed,
except for manufacturing. Manufacturing companies were also those with a wider reduction of the
interest coverage ratio, which remained relatively stable in the case of private companies (Chart 4.22).
Corporate lending criteria stabilise but remain restrictive
The continuation of negative expectations on the evolution of economic activity in general and the strong
inertia in the deleveraging process on non-financial corporations, especially in specific sectors, justify the
persistence of restrictive lending criteria. There are, however, signs that some banks have already begun
to ease their lending criteria. Spreads on new loans to non-financial corporations have stabilised at a
high level following the increasing trend noted over the course of the last three years. This conclusion
is also confirmed by the results of the Bank Lending Survey which indicate a progressive stabilisation of
lending criteria over the course of the year.
8 The indicators are calculated on the basis of a sample common to the two years.
Credit Risk
25
Chart 4.22
PERFORMANCE OF THE PRIVATE NON-FINANCIAL CORPORATIONS
I
15.0
16
15
10.0
14
Per cent
5.0
Per cent
0.0
13
12
-5.0
11
-10.0
10
-15.0
9
-20.0
8
2011 I 2011 II 2011 III 2011 IV 2012 I 2012 II 2012 III 2012 IV
2011 I 2011 II 2011 III 2011 IV 2012 I 2012 II 2012 III 2012 IV
Interest coverage ratio
11.0
10.0
9.0
8.0
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
52
Profitability ratio
Gross margin
(year-on-year rate of change)
7.0
6.0
5.0
4.0
3.0
2.0
2011 I 2011 II 2011 III 2011 IV 2012 I 2012 II 2012 III 2012 IV
Private
Exporters
Trade
Manufacturing
Source: Banco de Portugal (Quarterly Central Balance Sheet).
Notes: Gross margin= Turnover+Operating subsidies+Variation in production+Capitalized production-Costs of goods sold and material consumed-Supplies and external services; Profitability ratio=EBITDA/(Shares and other equity+Loans and debt securities); Interest
coverage ratio = EBITDA/Interest paid; EBITDA=Profit before taxation+interest paid+depreciation and amortization. The indicators are
derived from samples with the same corporations in the two years.
Marked divergence between the evolution of lending to companies by resident banking
sector and credit obtained from other sources
Total credit9 to non-financial corporations was down 0.6 per cent at the end of 2012 (Chart 4.23).
Lending by the resident banking sector, in the form of loans or securities, was down by around 6 per
cent in the same period. This evolution was mitigated by the positive contributions of lending by non-residents and, at a smaller extent, the non-financial resident sector. The intra-annual profile on both
the evolution of total credit and lending by the resident banking sector, showed an interruption of the
deterioration trend since the end of the year.
9 Total credit to non-financial corporations includes loans made by resident banks, loans made by non-residents,
debt securities (held by residents and non-residents), trade credit (made by residents and non-residents), loans
made by households and Treasury loans, the latter being relevant in the case of public corporations.
In the case of private companies, total credit and credit made by the resident banking sector fell 1.4
per cent and 7.0 per cent, respectively, at the end of 2012. In this case, the largest contribution to the
sectors also made a positive contribution until the end of 2012. In the case of state owned companies,
total credit and credit made by the resident banking sector grew 4.0 percent and 18.5 percent, respectively, at the end of 2012. In this case, the non-resident sectors made a negative contribution to credit
growth, with the resident banking sector making a broadly positive contribution, which was more marked
at the start of 2013 (Chart 4.25).
In the context of an across-the-board increase in risk perception, the divergences noted in the evolution
of credit largely reflect the different risk profiles associated with the various types of companies. Private
companies more reliant on the domestic market, such as those in the construction sector, the smaller
and the younger, have greater difficulty in the access to bank credit.
There is therefore a high level of heterogeneity in the evolution of credit by sector of activity. Total credit
(excluding trade credit) to “construction”, “real estate activities”, and, to a lesser extent, ”trade”, was
markedly down in 2012 (Table 4.1). In contrast, credit posted an increase in “information and communication”, “electricity, gas and water” and “non-financial holding companies”, in which larger companies
with access to the international financial markets predominate. Nevertheless, these sectors also registered
negative changes in lending from the resident banking sector.
Contraction of credit to smaller companies reflects the high share of “construction”, “real
estate activities” and “trade” in these categories
A breakdown of total credit (excluding trade credit) by corporate size indicates that lending to large firms
and holding companies maintained positive rates of change albeit decreasing slightly over the course of
2012 and the beginning of 2013 (Chart 4.26) (in March 2013, 3.0 per cent and 3.5 per cent for large
firms and holding companies, respectively). In the case of companies in the other size categories, the
rate of change of credit remained negative. The evolution of credit by corporate size is strongly associated with the importance of the different sectors of activity in the usually considered size categories10.
“Construction”, “real estate activities” and “trade” account for a much higher proportion of lending to
smaller companies than of lending to large companies. Therefore, the significant decline in total credit
to small and medium sized enterprises (including micro companies) (in February 2013, -3.7 per cent) is
likely to be related to the high share of “construction”, “real estate” and “trade” in smaller companies. Excluding these three sectors, the annual rate of change of total credit to small and medium-sized
enterprises (including micro companies) has been substantially less negative since the end of 2012 (in
February 2013, -1.2 per cent). Additionally, the quarterly flows are positive, since January 2013 (Chart
4.27). Similarly, the larger size and the lower dependence on bank credit of companies in “electricity,
gas and water”, “transportation and storage” and “non-financial holding” is likely to have contributed
to the fact that these sectors posted a higher annual rate of change of total credit.
Increased materialisation of credit risk, albeit less marked from mid 2012
Corporate profitability has remained low in the context of a sharp fall in domestic demand, more
sluggish external demand and high financing costs. Younger, smaller and more leveraged companies
were particularly affected by difficulties in access to financing, in particular to provide for their short
term liquidity management. The increasing trend in the materialisation of credit risk in the case of non-financial corporations therefore continued over the course of 2012. The default and non-performing
10 See “Box 2.4 Differentiation in credit to non-financial corporations with cross-referencing of size and sector of
activity”, Banco de Portugal, Annual Report 2012.
4
53
Credit Risk
evolution of total credit was again made by the non-resident sector (Chart 4.24). Resident non-financial
CREDIT GRANTED TO NON-FINANCIAL
CORPORATIONS | CONTRIBUTIONS TO THE ANNUAL RATE
CREDIT GRANTED TO PRIVATE NON-FINANCIAL
CORPORATIONS | CONTRIBUTIONS TO THE ANNUAL RATE
OF CHANGE
OF CHANGE
6
6
4
4
4
4
2
2
2
2
0
0
0
0
-2
-2
-2
-2
-4
-4
-4
-4
-6
-6
-6
-6
-8
-8
-8
2010
Q1
2010
Q3
2011
Q1
2011
Q3
2012
Q1
2012
Q3
Per cent
Percentage points
6
Percentage points
54
6
2013
Q1
Per cent
Chart 4.24
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
I
Chart 4.23
-8
2010
Q1
2010
Q3
2011
Q1
2011
Q3
2012
Q1
2012
Q3
2013
Q1
Chart 4.25
12
24
10
20
8
16
6
12
4
8
2
4
0
0
-2
-4
-4
-8
-6
-12
-8
-16
-10
-20
-12
Per cent
Percentage points
CREDIT GRANTED TO STATE-OWNED ENTERPRISES (NON-CONSOLIDATING IN GENERAL GOVERNMENT)
| CONTRIBUTIONS TO THE ANNUAL RATE OF CHANGE
-24
2010
Q1
2010
Q3
2011
Q1
2011
Q3
2012
Q1
2012
Q3
2013
Q1
Loans - resident banks
Loans - other resident financial institution
Debt securities - residents
Trade credit and loans by other resident sectors
Loans, debt securities and trade credit - non residents
Total credit (rhs)
Loans and debt securities - resident banks (rhs)
Source: Banco de Portugal.
Notes: See credit aggregates definitions in Monthly Economic Indicators. Last observation: March 2013.
Table 4.1
CREDIT TO NON-FINANCIAL CORPORATIONS BY SECTOR OF ACTIVITY | ANNUAL RATE OF CHANGE
Total credit
Memo
2012
2013
2011
2012
2013
Dec.
Dec.
Mar.
Dec.
Dec.
Mar.
-2.4
-5.8
-5.7
0.4
-0.3
0.2
100.0
50.6
-3.1
-7.8
-6.2
-2.4
-1.2
-0.7
10.9
66.4
Electricity, gas and water
5.3
-4.5
-13.1
5.6
6.5
-0.9
9.0
32.9
Construction
-3.2
-8.7
-9.5
-1.9
-6.8
-8.2
13.8
68.9
Wholesale and retail trade (incl.
repair of motor vehicles and
motorcy)
-6.0
-10.6
-8.1
-1.2
-5.0
-3.7
10.9
57.3
Transportation and storage
5.2
7.6
16.2
8.5
2.6
10.4
8.9
48.2
Accomodation and food service
activities
9.4
-4.0
-3.5
4.8
-3.1
-3.6
3.5
68.1
Information and communication
-23.5
-8.5
-10.0
-25.8
24.8
15.4
2.8
24.4
Non-financial holdings
-6.3
-6.6
-10.2
5.3
5.4
3.7
18.2
42.6
Real estate activities
-5.4
-3.8
-3.8
-2.0
-3.7
-4.5
10.7
62.0
Professional, scientific, technical
and administrative activities
4.5
-15.1
-13.4
7.3
-5.0
-3.9
6.4
50.6
Education, human health, social
work and other personal activities
-4.2
-4.7
-3.9
-7.0
-1.8
-2.2
3.0
65.6
Other activities
3.6
4.0
-2.6
-27.0
0.5
-0.8
1.7
60.3
Total
Total
Bank
credit
credit
sector
sector
(% total (% total
credit)
credit)
Dec.2012 Dec.2012
Sectors
Manufacturing (incl. mining and
quarrying)
Source: Banco de Portugal.
Notes: See credit aggregates definitions in Monthly Economic Indicators. Banking credit includes all credit extended by resident
banks, which includes in addition to loans (adjusted for securitisations), debt securities held by banks. There is no data available
regarding trade credit by sector of activity.
Chart 4.26
CREDIT GRANTED TO NON-FINANCIAL CORPORATIONS | ANNUAL RATE OF CHANGE
25.0
SME (including micro-corporations)
Large corporations
Non-financial holdings
20.0
Per cent
15.0
10.0
5.0
0.0
-5.0
-10.0
Dec-08 Jun-09 Dec-09 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-12
Source: Banco de Portugal.
Notes: The credit by non-financial corporation’s dimension includes loans and debt independently of who conceives the credit. The
data presented excludes trade credit given that this is not available by firm size. Last observation: March 2013.
4
55
Credit Risk
Banking credit
2011
Chart 4.27
CREDIT GRANTED TO SMALL AND MEDIUM NON-FINANCIAL CORPORATIONS (INCLUDING MICRO)
I
Annual rate of change (a)
20
Non-financial corporations (excl. construction, real estate and trade)
Construction, real estate and trade
4 000
Non-financial corporations (excl. construction, real estate and trade)
Construction, real estate and trade
3 000
15
2 000
EUR million
10
Per cent
5
1 000
0
0
-1 000
-5
-2 000
-10
Dec-08 Jun-09 Dec-09 Jun-10 Dec-10
Jun-11
Dec-11
Jun-12 Dec-12
-3 000
Dez-10
Jun-11
Dez-11
Jun-12
Dez-12
Source: Banco de Portugal.
Notes: Credit includes loans and debt securities independently of who conceives the credit. The data presented excludes trade credit
given that this is not available by sector of activity and firm size. (a) The annual rate of change is calculated as the rate of change of
the outstanding amounts of total credit to non-financial corporations. (b) The quarterly flow was not seasonally adjusted given that
there are very few observations available. Last observation: February 2013.
loans ratios, in 2012, reached successive maximum levels since the beginning of the euro area (Chart
4.28). The most recent evolution of these indicators shows, however, a smoother profile from mid year
onwards. Additionally, downwards trend in the annual flow of new overdue and other doubtful loans
was been observed in the same period. The results of an estimated model on the average probability
of default of non-financial corporations also indicate a downwards trend in the probability of default
(Chart 4.29). In the case of major exposures, which do not necessarily correspond to the exposures of
the largest companies, probabilities of default have been, since the middle of 2011, higher than in the
case of small exposures and above the values estimated based on the relevant macroeconomic variables11.
The increase in the default ratio, in 2012, was transversal to most sectors of activity. However, “construction”, “real estate activities” and “trade” made the largest contribution to the increase in the default
Chart 4.28
OVERDUE AND OTHER DOUBTFUL BANK LOANS TO NON-FINANCIAL CORPORATIONS
Annual flow of new overdue and other doubtful bank loans
Default ratio (rhs)
4.5
12.0
4.0
10.0
3.0
2.5
8.0
2.0
6.0
1.5
1.0
4.0
0.5
0.0
2.0
Jan-99 Jan-01 Jan-03 Jan-05 Jan-07 Jan-09 Jan-11 Jan-13
-0.5
-1.0
Source: Banco de Portugal.
Notes: See credit risk indicators definitions, footnote 1. Last observation: March 2013.
11 See A. Antunes, (2012), “Modelling default ratios with micro data”, mimeo.
0.0
Per cent
3.5
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
56
Quarterly flow (b)
Chart 4.29
CREDIT DEFAULT PROBABILITY FOR LOANS TO NON-FINANCIAL CORPORATIONS
Small exposures
Observed
Model
0.040
Large exposures
0.045
Observed
Model
0.040
0.035
0.035
0.030
0.030
0.025
0.025
0.020
0.020
0.015
0.015
0.010
0.010
0.005
0.005
0.000
0.000
Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar99 00 01 02 03 04 05 06 07 08 09 10 11 12
4
57
Credit Risk
0.045
Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar- Mar99 00 01 02 03 04 05 06 07 08 09 10 11 12
Source: Banco de Portugal.
Notes: Values in natural units (quarterly). The model uses, among other regressors, the GDP growth rate (quarter on quarter) and
the variation in the unemployment rate. Loans were classified based on each firm total exposure. Small exposures correspond to
exposures of less than 1 million euros and large exposures correspond to total exposures of more than 1 million euros.
ratio of non-financial corporations. These sectors, particularly “construction”, have higher indebtedness
and lower profitability than the average and much higher default indicators. Moreover as a whole they
represent around 45 per cent of total loans to non-financial corporations but account for around 70 per
cent of total credit in default (Chart 4.30). This disparity is more pronounced in “construction” which
has a share of less than 20 per cent on total loans and of almost 40 per cent on total credit in default.
Large enterprises have a much lower default level than other size categories (Table 4.2).
The differences in the risk profiles of companies in different sectors of activity and size categories appear
to translate into divergences in the evolution of lending. This is evident in the fact that more marked
decreases in lending were noted in “construction”, “real estate activities” and smaller companies. Surveys
on companies in “manufacturing”, “construction” and “services” accordingly show that the percentage
of companies which consider that the difficulty of obtaining bank loans is one of the main constraints
on their activity is much higher in the case of construction companies (Chart 4.31).
Table 4.2
DEFAULT INDICATORS ON LOANS TO NON-FINANCIAL CORPORATIONS, BY CORPORATION SIZE AND
TYPE OF FINANCIAL INSTITUTION | PER CENT
Number of debtors in default (a)
Memo
Weight
on total
(em Mar13)
Dec-12
Mar-13
Dec-11
Dec-12
Mar-13
22.9
27.1
28.4
6.7
10.0
10.9
100.0
Micro corporations
23.8
27.8
29.0
10.6
14.9
16.6
33.2
Small corporations
19.4
24.6
26.3
6.4
10.8
11.9
23.7
Medium corporations
18.4
24.2
25.7
4.6
8.1
8.9
24.6
Large corporations
12.2
15.3
17.1
1.5
2.1
2.3
18.5
58
Loans granted by monetary financial
institutions
Loans granted by non-monetary
financial institutions
28.5
36.5
37.2
16.5
23.3
24.5
100.0
Micro corporations
30.8
38.8
39.1
23.2
28.5
29.2
32.3
Small corporations
24.6
33.4
35.1
20.8
28.6
30.1
26.1
Medium corporations
21.7
28.5
30.1
13.6
21.5
22.8
25.5
Large corporations
10.6
15.0
15.3
2.3
7.6
8.4
16.1
Source: Banco de Portugal
Notes: Indicators based on information from the Central Credit Register (CRC). Includes loans granted by banks, savings banks,
mutual credit agricultural institutions, financial credit institutions, factoring companies, leasing companies, credit card issuing or
management companies and other resident financial intermediaries. Does not include loans granted to non-financial holdings. (a)
As a percentage of the number of non-financial corporations with debts to monetary financial institutions or non-monetary financial
institutions participating in the CRC. (b) As a percentage of the total credit from monetary financial institutions or non-monetary
financial institutions participating in the CRC to resident non-financial corporations.
Chart 4.30
Chart 4.31
OVERDUE AND OTHER DOUBTFUL BANK LOANS
TO NON-FINANCIAL CORPORATIONS | BY BRANCH
PERCENTAGE OF FIRMS WITH DIFFICULTIES IN
THE ACCESS TO CREDIT (a)
OF ACTIVITY
25
Default ratio (Mar 13)
Default ratio (Mar 12)
Share of total credit (rhs)
Share of total credit in default (rhs)
60
Manufacturing
Services
Construction
40
50
35
20
30
40
20
10
15
Per cent
25
15
Per cent
Per cent
30
10
5
5
20
0
Transport
Manufacturing
Restaurant. and
hotels
Trade
Real estate
Construction
0
Total
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
Overdue credit and interest (b)
Dec-11
Other
I
Source: Banco de Portugal.
Note: See credit risk indicators definitions, footnote 1.
10
0
2009-I 2009-IV 2010-II 2010-IV 2011-II 2011-IV 2012-II 2012-IV
Source: INE.
Note: (a) Percentage of firms declaring that difficulties in the
access to credit are one of the main limits to their activity.
BOX 4.1 | RESTRUCTURED CREDIT: BANCO DE PORTUGAL’S ACTION
AND A PRELIMINARY ANALYSIS OF RESULTS
4
Background
assessment of credit risk in the banking sector. Accordingly, Banco de Portugal has recently issued an
Instruction (Instruction nº18/2012) prescribing the obligation of credit institutions (and branches of credit
institutions operating in Portugal, with head offices in countries outside the European Union) to identify
and mark in their lending systems the loans whose contractual conditions have been changed (e.g., an
extending of the repayment period, reduction of interest rate and the introduction of grace periods)
owing to a customer’s financial difficulties. In accordance with this regulation, a customer is considered
to be in financial difficulty when a given commitment with the institution has not been met or is likely
not to be met, taking, among other, the following into account:
i. Overdue amounts registered in the Central Credit Register;
ii. Inclusion in the official list of misusers of checks due to the issuance of dishonoured checks and
the ensuing inhibition of further use of this payment mean;
iii. Activation of internal alerts (e.g., a marked deterioration of internal risk classification);
iv. Qualitative events (e.g., debts to the tax authorities and social security system, calling-in of bank
guarantees, bankruptcy, insolvency, judicial processes and legal disputes, arrears of wages, pledges
on bank accounts, etc.).
Instruction nº18/2012 also states that whenever a restructured credit operation represents more than 25%
of total exposure to a single customer, all credit operations with the said customer should be identified
and marked as restructured credit.1
Subsequently, the On-site Inspections Programme (OIP) endeavoured to monitor progress in the implementation of the requirements of Instruction nº18/2012 on the identification and reporting of restructured
credit in the construction and real estate sectors. The programme’s main objective was to assess the
adequacy of impairment levels on the sectors in question. In this context, Banco de Portugal has issued a
series of recommendations with the aim of eliminating the main defects found, over the medium term.
Restructured credit ratio by credit segment for the eight biggest banking groups
Based on the fact that Instruction nº18/2012 introduced the need to identify and mark restructured
credit and that it provides for the possibility that such a marking process could be realised in stages, the
first periods reported should be analysed with prudence, as the practice of marking this type of credit
for some banks had still not been established in the banks’ internal procedures. This box is therefore
centred on December 2012 results.
In December 2012, the restructured credit ratio totalled 5.8 percent, although there is a high level of
heterogeneity among segments of the credit portfolio. The restructured credit ratio in the credit for
consumption and other purposes segment was 11.5 percent, which compares to 1.8 percent for residential mortgages. In turn, the ratio for non-financial corporations was 9.8 percent.
The current demanding economic background calls for special attention on the developments of restructured credit, going forward.
1 Banco de Portugal latterly changed the Instruction regarding the information on non-performing loans (Instruction nº24/2012, which changed Instruction nº22/2011) for the purposes of incorporating the information on
restructured credit per segment of the loan book. The reporting of such information began in October 2012
with not only to September 2012, on a quarterly basis, but also some historical data (from June 2011 to June
2012).
59
Credit Risk
At the current juncture, the monitoring of the evolution of restructured loans plays a critical role in the
I
Chart 1
RESTRUCTURED CREDIT RATIO(a) | IN PERCENTAGE OF CREDIT GRANTED
14
12
10
8
In percentage
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
60
6
4
2
0
Non-financial
corporations
Source: Banco de Portugal.
Note: (a) On a consolidated basis.
Housing
Consumption
and other
purposes
Other
Non-residents Restructured
credit ratio
(Global)
BOX 4.2 | OVERVIEW OF THE HOUSE PRICE INDEX PRODUCED BY THE
INE
4
In the context of the current economic and financial crisis, statistics related with the real estate market,
for instance, residential and commercial property price indices have grown in importance in terms of
the monitoring of economic evolution and assessment of the stability of the financial system. The fact
that the European Commission has included the House Price Index – HPI in the restricted collection of
indicators to be monitored under the scope of the supervisory, early warning, detection and prevention
of excessive macroeconomic imbalances (Regulation (EU) 1176/2011 - Scoreboard for the surveillance
of macroeconomic imbalances) illustrates the overall relevance and appropriateness of these statistics.
House price indices available in Portugal
For a relatively long period, there was only one provider of price statistics in the housing market with
the characteristics of an index. This index, the Confidencial Imobiliário Index - Ci Index, is produced by a
private company and dates back to 1988 with monthly frequency. Since 2005, it has been constructed
on the basis of a hedonic prices methodology, i.e., using regression techniques which adjust for quality
effects, including, inter alia, geographical location, occupancy areas, number of rooms and the age
of properties. The Ci Index’s primary sources are asking prices as reported by agencies and other real
estate operators.
INE provides an index based on the appraisal prices reported by a sample of banks when processing
housing loans’ applications. This sample represents the overwhelming majority of housing mortgages
in Portugal. This index, the HPI index, is built by stratification based on Censos. In accordance with
Eurostat requirements for the construction of official national indices, ideally the primary sources should
be based on transaction prices, a situation with which neither INE, nor any other public or private entity
fully complies at the present time.
Comparison between HPI and CI Index
Notwithstanding the fact that the timeframe covered by the HPI is still short, the results obtained in
comparison to the Ci Index, suggest various conclusions: (1) the HPI index displays higher amplitude
of variation; (2) the year-on-year rate of change of the Ci Index lags the HPI by between one and two
quarters; (3) both indices display a consistent decrease since the period in which the sovereign debt
crisis in Portugal became more acute, more specifically with the inception of the economic and financial
assistance programme.This decrease is much more marked in the case of the HPI (Table 1 and Chart 1).
In contrast to other countries, there was no real estate bubble in Portugal in the period immediately
before the current recession. However, in the more recent period, a reduction of prices in the housing
market, related with the evolution of the fundamentals, has been noted. The marked drop in economic
activity has brought significant uncertainty regarding growth prospects over short, medium and long
terms. This situation has translated into a reduction of activity in the real estate market, partly for reasons
associated with supply in the mortgage market. In this context, the fact that the HPI index has a wider
range of variation with a much higher reduction, is likely to be associated with the fact that asking
prices (used as input for the Ci Index) tend to be resistant to downwards revisions in periods in which
transaction values are actually declining. In turn, it can be argued that valuations have been influenced
by banks’ lending policies, which are pro-cyclical in nature. However, the relevance of the mortgage
market to the functioning of the housing market sustains the idea that bank valuations should be one
of the most important factors in terms of the formation of prices in this market.
61
Credit Risk
Background
Table 1
HPI AND CI INDICES - YEAR-ON-YEAR RATE OF CHANGE | PERIOD 2008T1 - 2012T4
I
Periods
HPI Index
Ci Index
t; t+2
0.87
Maximum
2.3
3.0
t; t+1
0.86
Minimum
-8.3
-2.6
t; t
0.70
Max.-Min.
10.6
5.6
Sources: Confidencial Imobiliário and INE.
Chart 1
Chart 2
INDEX 2008Q1=100
YEAR-ON-YEAR RATE OF CHANGE
105
4.0
2.0
100
-4.0
90
-6.0
Ci Index
HPI
85
Mar-08
-8.0
Mar-09
Mar-10
Mar-11
Sources: Confidencial Imobiliário and INE.
Mar-12
Mar-13
Ci Index
HPI
-10.0
Sources: Confidencial Imobiliário and INE.
Mar-13
Mar-12
Mar-11
-2.0
95
Mar-10
0.0
Mar-09
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
62
Correlation HPI (t)
vs Ci (t+j)
BOX 4.3 | THE LOAN TO VALUE RATIO IN THE RESIDENTIAL MORTGAGE
MARKET IN PORTUGAL
sents the bulk of household wealth. Changes in the prices of these assets affect consumption and savings
decisions and these effects will be all the more relevant the higher the proportion of housing-related debt.
Properties are also used to guarantee housing loans and, as such loans represent a significant proportion
of the activity of Portuguese banks, developments in the housing market have a bearing on credit quality.
From an aggregate viewpoint, the part of the value of housing financed by credit (loan to value – LTV) at
the time of the acquisition of the property is an indicator of the tightness of the loan approval criteria.
Therefore, periods with higher levels of leverage on these operations, i.e., higher LTV, are associated
with increases in competition and strong levels of activity in the housing and mortgage markets. In this
context, the monitoring of this indicator through the cycle enables the detection of accumulations of
vulnerabilities in banks’ credit portfolios. To be sure, the existence of statistics on the distribution of
LTVs in banks’ portfolios is especially important, with situations of greater fragility tendentially found in
operations with the highest indicator.
As shown in Chart 1, the information collected from the eight major banks allows to conclude that
a progressive increase in the average LTV ratios, measured at origination, occurred between the early
2000’s and the onset of the financial crisis in 2007. The fact that this has been accompanied by lower
spreads on new loans, also suggests greater competition in this segment. The evolution noted since 2007
illustrates an increase in banks’ liquidity and capital restrictions and a more pessimistic assessment of the
outlook on the evolution of economic activity and the housing market, having an effect on households’
debt servicing capacity.
The available qualitative information in the Bank Lending Survey illustrates more restrictive loan criteria
starting mid 2007, in addition to lower demand for bank lending, in a context of high levels of uncertainty and a significant contraction of economic activity. These, together with higher unemployment and
tighter fiscal policy, translated into a strong decline in purchasing power and less favourable prospects
for the evolution of house prices (Chart 2).
As shown in Chart 3, the least restrictive period in terms of criteria for making housing loans coincided
with a decline in the materialisation of credit risk. Following the onset of the current financial crisis, the
more restrictive lending criteria, as well as the decline in average LTV ratios, coincided with an increase
in the annual flow of new loans in default.
In December 2012, around half of the portfolio of loans to households for house purchase of the largest
banking groups had an LTV ratio of less than 70 per cent (Chart 4) and the average LTV is around 65
per cent. Although these values may be considered relatively low, it should, nevertheless be pointed
out that 35 per cent of the portfolio had an LTV of more than 80 per cent (16 per cent of the portfolio
with an LTV of more than 90 per cent). Although no instances of across-the-board overvaluation in the
Portuguese housing market have been recorded over the last few years, the available indicators point
to a situation of falling house prices since mid 2011, a period which coincides with the inception of
the economic and financial assistance programme. Against the background of a deep and protracted
recession affecting the Portuguese economy, there is the risk of an additional reduction of house prices,
and the significant proportion of loans with relatively high LTVs represents a challenge for banks’ risk
management and for the recovery of housing loans in default, even though delinquency in this portfolio
remains at relatively contained levels.
4
63
Credit Risk
The functioning of the housing market is important from a macroeconomic viewpoint, as housing repre-
Chart 2
LOAN TO VALUE RATIO AND SPREAD ON NEW
LOANS TO HOUSEHOLDS FOR HOUSE PURCHASE
DEMAND AND SUPPLY OF LOANS TO
HOUSEHOLDS FOR HOUSE PURCHASE
100
LTV at origination
74
Spread on new operations (r.h.s.)
3.5
80
3.0
60
73
Restrictiveness of supply
Demand
72
2.0
69
1.5
68
67
1.0
Diffusion index (%)
70
In percentage points
In percentage
40
2.5
71
20
0
-20
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
-40
66
-60
0.5
65
-80
64
0.0
2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
-100
Source: Banco de Portugal.
Source: Banco de Portugal (Bank Lending Survey).
Note: Data for the 8 largest banking groups. The period before
2008 includes estimated data.
Notes: Diffusion index calculated using a scale to aggregate individual replies, based on the direction and magnitude of each
bank’s reply. The index ranges from -100 to 100, and equals
zero when the situation is reported as unchanged. Positive values of the index correspond to a tightening of the supply of
loans or to an increase in demand.
Chart 3
Chart 4
OVERDUE AND OTHER DOUBTFUL BANK LOANS
TO HOUSEHOLDS FOR HOUSE PURCHASE
DISTRIBUTION OF THE LOAN TO VALUE RATIO
IN THE PORTFOLIO OF LOANS TO HOUSEHOLDS
FOR HOUSE PURCHASE IN DECEMBER 2012
0.5
25
Annual flow of new overdue and other doubtful loans
0.4
20
0.3
Per cent
15
Per cent
0.2
10
0.1
5
-0.1
Source: Banco de Portugal.
Jan-13
Jan-12
Jan-11
Jan-10
Jan-09
Jan-08
Jan-07
Jan-06
Jan-05
Jan-04
0.0
Jan-03
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
64
Q1
Q2
Q3
Q4
Q1
Q2
Q3
Q4
Q1
Q2
Q3
Q4
Q1
Q2
Q3
Q4
Q1
Q2
Q4
Q4
Q1
Q2
Q4
Q4
Q1
Q2
Q4
Q4
Q1
Q2
Q4
Q4
Q1
Q2
Q4
Q4
Q1
Q2
Q4
Q4
I
Chart 1
0
<50%
50%-60% 60%-70% 70%-80% 80%-90% 90%-100%
Source: Banco de Portugal.
Note: Data for the 8 largest banking groups.
>100%
5
The Portuguese banking system’s liquidity situation, at the end of 2012, was more favourable than at
65
the end of the preceding year. This is visible in the improvement of summary liquidity indicators such as
the liquidity gaps or the increase in the collateral buffers eligible for Eurosystem refinancing operations.
In parallel, through 2012 there was a progressive reduction in banks’ risk premia, whose decline kept
pace with the favourable developments noted on a level of Portuguese sovereign debt. At the end of
2012 and start of 2013, this evolution permitted some of the biggest Portuguese banking groups to
resume their issues of medium term senior debt securities. Notwithstanding, the banks’ access to the
international financial markets cannot be described as normal, taking into account both the reduced
volumes issued and the maintenance of a high yield spread in comparison to euro area benchmarks,
which is consistent with the situation relating to Portuguese sovereign debt.
The situation was improved by a combination of several self-reinforcing factors. Firstly, the measures
announced by the ECB and the European Union, which mitigated fragmentation in the financial markets
deriving from the sovereign debt crisis in the euro area. Secondly, the more favourable evolution of the
perception of sovereign risk, also resulting from the Portuguese authorities’ fiscal consolidation efforts,
diminishing the risk premia on domestic issuers. Finally, the continuation of the structural adjustment on
the banks’ balance sheets, particularly evident in the reduction of the loans-to-deposits ratio, compatible
with the gradual convergence to a more sustainable financing structure, less sensitive to changes in
international investors’ perceptions of risk.
A marked development during the course of 2012 was the substantial decline in banks’ short term refinancing needs, translating into an improvement of their liquidity gaps which, in the case of domestic
institutions, became positive in all maturities of up to 1 year. This evolution was fundamentally associated
with the ECB’s 3-year maturity monetary policy operations and the widening of the definition of assets
eligible as collateral with the Eurosystem, and signals an increase in the banking system’s resistance
to potential negative shocks on its financing capacity. The reinforcement of assets available for use as
collateral for liquidity injection operations is particularly important in a context of the persistence of risks
on the sustainability of the diminishing tensions in the international financial markets.
It should be noted that this evolution is consentaneous with the adoption of more demanding rules
under the scope of the future EU regulation on liquidity requirements.1 Notwithstanding such requirements, but given the context of the persistence of considerable uncertainty on an external and internal
level, it will be necessary for Portuguese banks to protect themselves from eventual shocks which may
compromise their capacity to ensure the financing of their assets and comply with their obligations as
and when they fall due.
1 The proposals for the new regulatory framework were set out in “Box 2.1 The main Basel III proposals”, Banco
de Portugal, Financial Stability Report-November 2010. At the beginning of 2013, the Basel Committee made
several changes to the Liquidity Coverage Ratio (LCR). Changes were made to the concepts of high quality liquid
assets and inflow and outflow rates, endeavouring to incorporate the recent evidence provided by the crisis
in the financial markets. Similarly, the prudential treatment of financing obtained from the central bank was
changed for the purpose of improving its stability, given the widening of the collection of assets afforded more
favourable treatment in terms of eligibility as stable collateral. A longer period for the ratio to be fully adopted,
a part of which beginning in 2015 and to be gradually implemented by 2019 was also defined. This gradual
approach was to ensure that the introduction of the LCR would not disturb the strengthening of banking systems and financing flows to the economy. For further details see http://www.bis.org/publ/bcbs238.htm and
associated documentation.
Liquidity Risk
5. Liquidity Risk
The Portuguese banking system’s funding conditions improved in 2013, but access to
international financial markets continued to be conditioned, notwithstanding the appearance
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
66
of several positive signs at the end of 2012 and start of 2013
The measures announced by the ECB and European authorities were interpreted as conducive to the
development of resolution mechanisms on the sovereign debt crisis in the euro area, helping to re-establish
investor confidence and reduce redomination risk. Such measures accordingly favoured a reduction of
systemic risk in the euro area and a gradual improvement of financing conditions. In this context, the
upwards path on the risk premia on Portuguese public debt was reversed in the first quarter of 2012,
with positive consequences on banks’ risk premia. There was a decline in the yield spread in the secondary market for the debt securities issued by resident banks vis-à-vis the Iboxx index2 (Chart 5.1). This
evolution enabled two of the main domestic banking groups to issue senior debt securities with medium
term maturities at the end of 2012. The issues, which proceeded through early 2013, recorded highly
significant demand-to-supply ratios and attracted a diversified range of investors, mostly international,
representing an initial step to the return to markets, albeit at a relatively high cost.
The reduction of banks’ financing costs was also noted on a deposits level. November 2011 saw the
start of a downwards path on interest rates on term deposits balances. The reduction for 2012 as a
whole was around 60 bp, plus 13 bp in first quarter 2013. This reduction was particularly marked in the
case of the term deposits of non-financial corporations, as opposed to the greater stability of the rates
associated with household deposits, which is also likely to be related to differences in the operations’
agreed maturities. During the course of the last few years, term deposits for more than a year represented
between 20 and 30 per cent of total new household operations but were of marginal importance in
the case of non-financial corporations, so that there is a certain inertia in the decline of interest rates on
balances for the first of these segments.
Interest rates on new term deposit operations with an agreed maturity registered a marked reduction
over the course of 2012, both in the case of non-financial corporations and households (1.6 and 1.3
percentage points, respectively). These developments are likely to be a reflection of the decline of interest
Chart 5.1
INTEREST RATES
14
12
Yields on senior bonds issued by Portuguese banks
Yields on covered bonds isued by Portuguese banks
Iboxx Euro Covered 1-10 years yields
Term deposits of the non financial private sector (new operations)
Term deposits of the non financial private sector (outstanding ammounts)
6 month Euribor (monthly average)
ECB main refinancing operations
10
Per cent
I
8
6
4
2
0
Dec-08
Dec-09
Dec-10
Dec-11
Dec-12
Sources: Bloomberg, Thomson Reuters and Banco de Portugal.
Notes: The yields on senior and covered bonds isued by portuguese banks were computed as the weighted average of yields observed in the secondary market for bonds issued by the banking groups BCP, BPI, BST, CGD, ESFG and MG with a residual maturity
between 1 and 10 years. Last observation: March 2013.
2 This index is made up of securities guaranteed by investment grade mortgages issued in euros. The yields on
the securities issued by the banks in the secondary market constitute an indicator of investors’ risk perceptions
and do not necessarily represent banks’ effective financing costs in wholesale debt markets which will remain
restricted for several issuers.
rates in the money market, easing of pressure on bank liquidity deriving from the ECB’s non-conventional
monetary policy measures and Banco de Portugal’s prudential measures, designed to penalise excessive
should, however, be noted that interest rates on operations with non-financial corporations were also
conditioned by operations performed by branches located in the Madeira and Azores off-shores. The
volatility associated with these operations made a decisive contribution to the reduced level of these
rates at the end of 2012 and their latter increase during the course of first quarter 2013.
During the course of 2012, reference should be made to the reduction of financing obtained
by the Portuguese banking system from the international wholesale debt markets and the
slowdown in customer resources
In the context of the deleveraging in progress in the banking system, a reduction of the customer
resources to credit gap, and increase in capital have been witnessed. This has been accompanied by a
significant reduction of financing from the international wholesale debt markets (Chart 5.2). More visible
in the first half year were the declines in the debt securities liabilities and net resources of other credit
institutions, especially non-residents. It should, however, be noted that the reduction of financing by
securities reflected not only the adverse conditions in the primary debt markets, which limited new debt
issues, but also debt repurchasing operations in the secondary market. As stated, it was only from the
end of the year that several Portuguese banks realised issues in these markets.
There was a very slight increase in financing (net of investments) from central banks for the year as a
whole, although there was a reduction in the case of domestic banks. A first half year increase (similar
for domestic and non-domestic banks) was followed by a significant reduction of domestic banks’ use
of this source of financing in the second half year.
There was a significant increase in subordinated liabilities in the context of several domestic banks’ issues
of contingent capital instruments (CoCos), subscribed for by the Portuguese state. This was associated
with the capitalisation needs deriving from domestic (Core Tier 1 ratio of at least 10 per cent) and international regulation, in the context of the European Banking Authority (EBA) exercise (core capital ratio
of at least 9 per cent). In the same vein, there were also capital increases by some banks.
Customer resources continued to make a positive contribution to the financing of the banking system,
albeit less markedly so than in 2011. Reference should, however, be made to the fact that in 2012 the
increase in customer resources was associated with the evolution of resources obtained in the context of
banks’ international activities. On a level of activities with residents, customer resources were slightly down.
In contrast to the notable dynamism of international activity, there was a slowdown in
deposit-taking from domestic customers, which was, to a large extent, foreseeable
There was a negative rate of change in residents’ deposits, at the end of 2012 (-2.3 per cent), with a
significant slowdown over the course of the year.3 The main contribution to this slowdown derived from
the evolution of household deposits, which accounted for around two thirds of residents’ total deposits
and recorded a virtually null rate of change in 2012, following the significant growth in 2011 (5.8 per
cent) (Chart 5.3).
3 Reference should be made to the fact that in the analysis of the evolution of residents’ deposits, the deposits of non-monetary financial institutions with a maturity of more than 2 years are excluded, as they largely
correspond to the accounting recognition of non-derecognised securitisation operations. This adjustment is
particularly relevant in 2012, as reversals of securitisation operations by several banking groups translated into
a significant decline in investments of resident non-monetary financial institutions with a maturity of more than
2 years.
5
67
Liquidity Risk
competition in deposit-taking from customers, potentially harming the whole of the banking system. It
Chart 5.3
BANKING SYSTEM SIX-MONTH FINANCING
FLOWS ON A CONSOLIDATED BASIS
DEPOSITS BY THE NON-MONETARY SECTOR DOMESTIC ACTIVITY | ACCUMMULATED CHANGE AS
FROM JANUARY 2007 - SECTOR CONTRIBUTIONS
68
50
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
40
50 000
Resident private individuals and emigrants
Resident non-financial corporations
Other resident sectors (a)
Non-residents
40 000
30
EUR Billion
20
30 000
EUR million
I
Chart 5.2
10
0
-10
20 000
10 000
-20
0
-30
H2 H1 H2 H1 H2 H1 H2 H1 H2 H1 H2 H2 H1 H2 H1 H2 H1 H2 H1 H2 H1 H2
07 08
09
10
11
Banking system
12 07 08
09
10
11
12
Domestic banks
-10 000
Jan-07
Jan-08
Jan-09
Jan-10
Jan-11
Jan-12
Jan-13
Resources from central banks (net)
Subordinated liabilities
Customer resources and other loans
Capital
Resources from other credit institutions (net)
Debt securities
Total financing
Source: Banco de Portugal.
Source: Banco de Portugal.
Notes: (a) Includes deposits by non-monetary financial institutions with a maturity below 2 years and deposits by insurance
corporations, pension funds and the general government. The
dashed lines exclude the deposits in banks’ branches located in
Madeira and Azores off-shores. The vertical dashed line signals
the financial assistance request by Portugal. Last observation:
March 2013.
To a large extent, the deceleration of household deposits in 2012 was foreseeable, taking into account
the occurrence of portfolio adjustments in 2011. During this year, the materialisation of market risk,
increase in interest rates on deposits and continued confidence in the banking system, led to a significant recomposition of households’ financial assets portfolios.4 This took the form of the replacement
of investments in mutual funds, life insurance, savings certificates and other public debt securities by
bank deposits. The effect of this type of portfolio adjustment on the rate of change of deposits tends
to dissipate in line with the recomposition of assets.
The favourable evolution of financial markets over the course of 2012 also re-established a certain attractiveness of products whose interest is associated with the evolution of prices in these markets, helping
to moderate the recomposition of the portfolio to deposits. This appeal has also motivated several large
non-financial corporations’ issues of significant volumes of bonds over the medium term, at relatively
attractive yields, a substantial part of which was sold on by the banks to their retail customers. In parallel
and in addition to capital increases, several banks resumed their issues of own bonds for sale to their
customers, although the amounts of such issues were less significant and with shorter maturities on
the securities.5 It should be noted that this migration of deposits to other investments is also likely to
4 See “Section 4 Credit Risk”, of this Report.
5 See “Section 4 Credit Risk”, of this Report.
have reflected, inter alia, Banco de Portugal’s rules relating to interest payable on deposits.6 This context
witnessed a significant flow of investments in debt securities by households.7
evolution of the annualised quarterly rate of change (Chart 5.4). It should, however, be borne in mind that
the intra-annual evolution profile on deposits, this year, is likely to be affected by changes to payments
of holiday and Christmas subsidies, which will tend to change the usual seasonal patterns.
Deposits made by non-financial corporations also made a significant contribution to the slowdown
of bank deposits. However, the decline in this aggregate’s rate of change, which, at the end of 2012,
accounted for around 15 per cent of residents’ deposits, was determined by the evolution of deposits
made with bank branches in the Madeira and Azores off-shores. Excluding such deposits, the deposits
of this sector, in December 2012, were down by around 10 per cent over the preceding year, albeit
having stabilised since mid 2012. Given the constraints in the credit market, companies have had added
incentives to implement more demanding liquidity management. On the other hand, in the context of the
uncertainty associated with the sovereign debt crisis in the euro area, particularly in first half 2012, the
possibility that several specific customer segments, including large enterprises, may have geographically
diversified their deposits portfolio cannot be excluded.
Reference should also be made to the fact that, to a certain extent, the slowdown of deposits reflects
the more comfortable liquidity situation of most banks. This derived, on the one hand, from the non-conventional monetary policy measures adopted by the Eurosystem with the aim of normalising financing
conditions in the euro area and, on the other, the progress on a level of the structural adjustment of
the banks’ liquidity position.
The deposits of non-monetary financial institutions (excluding deposits for agreed periods of more than 2
years8) and insurance companies and pension funds, as in the case of general government deposits, tend
to be more volatile, with a positive contribution to the growth of deposits of the resident non-monetary
sector having been made at the end of 2012.
Continued decline in the loans-to-deposits ratio, with a larger contribution of the decline in
credit in 2012
As stated, the loans-to-deposits ratio continued to be adjusted in 2012 (Charts 5.5 and 5.6). The main
contribution to this evolution derived from the evolution of credit, owing to moderation in the expansion
of customer resources, in contrast to 2011. It should be noted that the decline in the credit portfolio
reflected not only the reduction of net credit flows but also loan disposals by several of the main banking
groups, notably loans to resident non-financial corporations.
At the end of 2012, the loans-to-deposits ratios of the biggest resident banking groups on a consolidated basis was around 120 per cent, down 8 pp over the end of 2011 (Chart 5.7). Three of the eight
institutions asked to submit funding and capital plans under the scope of the economic and financial
assistance programme, representing as a whole around one third of total banking system assets, had
6 This evolution appears in a context in which the deduction applied by Banco de Portugal on the own funds of
the banks offering higher interest rates on deposits translates into a significantly higher penalty for the shorter
periods. The spreads considered in the definition of key reference rates on the basis of which deductions to own
funds are made are higher in the case of operations with longer maturities with the deduction from own funds,
on the other hand, applying for a year, notwithstanding the period of the deposit.
7 More recently, in third quarter 2012, the rules governing interest on savings certificates were changed, with
the aim of increasing this product’s savings appeal, contributing to stabilising the financing obtained by general
government from this instrument.
8 These deposits are excluded from the analysis owing to the fact that, as already stated, they are associated with
the accounting of the financing associated with credit securitisation operations.
5
69
Liquidity Risk
Data for first quarter 2013 indicate a certain acceleration of household deposits, as suggested by the
Chart 5.5
DEPOSITS BY PRIVATE INDIVIDUALS | RATES OF
RATIO OF CREDIT TO CUSTOMER RESOURCES (a) |
CHANGE
BANKING SYSTEM
35
30
Year-on-year rate of change
Annualised quarterly rate of change
180
25
Gross credit (including securitised and non derecognised credits)
– customer resources ratio
Credit (including securitised and non derecognised credits) net of
impairments – customer resources ratio
Gross credit (including securitised and non derecognised credits)
– customer resources ratio (b)
170
20
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
70
15
10
Per cent
I
Chart 5.4
160
150
5
140
0
-5
-10
Dec-98Jun-00Dec-01Jun-03Dec-04Jun-06Dec-07Jun-09Dec-10Jun-12
130
120
Jun-05 Jun-06 Jun-07 Jun-08 Jun-09 Jun-10 Jun-11 Jun-12
Source: Banco de Portugal.
Source: Banco de Portugal.
Notes: The annualised quarterly rate of change is calculated on
seasonally adjusted data. Last observation: March 2013.
Notes: (a) Data on a consolidated basis. The concept of customer resources includes mostly deposits and does not account
for debt securities issued by the banks and placed with their
customer base. The break in the series in 2007 comprises an
increase in the number of institutions under analysis. (b) Information obtained under the report set by Banco de Portugal
Instruction no. 13/2009, which considers only the set of institutions which collect customer deposits.
ratios of less than 120 per cent, with only one institution having a ratio of more than 130 per cent (when,
at the end of 2011, there were 4 institutions in this situation).
The decline in the loans-to-deposits ratio was across-the-board to most banks, translating into a movement
of the ratio’s empirical distribution curve to the left (Chart 5.8). The convergence noted eliminated the
bimodal character of the distribution, indicating that the adjustment was tendentially more significant
for institutions with the highest ratios. Reference should, notwithstanding, be made to the importance
of proceeding gradually with the adjustment, in the case of institutions with the highest ratios. This is
particularly important in the context of the future prudential regulation on liquidity.
Reduction of Eurosystem financing to domestic banks but an increase in the use of this
source of financing by non-domestic banks
As already stated, there was a very slight increase in the Portuguese banking system’s use of financing
(net of investments) from central banks in 2012, although reference should be made to the difference
between domestic and non-domestic institutions and the intra-annual profile. The ECB’s second liquidity
injection operation with a maturity of 3 years and the widening of the collection of assets eligible as
collateral, were followed by the Portuguese banking system’s use of Eurosystem financing during the
course of the first half, both by domestic and non-domestic institutions (Chart 5.9a). This evolution was
in line with the rest of the euro area (Chart 5.9b). There was, however, a reduction of the domestic banks’
use of this source of financing in the second half year. In the case of non-domestic banks, the decline
occurring in the second half year was inexpressive, partly reflecting the evolution noted in their financing
strategy, towards a greater level of autonomy from their respective parent companies, translating into
a substantial increase of the use of Eurosystem liquidity operations and significant endeavours to adjust
their balance sheets, either in the form of deposit-taking or based on lower lending levels.
Chart 5.6
RATIO OF CREDIT TO CUSTOMER RESOURCES
(a)
| DOMESTIC BANKS
5
165
Gross credit (including securitised and non derecognised
credits) – customer resources ratio
Credit (including securitised and non derecognised credits) net
of impairments – customer resources ratio(b)
Gross credit (including securitised and non derecognised
credits) – customer resources ratio(b)
Gross credit – customer resources ratio (international activity)
Credit net of impairments – customer resources ratio (international activity)
145
Per cent
135
125
115
105
95
85
Jun-05 Jun-06 Jun-07 Jun-08 Jun-09 Jun-10 Jun-11 Jun-12
Source: Banco de Portugal.
Notes: (a) Data on a consolidated basis. The concept of customer resources includes mostly deposits and does not account for debt
securities issued by the banks and placed with their customer base. The break in the series in 2007 comprises an increase in the
number of institutions under analysis. (b) Information obtained under the report set by Banco de Portugal Instruction no. 13/2009,
which considers only the set of institutions which collect customer deposits
Chart 5.7
RATIO OF CREDIT TO CUSTOMER RESOURCES
FOR THE LARGEST RESIDENT BANKS
Chart 5.8
RATIO OF CREDITO TO CUSTOMER RESOURCES DOMESTIC BANKS | EMPIRICAL DISTRIBUTION
170
Dec-11
Jun-12
Mar-13
160
Per cent
150
140
130
120
110
100
Dec-07 Dec-09 Mar-10 Sep-10 Mar-11 Sep-11 Mar-12 Sep-12
25
75
125
175
225
Source: Banco de Portugal.
Source: Banco de Portugal.
Notes: Data on a consolidated basis.The concept of credit is
net of impairment and includes securitised and non derecognised credits. The concept of customer resources includes mostly
deposits, does not include debt securities issued by the banks
and placed with their customer base and comprises stable funding lines obtained from the parent company, qualified shareholders or multilateralinstitutions.
Notes: The concept of customer resources includes mostly deposits and does not account for debt securities issued by the
banks and placed with their customer base. Information obtained under the report set by Banco de Portugal Instruction no.
13/2009, on a consolidated basis. Empirical distribution obtained through recourse to non-parametric methods, namely to a
Gaussian kernel that weights institutions by their assets.
Liquidity Risk
71
155
Chart 5.9b
OUTSTANDING AMOUNTS OF MONETARY
POLICY OPERATIONS OF RESIDENT BANKS
OUTSTANDING AMOUNTS OF MONETARY
POLICY OPERATIONS OF THE EUROSYSTEM
1500
60
1000
50
40
EUR Billion
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
72
500
EUR Billion
I
Chart 5.9a
30
20
10
0
-500
0
-1000
-10
-20
Jan-08
Jan-09
Jan-10
Jan-11
Jan-12
Jan-13
-1500
Jan-08
Jan-09
Jan-10
Jan-11
Jan-12
Jan-13
Main refinancing operations
Longer-term refinancing operations
Marginal lending facility
Other liquidity provision operations(a)
Deposit facility
Other liquidity absortion operations(b)
Source: Banco de Portugal.
Notes: (a) Includes “Fine-tuning operations” and “Structural operations”. (b) Includes “Fixed-term deposits” and “Reverse transactions”. Last observation: March 2013.
In the case of domestic institutions, a progressive reduction of resources obtained from other non-resident
credit institutions continued to be noted and should also be viewed in the context of fragmentation
in financing markets in the euro area and the maintenance of a particularly reduced level of activity in
non-collateralised interbanking markets on an international level. The situation continues to translate
into a significant use of the deposits facility by euro area banks as a whole.
In this context, it should be remembered that owing to the disturbances in the euro area money market
and to stimulate the efficient working of the domestic interbank money market, from 3 September 2012,
Banco de Portugal provided resident institutions with a platform for the registration and processing of
non-collateralised operations in the interbank money market by means of which institutions can swap
euro denominated funds among each other for periods of up to a year. A similar platform for the domestic
money market with a guarantee on assets was also provided at the start of May 2013. The aim was to
improve the effectiveness of the monetary policy transmission mechanism.
The financing obtained by the Portuguese banking system from the Eurosystem at the end of 2012
represented around 5 per cent of the total use of Eurosystem monetary policy operations and just over
10 per cent of resident banks’ balance sheets in Portugal (with domestic and non-domestic banks posting
similar levels). The amount has remained relatively stable since the settlement of the second liquidity
injection operation with a maturity of 3 years, in March 2012.
Across-the-board improvement in liquidity gaps following the 3 year liquidity injection
operation9
There was a substantial improvement in the Portuguese banking system’s liquidity gaps over the course of
5
In the case of domestic institutions, the main factors underpinning this evolution were the lengthening of the
73
average maturity of financing obtained from the Eurosystem, deriving from the ECB’s second 3 year liquidity
Liquidity Risk
2012, transversal to the various time horizons considered and most institutions (Charts 5.10, 5.11 and 5.12).
injection operation, and decline in debt securities (Chart 5.13). In the case of the non-domestic banks, the
improved gaps fundamentally reflected the improvement of the net interbanking position, in the context of
the already referred to change in these banks’ financing strategy and an increase in claims and short term
investments in credit institutions abroad.
Reinforcement of asset portfolios eligible as collateral for Eurosystem liquidity injection
operations, following the resolutions of the Governing Council of the ECB
The improvement of the Portuguese bank’s liquidity situation also reflected the increase in the size of the
collateral pool and, in general, the assets eligible as collateral for the ECB’s monetary policy operations, which
also consider assets that, although eligible for inclusion in the said pool, have not, as yet, been included. This
evolution was made possible by the non-conventional monetary policy measures approved by the Governing
Council of the ECB in December 2011. They particularly include, inter alia, on account of their importance,
those relating to the collateral eligibility rules including, on the one hand, the lowering of the minimum rating
for the eligibility of asset-backed securities (ABS) and, on the other, the possibility of domestic central banks’
temporary acceptance of bank loans fulfilling specific eligibility criteria and with rules subject to approval
by the Governing Council10, as a guarantee. In particular, this last measure made it possible to significantly
increase the banks’ capacity to generate collateral, diminishing their sensitivity to international investors’ risk
perceptions and rating fluctuations, reducing uncertainty over collateral values.
Chart 5.10
LIQUIDITY GAPS IN CUMULATIVE MATURITY LADDERS
Per cent of total assets minus liquid assets
12
9
Up to 1 month
Up to 3 months
Up to 1 year
6
3
0
-3
-6
-9
-12
-15
-18
-21
Dec-08
Dec-09
Dec-10
Dec-11
Dec-12
Source: Banco de Portugal.
Notes: The liquidity gap is defined as (Liquid Assets – Volatile Liabilities)/(Assets – Liquid Assets) x 100 for each cumulative ladder
of residual maturity. Information obtained under the report set by Instruction no. 13/2009 of Banco de Portugal, on a consolidated
basis. The dashed lines show domestic institutions.
9 Liquidity gaps are defined by the ratio (net assets – volatile liabilities)/(assets – net assets )*100, in each cumulative maturity ladder.
10 For further details on the changes to collateral eligibility for liquidity injection operations, see Section 4.3 “Liquidity Risk”, Banco de Portugal, Financial Stability Report-May 2012.
I
Chart 5.11
Chart 5.12
LIQUIDITY GAP UP TO 1 MONTH - DOMESTIC
INSTITUTIONS | EMPIRICAL DISTRIBUTION
LIQUIDITY GAP UP TO 12 MONTHS - DOMESTIC
INSTITUTIONS | EMPIRICAL DISTRIBUTION
-40
-30
-20
-10
0
10
20
30
Dec-11
Dec-11
Jun-12
Jun-12
Mar-13
Mar-13
40
-40
-30
-20
-10
0
10
20
30
40
Source: Banco de Portugal.
Source: Banco de Portugal.
Notes: Information obtained under the report set by Banco
de Portugal Instruction no. 13/2009 , on a consolidated basis. Empirical distribution obtained through recourse to non-parametric methods, namely to a Gaussian kernel that weights
institutions by their assets.
Notes: Information obtained under the report set by Banco
de Portugal Instruction no. 13/2009, on a consolidated basis.
Empirical distribution obtained through recourse to non-parametric methods, namely to a Gaussian kernel that weights
institutions by their assets.
In this context, 2012 saw an increase in the size of the Portuguese banking system’s collateral pool for
Eurosystem monetary policy operations, enabling greater use of refinancing operations in addition to maintaining a higher level of overcollateralisation than noted at the end of 2011 (Chart 5.14). A contributory
factor was the inclusion of significant volumes of loans and advances to customers in the collateral pool.
Significant volumes of public debt securities, credit institutions’ debt securities and covered bonds were also
included in the collateral pool. This was accompanied by a decline in the value of asset-backed securities (ABS)
in the collection of assets included in the collateral pool, reflecting the reversal of securitisation operations
Chart 5.13
LIQUIDITY GAPS FOR DOMESTIC BANKS - MATURITIES UP TO 12 MONTHS | MAIN CONTRIBUTIONS
30
20
10
0
-10
-20
-30
-40
Up to 1 month
Up to 1 year
Up to 1 month
Up to 1 year
Up to 1 month
Up to 1 year
Up to 1 month
Up to 1 year
Up to 1 month
Up to 1 year
Up to 1 month
Up to 1 year
Up to 1 month
Up to 1 year
Up to 1 month
Up to 1 year
Up to 1 month
Up to 1 year
Up to 1 month
Up to 1 year
Up to 1 month
Up to 1 year
Per cent of total assets minus liquid assets
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
74
Securities eligible as collateral in credit operations with central
banks (unencumbered)
Derivatives
Commitments to third parties
Other assets/liabilities
(Net) resources from central banks
(Net) resources from other credit institutions
Liabilities in the form of securities
Liquidity gap - up to 1 month
Liquidity gap - up to 3 months
Liquidity gap - up to 1 year
Dec- Dec- Mar- Jun- Sep- Dec- Mar- Jun- Sep- Dec- Mar09
10
11
11
11
11
12
12
12
12
13
Source: Banco de Portugal.
Note: Information obtained under the report set by Banco de Portugal Instruction no. 13/2009, on a consolidated basis.
Chart 5.14
EUROSYSTEM FINANCING AND COLLATERAL POOL OF RESIDENT BANKS
60
10
0
0
Source: Banco de Portugal.
Notes: (a) Outstanding amounts on main refi nancing operations, on longer-term r efinancingoperations and on occasional regularization operations. From 4 July 2011 it also includes intraday limit credit operations. From that date Banco de Portugal only has a
single collateral pool for monetary policy and intraday credit operations.
by several banking groups following the changes in the eligibility criteria on the use of assets as collateral
for monetary policy operations.
Lastly, it should be noted that the banks still have eligible assets which are not included in the pool and
reference should also be made to their capacity to generate additional collateral based on loans and advances
to customers.11 This gives them an additional short term capacity to accommodate adverse liquidity shocks
such as those deriving from the adverse evolution of ratings on assets whose eligibility depends on them.
It should, however, be noted that the measures adopted by the Eurosystem are likely to be reversed over
the medium to long term and it is therefore extremely important that Portuguese banks should continue to
implement the gradual structural adjustment process on their balance sheets, adjusting the liquidity of assets
to the requirements deriving from their liabilities.
11 For further details on the additional collateral generating capacity and respective risk control measures, see
“Section 4.3 Liquidity Risk”,Banco de Portugal, Financial Stability Report-May 2012.
Liquidity Risk
15
Abr-13
20
Dez-12
30
Jun-12
30
Dez-11
45
Jun-11
40
Dez-10
60
Jun-10
50
Percentage of collateral pool
75
75
Jan-10
EUR Billion
5
Credit operations (a)
Collateral pool
Overcollateralization (r.h.s.)
90
6. Market Risk
6
The main market risks to the Portuguese banking system derive from the interaction
between weak economic growth prospects on a European level and tensions in the sovereign
debt markets, in a context of the banks’ higher exposure to sovereign risk
The securities and financial investments portfolio exposes the banks to losses on the value of their securities. After a favourable evolution during the course of 2012, an eventual heightening of tensions in
the international financial markets and particularly in the sovereign debt markets and their interaction
with the real economy comprise the main sources of market risk to the Portuguese banking system. In
particular, losses associated with the depreciation of securities may translate into significant pressure on
banks’ profitability and equity.
It is therefore crucial that the commitments made on a euro area level over the course of the last months
be reinforced. Their respective implementation should ensure greater financial and fiscal integration,
making it possible to create the mechanisms required to curb the interaction effects between sovereign
risk and financial stability.
The increase in the available for sale assets portfolio, in 2012, resulted from the net
acquisition of public debt securities in the first half of the year and appreciation of the
respective portfolio during the course of 2012
There was an increase in the value of the Portuguese banking system’s securities and financial investments
portfolio, in 2012.1 This evolution occurred in first half 2012, especially in the first quarter of the year
and particularly reflecting net acquisitions, but also portfolio appreciation. This was offset by a decline
in portfolio value in second half 2012, which was associated with the capital gains on sales made at the
end of the year. The increase noted, in 2012, was particularly significant, in a context in which the assets
total was significantly down. It should be remembered that the Portuguese banking system’s securities
and financial investments portfolio lost value in 2011 – reflecting unfavourable developments in the
international financial markets having a negative impact on the value of the securities held – notwithstanding the full year to have observed an increase in the Portuguese public debt and other resident
public issuers’ securities portfolio.
The increase in the size of the portfolio essentially translated the evolution of available for sale financial
assets, which largely reflected acquisitions of Portuguese public debt securities and appreciation of the
respective portfolio during the course of 2012 (Chart 6.1).2 This evolution benefited from the ECB’s
non-conventional monetary policy measures.3 Reference should also be made to the fact that under
the scope of the capital increases associated with the requirements defined by the European Banking
Authority (EBA) for June 2012, several of the major Portuguese banking groups increased the level of
their investments in public debt securities. This was offset by a reduction in the assets held to maturity
1 The securities and financial investments portfolio comprises financial assets at fair value through profit or loss,
including trading derivatives (net of liabilities held for trading), available for sale financial assets, investments
held to maturity, investments in subsidiaries and the net value of hedge derivatives registered in the Portuguese
banking system’s balance sheet, on a consolidated basis.
2 See “Box 6.1 Evolution of residents’ Portuguese public debt portfolios”, of this Report.
3 For further details see “Box 1.2 Non-standard monetary policy in the major advanced economies” Banco de
Portugal, Economic Bulletin - Autumn 2012.
Market Risk
77
portfolio, which translated a decline in Portuguese and Greek public debt securities and the redemption
effect on the bonds of national private issuers relative to one of the biggest Portuguese banking groups
There was a decline in the value of the Portuguese banking system’s securities and financial investments
portfolio in second half 2012. This decline translated, on the one hand, the sale of bearer insurance
certificates classified in the other financial assets at fair value through profit or loss portfolio by one of
the biggest Portuguese banking groups and, on the other, the reduction of investment in subsidiaries,
associated with sales of preference shares held by one of the main non-domestic banking groups.
When analysed in terms of source of risk, the increase in the size of the portfolio was associated with
the maturity of financial liabilities held for trading – recognised as other financial instruments in financial
assets at fair value through profit or loss – by one of the biggest Portuguese banking groups. Another
contributory factor in the change in the portfolio was the increase in other financial instruments, notably
investment units in venture capital funds, to which a contribution was made by two of the biggest Portuguese banking groups’ loan portfolio sales. One of these groups also increased its exposure in investment
units in real state funds. Similarly, the equity securities portfolio which, on the same date accounted for a
proportion of 1 per cent of the assets, registered a slight increase. By contrast, interest rate instruments,
which are the main source of risk component of the portfolio of securities and financial investments
Chart 6.1
Chart 6.2
OTHER MONETARY FINANCIAL INSTITUTIONS
PORTFOLIO | IN PORTUGUESE PUBLIC DEBT SECURITIES
SECURITIES AND FINANCIAL INVESTMENTS
PORTFOLIO | IN ACCORDANCE WITH THE IAS
CLASSIFICATION
10 000
Transactions
Price changes
18
16
8 000
14
6 000
Percentage of assets
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
78
in first half 2012 (Chart 6.2).4
EUR million
I
4 000
2 000
12
10
8
6
4
0
2
-2 000
0
-2
-4 000
2010 H1
2010 H2
2011 H1
2011 H2
2012 H1
2012 H2
Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec- Jun- Dec07 07 08 08 09 09 10 10 11 11 12 12
Financial assets at fair value through profit or loss (net)
Available for sale financial assets
Investments held to maturity
Investments in subsidiaries
Trading derivatives
Hedge derivatives
Total securities and investments portfolio
Source: Banco the Portugal.
Source: Banco de Portugal.
Note: The information in this chart should be seen as indicative
of the relative importance of data on transactions and price
changes, since the source of information used, i.e., Sistema
Integrado de Estatísticas de Títulos (SIET) is distinct from that
considered in this section.
Note: The securities and financial investments portfolio comprises financial assets at fair value through profit or loss, including
trading derivatives (net of liabilities held for trading), available
for sale assets, investments held to maturity, investments in
subsidiaries and the net value of hedge derivatives registered
in the Portuguese banking system’s balance sheet on a consolidated basis.
4 Reference should be made to the agreement reached pursuant to the scope of the plan for the involvement of
the private sector in financial support to Greece at the end of 2011. This initiative implied a haircut of 53.5 per
cent of the value of securities, in addition to the conversion of 15 per cent into debt securities of the European
Financial Stabilisation Fund and the remaining 31.5 per cent in new Greek sovereign securities with maturities
of between 11 and 30 years.
portfolio – representing around 12.6 per cent of total banking system assets on a consolidated basis –,
fell slightly at the end of 2012. This decline reflected a reduction in the value of debt securities, public
Around 65 per cent of the debt securities portfolio is made up of sovereign debt securities, in particular
medium and long term, translating into one of the transmission channels between the banking system
and sovereign risk (Chart 6.3). At the end of 2012, the proportion of public debt securities registered in
each of the different assets portfolios was 82 per cent of available for sale assets, 11 per cent of assets
held to maturity and 7 per cent of assets at fair value. As already stated, the increase in public debt
securities was concentrated in the available for sale assets portfolio in which assets are assessed on a
mark-to-market basis and value fluctuations accounted in revaluation reserves.5
In comparison with other banks in the euro area, particularly in other countries subject to strong pressures
in the sovereign debt markets, Portuguese banks continued to occupy an intermediate position in terms
of their exposure to public debt securities over the course of 2012 (Chart 6.4).6 The public debt securities
portfolio is, to a large extent, heavily populated by Portuguese securities. As stated, the increase in the
size of the Portuguese public debt portfolio reflected not only acquisitions of securities – occurring in the
first half of the year and only partly reversed in second half of 2012 – but also the increase in the value
market of securities during the course of 2012. As regards the public debt securities of other countries,
2012 saw an increase in the proportion of Spanish and Italian securities and a decline in Greek securities.
Chart 6.3
Chart 6.4
BREAKDOWN OF DEBT SECURITIES PORTFOLIO
GOVERNMENT BONDS HELD BY MONETARY
FINANCIAL INSTITUTIONS IN SELECTED EURO
AREA COUNTRIES
Portuguese public debt and other public resident issuers
Issued by foreign public entities and international organisations
Other debt securities
10
9
80
Ireland
Spain
8
Percentage of assets
70
EUR billion
Portugal
Greece
Italy
60
50
40
30
20
7
6
5
4
3
2
10
1
0
Dec-07
Dec-08
Dec-09
Dec-10
Dec-11
Dec-12
0
Jan-07
Jan-08
Jan-09
Jan-10
Jan-11
Source: Banco de Portugal
Source: ECB.
Note: Debt securities portfolio in the balance sheet of the
banking system, on a consolidated basis.
Note: Last observation – February 2013.
Jan-12
Jan-13
5 In accounting terms, changes in the financial assets assessed at fair value portfolio through profit or loss are
fully reflected in income accounts, whereas changes in other components of the securities and financial investments portfolio only affect income for the year when related with the sale of instruments or when they are
underpinned by value changes which imply the recognition of impairment. Value changes which do not require
such recognition are processed in the revaluations reserves component in shareholders’ equity. In addition, value
changes of available for sale financial assets, also valued at mark-to-market, have an impact in prudential terms,
i.e. on institutions’ regulatory capital, albeit differentiated in accordance with the type of instrument. In particular, whereas potential capital gains and losses on equity securities are considered for the own funds assessment,
the effect of changes in the value of debt securities is neutral.
6 It should be remembered that the significant decline in public debt securities held by Greek banks in March 2012
was associated with the private sector’s participation in the restructuring of Greek public debt.
6
79
Market Risk
and private, issued by non-resident entities classified in the investments held to maturity portfolio.
Improved results associated with financial operations, to a large extent reflecting own bond
repurchase operations by the major resident banking groups in addition to gains on the
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
80
appreciation of debt securities portfolios
Income from financial operations, net of impairment in securities and financial investments, were up
in 2012 over the preceding year, making a positive contribution to returns on assets (Chart 6.5).7 This
increase largely reflected the result of own bonds repurchase operations by the major Portuguese banking
groups, also observed over the course of 2011, which benefited from the fact that they were undervalued
in comparison to their nominal value. Reference should also be made to the contribution made by the
reduction of yields on medium and long term debt securities issued by Portuguese entities – particularly
Portuguese public debt securities – to income from financial operations in which the banks exploited
the price increases of these securities to make capital gains at the end of the year.
In this context, the appreciation of the financial assets portfolio made a positive contribution to institutions’
equity, reflecting the decline in the negative value of reserves measured at the fair value of debt securities.
Positive evolution of pension funds, reflecting portfolio appreciation and higher
contributions in 2012
There was an improvement in the global level of coverage of bank employees’ pension funds in 2012,
which contributed to a large extent the observed increase in assets (7.7 per cent). The evolution of
assets reflected the appreciation of the fund’s portfolio and the improved net balance between pensions
contributions received and paid (Table 6.1). It should be remembered that 2011 witnessed a significant
decline in the portfolio values of pension funds and their respective liabilities, following the transfer to
Social Security of liabilities for pensions in payment.8
In aggregate terms, the impact in equity of the financial and actuarial deviations, through reserves or via
results, is reduced, as the increase in liabilities arising from the revision of the calculation assumptions
Chart 6.5
INCOME FROM FINANCIAL OPERATIONS AND IMPAIRMENTS ON SECURITIES AND FINANCIAL
INVESTMENTS | CONTRIBUTIONS TO RETURN ON ASSETS - IN ACCORDANCE WITH THE IAS CLASSIFICATION AND THE SOURCE OF
RISK
0.40
10 H1
10 H2
11 H1
11 H2
12 H1
12 H2
0.30
Percentage of average assets
I
0.20
0.10
0.00
-0.10
-0.20
-0.30
-0.40
-0.50
-0.60
InterestShares Other InterestShares Other
rate
risks rate
risks
Financial
assets
at fair value
through
profit or loss
(net)
Available
for sale
financial
assets
Hedge Subsid- Ex- Other Impair- Total
and iaries change
ments
trading
differdeences
rivatives
Source: Banco de Portugal.
7 It should also be remembered that, in 2011, the value of impairment associated with the securities and financial
investments portfolio increased very highly in the context of the private sector’s participation in the restructuring
of Greek public debt.
8 For further details see “Box 4.2 Accounting and prudential impact of the partial transfer of banking sector pension funds to the Social Security System”, Banco de Portugal, Financial Stability Report – May 2012.
Table 6.1
PENSION FUNDS - BANKING SYSTEM | ON AN INDIVIDUAL BASIS; EUR MILLION
2010
2011
2012
6
81
Total liabilities
13 991
14 018
7 510
7 824
Minimum level of liabilities to be covered
13 410
13 506
7 190
7 599
13 268
14 388
14 037
7 519
1 190
-209
-652
429
464
481
790
508
Pension fund
Value of pension fund at the beginning of the year
Net income of fund
Contribution made to fund
Contribution paid by beneficiaries
53
53
52
48
Retirement pensions paid by the fund
633
648
655
172
Survivors' pensions paid by the fund
36
32
36
14
0
0
-4 490
-235
Changes in fund value resulting from cuts or sales
Other net changes
Value of pension fund at the end of the year
Coverage of fund: Value of pension fund at the end of the year (including
other forms of coverage) - Minimum liabilities level to be covered
60
10
-1 354
12
14 365
14 043
7 693
8 096
1 336
911
858
802
Source: Banco de Portugal.
Notes: The value of the Pension Fund (PF) at the beginning of a year does not necessarily coincide with the value of PF at the end of
last year, since the institutions that enter into the calculation of the PF in each year may not exactly match the ones that entered in
the previous year. This was particularly evident when comparing the value of the PF in early 2012 and late 2011. This result primarily
from the transfer of BPN PF to CGD in April 2012. Decree-Law No. 88/2012, of April 11 defines the conditions and group entities
covered by the transfer. These entities are not considered in calculating the value of pension funds in 2012. Thus, the comparison of
responsibilities between two years for PF should not be performed.
was largely mitigated by the reduction in liabilities from the change/repayment of benefits plans 9 and
financial positive deviations.
The end of 2012, due to the reduction of market interest rates for the relevant maturities, there was a
downward revision of the average discount rates used to value the funds liabilities.10
9 Change in the formula for calculation the death benefit (Decree - Law No. 133/2012) and settlement of a complementary retirement plan.
10 It should be noted that the average change in the discount rate, weighted by pension fund liabilities, decreased
by 73 basis points over the end of the preceding year. This information corresponds to about 72 percent of the
value of the liabilities of the pension funds of the banking system.
Market Risk
2009
Liabilities
BOX 6.1 | EVOLUTION OF RESIDENTS’ PORTUGUESE PUBLIC DEBT
PORTFOLIOS1
last few years (Chart 1). Until the emergence of the international financial crisis, in the summer of 2007,
the Portuguese government, like the other resident sectors, enjoyed regular access to financing from
the international wholesale debt markets. In fact, immediately preceding the sovereign debt crisis, non-residents held around three quarters of total Portuguese public debt, almost all of which comprised
debt securities.
However, the crisis caused a paradigm shift in the existence of liquidity in abundance at reduced cost,
which initially affected bank financing and latterly in line with the developments in the sovereign debt
crisis in the euro area, the public sector. Difficulties in access to the international financial markets by
the public sector were initially accommodated by the Portuguese financial system, particularly by banks.
However, the deterioration of the situation justified a request for external assistance in the second
quarter 2011. External financing was accordingly provided on the basis of loans made to the Portuguese
Republic under the scope of the economic and financial assistance programme, in which the proportion
of financing obtained from the international markets diminished to around one third of total general
government debt.
Starting 2010, the financing of general government was essentially provided by increased financing
from the resident banking system (Chart 2). In terms of instruments, financing has essentially taken the
form of the acquisition of debt securities, whose portfolio, held by the banks increased by around 250
per cent between the end of 2009 and end of first quarter 2012 (when it was around EUR 36 billion),
having latterly registered a reduction trend which was particularly significant in the last quarter of the
year. There was also an upwards trend in lending to mid 2011 (to around EUR 16 billion), which latterly
Chart 1
PORTUGUESE PUBLIC DEBT | BY INSTITUTIONAL SECTOR
210 000
180 000
Non-residents
European Investment Bank (EIB)
Economic and financial assistance programme
Other non-residents
EUR million
150 000
120 000
Residents
Other resident sectors
Other financial institutions
Banks
90 000
60 000
30 000
0
Q4 Q4 Q4 Q4
2007 2008 2009 2010
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2011 2011 2011 2011 2012 2012 2012 2012
Source: Banco de Portugal.
Notes: Data considering nominal value and excluding accrued interest. Other financial institutions include the insurance corporations
and pension funds, financial auxiliaries and other financial intermediaries. The other resident sectors correspond to the non-financial
private sector, i.e., households and non-financial corporations.
1 Concept of Maastricht debt, which includes all entities classified, for statistical purposes, in the general government institutional sector. A consolidated approach has been adopted, i.e., excluding general government
assets in liabilities issued by general government itself and in which the accumulated capitalisation of savings
certificates is not considered. Data consider nominal value and excluding accrued interest.
4
83
Market Risk
The accumulation of debt by Portuguese general government has been a marked characteristic of the
declined up to the end of 2012 (to around EUR 8 billion). At the end of 2012, the banking system’s
exposure to Portuguese general government represented approximately 9 per cent of assets, having
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
84
increased by around 5 percentage points over the end of 2009.
There was a significant increase in financing from other resident financial institutions over the course
of 2010 and first quarter 2011 (of around EUR 6.7 billion), which remained relatively stable up to first
quarter 2012. The evolution of this aggregate was latterly affected by the impact of the processing of
the privatisation operations of two major enterprises in the energy sector (involving Parpública). The
last quarter of 2012 witnessed an increase in this sector’s portfolio, partly offsetting the referred reduction in the banking sector. Accordingly, the proportion of the exposure of other financial institutions
to Portuguese general government has increased significantly since the end of 2009. This increase was
accompanied by disinvestments of insurance corporations and pension funds and, to a lesser extent, of
investment funds in medium and long term debt securities issued by non-residents. An increase in the
interaction between sovereign and banking risk and the financial sector in general was therefore noted
in relation to 2009.
In turn, funding from other resident sectors registered a slight downward trend, strongly conditioned by
the reduction in savings and Treasury certificates held by households. At the end of 2010, the aggregate
was affected by the accounting of the impact of the renegotiation of several public-private partnership
agreements and debt settlement agreements for the Autonomous Region of Madeira.
Chart 2
PORTUGUESE PUBLIC DEBT HELD BY RESIDENTS | BY INSTITUTIONAL SECTOR AND FINANCIAL INSTRUMENT
80 000
70 000
Other resident sectors
Currency and deposits
Debt securities
Loans
60 000
EUR million
I
50 000
Other financial institutions
Currency and deposits
Debt securities
Loans
40 000
30 000
20 000
Banks
Currency and deposits
Debt securities
Loans
10 000
0
Q4 Q4 Q4 Q4
2007 2008 2009 2010
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2011 2011 2011 2011 2012 2012 2012 2012
Source: Banco de Portugal.
Notes: Data considering nominal value and excluding accrued interest. Other financial institutions include the insurance corporations
and pension funds, financial auxiliaries and other financial intermediaries. The other resident sectors correspond to the non-financial
private sector, i.e., households and non-financial corporations. The concept of “currency and deposits” includes primarily saving certificates and deposits in the Direção Geral do Tesouro (DGT) and coins with legal tender in the country, which are not in possession
of the respective issuing authorities (DGT).
ARTICLES
IS THERE A RISK-TAKING CHANNEL OF MONETARY POLICY IN PORTUGAL?
INVESTMENT DECISIONS AND FINANCIAL STANDING
OF PORTUGUESE FIRMS – RECENT EVIDENCE
BANK INTEREST RATES ON NEW LOANS TO NON-FINANCIAL
CORPORATIONS– ONE FIRST LOOK AT A NEW SET OF MICRO DATA
II
IS THERE A RISK-TAKING CHANNEL OF MONETARY POLICY IN
PORTUGAL?*
87
Diana Bonfim** | Carla Soares**
Articles
Abstract
It is well established that when monetary policy is accommodative, banks tend to grant
more credit. However, only recently attention was given to the quality of credit granted
and, naturally, the risk assumed during those periods. This article makes an empirical
contribution to the analysis of the so-called risk-taking channel of monetary policy,
by testing whether Portuguese banks grant more risky loans when monetary policy
interest rates are lower. Our results show that banks grant more loans to non-financial
corporations with recent defaults or without credit history when policy interest rates are
lower. Even though these loans turn out to have higher ex-post default probabilities,
as expected, the overall loan portfolio does not show an increase in the likelihood
of default in the aftermath of a period of lower monetary policy rates. All in all, the
evidence on the risk-taking channel in Portugal is not as strong as in other countries
where similar methodologies were implemented. The results obtained are generally
more supportive of the credit channel hypothesis than of a pure risk-taking channel.
1. Introduction
Since the onset of the financial crisis, there has been an increasing interest on the links between the financial system and monetary policy. One of the recent avenues of research has focused on the transmission
of monetary policy through banks’ risk-taking behaviour, usually labelled as the risk-taking channel. The
basic idea is that in an environment of low policy interest rates, the incentive for banks to take more risk
into their balance sheets increases. In the last few years, the literature on this channel has flourished,
most notably in what concerns empirical studies. Several authors have found a negative relationship
between the level of monetary policy interest rates and bank risk-taking. Generally, the results suggest
that in the short-run lower policy interest rates decrease the total credit risk of the banking sector, since
the impact via the increase in borrowers’ repayment capacity for outstanding loans is more significant.
However, in the medium-term, the increased risk-taking may eventually materialize in a deterioration of
banks’ asset quality, especially when a period of low policy interest rates is followed by a recession or
by a severe monetary policy contraction.
This article intends to test whether there is a risk-taking channel in Portugal, adapting the methodology proposed by Jiménez, Ongena, Peydró and Saurina (2008). Using data on loans to non-financial
corporations from the Portuguese Central Credit Register for the period between 1999 and 2007, we
*
We are much indebted to Isabel Gameiro and João Sousa, who were actively involved in earlier stages of this
project, which would not have been possible without their support. The authors would also like to thank
Nuno Alves, António Antunes, Sandra Gomes, Ana Cristina Leal, and Nuno Ribeiro for insightful comments
and suggestions. The opinions expressed in the article are those of the author and do not necessarily coincide
with those of Banco de Portugal or the Eurosystem. Any errors and omissions are the sole responsibility of the
authors.
** Banco de Portugal, Departamento de Estudos Económicos.
assess whether banks grant riskier credit when policy interest rates are lower. This is a relevant issue for
central banks, as it allows analyzing the impact of its policy decisions in a broader perspective, while
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
88
also illustrating the links between financial stability and monetary policy. Our results partly support the
existence of a risk-taking channel in Portugal, at least ex-ante. When monetary policy interest rates are
lower, banks are more likely to grant credit to borrowers currently perceived as riskier. However, the
loans granted during periods of low interest rates do not show overall higher default probabilities, thus
not supporting the existence of a fully-fledged risk-taking channel in Portugal. As such, although there
is some increased risk-taking behaviour of banks when policy rates are at a lower level, this does not
translate into a worse ex-post performance of overall loan quality, thus suggesting that Portuguese banks
were not less prudent in their lending decisions. All in all, our results are more supportive of the credit
channel hypothesis than of a pure risk-taking channel.
The paper is organized in the following way. In section 2 we briefly summarize the theoretical and empirical
discussions in the literature on the risk-taking channel. Section 3 describes the dataset used and section
4 details the identification strategy and methodologies followed. Section 5 presents and discusses the
results. These are built on three blocks. First, we use discrete choice models to assess the probability of
borrowers with bad credit history or no credit history being granted loans when policy interest rates are
lower. Second, we test whether smaller banks are more prone to risk-taking in these periods. Third, we
conduct a survival analysis to assess the impact of monetary policy rates at the time of loan concession
on the time until a firm defaults. Section 6 summarizes our main findings.
2. Literature review
Since the onset of the financial crisis, there has been an increasing interest on the links between financial
stability and monetary policy.1 One of the recent avenues of research has focused on the transmission
of monetary policy through banks’ risk-taking (risk-taking channel). The basic idea is that in an environment of persistently low policy interest rates, the incentive for banks to take more risk into their balance
sheets increases.
The theoretical research on this channel has been expanding significantly during the last few years,
with some contributions coming from Dell’Ariccia et al. (2011), Borio and Zhu (2012), Adrian and Shin
(2008, 2010), De Nicolò et al., (2010). These authors have identified some mechanisms through which
this channel operates. One of these mechanisms is the search for yield, which occurs mainly through
the asset side of financial institutions’ balance sheet. A decrease in policy rates decreases their portfolio
income and then decreases the incentive to monitor, or similarly, increases search for yield and then
risk-taking (Dell’Ariccia et al., 2011). This is especially the case for financial institutions with long-term
commitments such as pension funds. When policy interest rates are low and expected to remain low
for an extended period of time, these institutions have incentives to invest in riskier assets in order to
increase their return and be able to meet their commitments. Otherwise they would have to renegotiate
or default (Brunnermeier, 2001, and Rajan, 2006). For instance, a bank may increase loan spreads, thus
ending up with a larger percentage of riskier borrowers (Freixas and Rochet, 2008).
The risk-taking channel may also operate through risk-shifting, occurring mainly via the liability side of
financial institutions’ balance sheet. A decrease in policy rates decreases the cost of banks’ liabilities.
The lower cost of funding gives them an incentive to increase leverage, the degree of which depends on
whether the capital structure is determined endogenously (because higher leverage increases funding
costs) (Dell’Ariccia et al., 2011, Valencia, 2011). Moreover, a prolonged period of low interest rates can
affect asset and collateral valuations, as it is associated to lower market volatility, thus reducing risk
perception (Gambacorta, 2009). Adrian and Shin (2008, 2010) argue that banks that actively manage
1 See Gameiro et al. (2011) for a literature review on these issues.
their balance sheets target a leverage ratio. When asset prices increase, the balance sheet gets stronger
and the leverage ratio decreases. This can be considered equivalent to “surplus capacity” relative to
manufacturing firms. Then, banks use their surplus capacity by increasing their market funding and by
expanding credit. With low policy rates, short-term funding is cheaper. In this setting, banks tend to
an increase in the risk they assume. This mechanism reinforces itself, since banks increase demand for
assets, increasing their price and consequently expanding further their balance sheet and lowering the
leverage ratio. In the Diamond and Rajan (2012) model, this mechanism operates solely with expectations
of low interest rates at times of financial stress, raising the same need for central bank intervention.
Such expectations create incentives to increase short-term leverage and illiquid loans, which increases
banks’ vulnerability in case households’ deposits withdrawals increase in the future. This happens in a
model where banking sector liquidity difficulties come from the mismatch between the long maturity of
loans and the demandable nature of households’ deposits, together with uncertainty about households’
future endowments (Diamond and Dybvig, 1983).
Other authors highlight a distortion of incentives in an environment of very low interest rates. In the
model of Acharya and Naqvi (2012), an agency problem between the bank manager and the principal
induces the bank manager to take excessive risk when the bank is awash with liquidity. This usually
occurs in situations of high macroeconomic risk, which may also lead the central bank to loosen its
monetary policy. When macroeconomic risk is high, there is a ‘flight to quality’ in the sense that agents
prefer deposits in banks instead of direct investment in projects, flooding the bank with liquidity. In this
situation, the bank manager sensitivity to the credit risk of loans decreases, leading to excessive credit.
This is equivalent to loans’ rate falling below its first best and asset prices rising above fundamentals. If
the central bank loosens monetary policy in this scenario, it is fuelling the asset price bubble and inducing
the excessive risk-taking by banks.
It should be noted that the risk-taking channel differs from the credit channel in several important
dimensions. The credit channel encompasses two different transmission mechanisms: the bank lending
channel and the balance sheet channel. In the former, a loosening of monetary policy via an expansion
in bank reserves would raise deposits and, consequently, the amount of bank loans. As more loans are
granted, more risky projects get financed, so the risk taken into banks’ balance sheet rises (Bernanke and
Blinder, 1988, Disyatat, 2011). In turn, the balance sheet channel is based on the financial accelerator
concept (Bernanke and Gertler, 1989, 1995). In this case, a monetary policy contraction reduces the net
worth of borrowers, amplifying the spending and production effects of the initial shock.
During the last few years, there were several relevant empirical contributions to the literature on the
risk-taking channel. Most of these empirical studies have found evidence that banks increase lending to
riskier borrowers when interest rates are low. For instance, using an extensive database on loans granted
by Spanish credit institutions, Jiménez et al. (2008) find robust evidence that low short-term interest
rates imply a softening in lending standards and an increase in loans to borrowers with bad or no credit
history. Moreover, they find that banks approve loans that have a higher ex-ante and ex-post probability
of default. Using a similar methodology for a Bolivian loans database, Ioannidou et al. (2009) also find
evidence that banks increase risk-taking when monetary policy rates are lower. This behaviour is apparent
in the increase in new loans with a higher probability of default, granted to riskier borrowers and with
lower loan spreads. There is also evidence of a risk-taking channel in the US, as Paligorova and Santos
(2012) show that banks offer relatively lower spreads when lending to riskier borrowers in periods of
lower short-term rates. In contrast, Buch et al. (2011) do not find evidence of increased risk-taking during
such periods in the US, for the banking sector as a whole, even though they find important differences
between different types of banks. Altunbas et al. (2010) use an interest rate gap in order to measure the
effect of monetary policy stance on banks risk-taking, using balance sheet data for a sample of banks
from 16 countries. They find that banks indeed tend to take more risk when interest rates are below the
89
Articles
increase the reliance on short-term funding, while expanding credit to cover riskier projects, thus implying
rate given by a Taylor rule. Using data from bank lending surveys of the euro area and the US, Maddaloni
and Peydró (2011) conclude that low short-term interest rates induce a softening in lending standards and
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
90
that this effect is more pronounced the longer is the period of low interest rates. Gaggl and Valderrama
(2011) use data on Austrian firms and banks to find that in relatively long periods of low policy interest
rates banks loan-portfolio risk increases, controlling for macroeconomic conditions, bank and industry
characteristics. Finally, Delis and Kouretas (2011) also find a negative relationship between the level of
interest rates and bank risk-taking.
From a broad risk perspective on the economy, there is evidence that in the short-run lower interest
rates decrease the total credit risk of the banking sector, since the impact via the decrease in the credit
risk of outstanding loans is more significant (Jiménez et al., 2008, Altunbas et al., 2010). However, in
the medium-term, the total credit risk may increase, especially when a period of low interest rates is
followed by a severe recession or monetary contraction (Jiménez et al., 2008, Altunbas et al., 2010).
Available empirical evidence suggests that there is some heterogeneity in bank risk-taking behaviour.
Jiménez et al. (2008) find that this behaviour is more pronounced for small and commercial banks, while
banks with more own funds and more liquidity are usually more precautionary regarding the loans granted.
Brissimis and Delis (2010) find that the reaction of credit risk of US and euro area banks with higher
liquidity and capitalization to monetary policy changes is approximately null, while on average banks’
credit risk increases (although marginally) with expansionary monetary policy. Altunbas et al. (2010) also
find that banks that are involved in more non-traditional banking activities take more risk. Buch et al.
(2011) find that only small domestic banks adopt risk-taking behaviours during periods of low interest
rates, while foreign banks decrease their risk-taking and large banks do not show a meaningful change
in behaviour. Ioannidou et al. (2009) observe some heterogeneity among Bolivian banks. They find that
the risk-taking effect when policy interest rates are low is stronger for banks more prone to agency
problems, i.e., larger banks, banks with a lower capital ratio or a higher non-performing loans ratio,
as well as banks with more liquid assets. Furthermore, Maddaloni and Peydró (2011) find relevance of
agency problems in excessive risk-taking, given that the impact of low monetary policy rates on lending
standards is amplified when supervision standards for bank capital are weaker.
Financial innovation also seems to impact on banks’ lending standards. Maddaloni and Peydró (2011)
find that securitization leads to softer lending standards in both the euro area and the US, amplifying
the effects coming from low policy rates (see also Delis and Kouretas, 2011).
Finally, there has also been some literature more focused on macro data. Angeloni, Faia and Lo Duca
(2010) present time series evidence for the US and the euro area about the effect of monetary policy on
measures of banks’ leverage and balance sheet risk. They found stronger evidence for the US than for
the euro area on the negative effect of monetary policy on banks’ risk.
Our article contributes to this literature by empirically testing the existence of a risk-taking channel in the
Portuguese banking system, using micro data on bank loans to non-financial corporations.
3. Data
We collect data for the period between 1999 and 2007. As discussed below, the identification strategy
used relies on the exogeneity of monetary policy, thus requiring using only data for the period after
Portugal joined the euro area. We chose to use data only up to 2007, as the transmission of monetary
policy has been severely impaired by the global financial crisis (and, more importantly, by the euro area
sovereign crisis). As such, we want to test the existence of a risk-taking channel of monetary policy in
“normal” conditions, while exploring the exogeneity of the interest rates set by the ECB Governing Council.
The most important data source for this article is the Portuguese Central Credit Register (CRC), which
is a database managed by Banco de Portugal, covering virtually all bank loans granted in Portugal (all
financial institutions granting credit in Portugal are required, on a monthly basis to report to the CRC all
loans granted above 50 euros). The register includes loans granted to firms and households, as well as
potential credit liabilities associated with irrevocable commitments. In this article, we consider only loans
granted to non-financial corporations, as default rates tend to be more cyclical than for households. All
their previous consent, thus making the CRC a key information-sharing mechanism between banks. The
CRC has information on the type of loan, the debtor and the amount, while also including information
on loan defaults and renegotiations.
To address our research question, we have to identify episodes of default. We consider that there is a
default when a loan is overdue or in litigation during an entire quarter. This avoids mining the data with
very short-lived episodes, possibly related to reporting errors or problems in bank payments, for instance.
We also use information on banks’ characteristics using supervisory quarterly balance sheet data. From
all credit institutions with activity during at least one year between 1999 and 2007, we select institutions
with a market share of at least 0.1 per cent in the corporate loan market. After this first selection, we have
a sample of 89 out of 346 credit institutions. From these, we select only monetary financial institutions,
keeping in the end 52 institutions, including 30 banks, 10 mutual agricultural credit banks (Caixas de
crédito agrícola mútuo), 1 savings bank (Caixa económica) and 11 branches of credit institutions with
head office in the EU.
Our unit of observation is a firm-bank relationship in a given quarter. We consider that there is a new
loan when there is an increase in the amount of credit granted by a bank to a firm or when there is a
new firm-bank relationship2. Using quarterly data for the period 1999-2007, we have almost 12 million
observations, representing 933 611 different firm-bank relationships. Default episodes account for 7.95
per cent of total observations. On average, each firm has a relationship with three banks and has credit
history for 12 quarters3. The average amount of each firm’s credit per bank is around 234 thousand euro,
thus suggesting that we are dealing mainly with micro and small enterprises.
Table 1 presents the definitions of all the explanatory variables considered in the analysis, as well as some
descriptive statistics. As discussed above, our analysis relies on several different methodologies, in order
to ensure the robustness of the results. These methodologies consider on different dependent variables,
all of which related to borrower’s credit quality: having recent default history (bad_hist), currently being in
default in any loan (D_default) or currently being in default with that specific bank (D_default_bank). The
most relevant explanatory variable for our analysis is the monetary policy interest rate. Several concepts
are considered: the ECB main refinancing rate at the end of each quarter, its quarterly average, and the
quarterly average of the EONIA.
We also control for a broad set of bank, borrower and loan characteristics. Regarding bank characteristics, we control for bank size (ln(assets)), liquidity (defined as liquid assets as a percentage of total
assets - liq ratio), credit quality (the non-performing loans ratio of the bank relative to the ratio for the
entire banking sector – rel npl/assets), solvency (capital/assets). We also control for the bank type (deposit
taking financial institution, savings bank, agricultural cooperative banks (CCAM) and subsidiaries from
EU countries (ICUE)), for mergers and acquisitions (M&A) and for the change to International Accounting
Standards (IAS). Borrower characteristics are based on the information available in the CRC: number
of bank relationships (#rel), total amount of credit granted to the firm (credit), and number of quarters
with credit history (age). Further, we control for the logarithm of loan amount (loan) and for the share
of long term credit (Credit_LT_prop). Finally, besides including a time trend in many regressions, we also
2 Unlike Jiménez et al. (2008), we do not have individual loans data, i.e., we cannot exactly identify when a new
loan contract is established or when an old one matures. Nevertheless, we consider that the relevant unit of
analysis would still be the relationship between the bank and the firm and not strictly the loan contract.
3 To compute the duration of credit histories we used data since 1995.
91
Articles
financial institutions are allowed to consult information on their current and prospective borrowers, with
Table 1 (continue)
VARIABLES DESCRIPTION AND SOME DESCRIPTIVE STATISTICS
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
92
Description
Unit
Obs.
Mean
Std. Dev.
Min
Max
Dummy =1 if the borrower had overdue credit
in the current and in the previous quarter; =
0 otherwise
-
11772002
0.109
0.311
0
1
D_default
Dummy =1 if the borrower had overdue credit
in the current quarter; = 0 otherwise
-
10806094
0.155
0.362
0
1
new_rel
Dummy =1 if the borrower started a new
bank relationship with the specific bank; = 0
otherwise
-
11772002
0.057
0.233
0
1
Dummy =1 if the borrower had an increase in
the total amount or a new bank relationship;
= 0 otherwise
-
11772002
0.305
0.461
0
1
Dummy =1 if the borrower had overdue credit
in the current quarter with the specific bank;
= 0 otherwise
-
11772002
0.080
0.271
0
1
Dependent variables
Probit
bad_hist
Condition
new_loan
Survival
D_default_
bank
Independent variables
Monetary Policy Rates
i ECB eoq
ECB main refinancing rate at the end of the
quarter
%
11772002
2.978
0.885
2
4.75
i ECB av
Quarterly average of the ECB main refinancing
rate
%
11772002
2.963
0.869
2
4.75
i EONIA av
Quarterly average of the EONIA
%
11772002
3.025
0.877
2.01
4.84
EUR
11536811
23.419
1.662
16.70
25.19
%
11536811
18.475
10.809
0.00
82.87
rel npl/assets Difference between the bank ratio of non
performing loans over total assets and the
average ratio for all banks
%
11536811
-1.953
2.250
-3.79
22.55
capital/assets Ratio of the balance sheet capital over total
assets
%
11432772
4.819
2.462
0.07
37.99
-
11772002
0.033
0.179
0
1
-
11772002
0.023
0.150
0
1
Dummy = 1 if the bank is a branch of a credit
institution with head office in the EU; = 0
otherwise
-
11772002
0.037
0.189
0
1
Dummy = 1 if the banks was involved in
a merger in the respective quarter; = 0
otherwise
-
11772002
0.051
0.220
0
1
Dummy = 1 for the quarter in which the bank
switched from the old accounting standars
to the IAS
-
11772002
0.032
0.175
0
1
Bank characteristics
ln(assets)
Logarithm of the total assets of the bank
liq ratio
The amount of liquid assets over total assets.
Included in total assets: cash, balances with
the central bank, loans and advances to credit
institutions, loans and advances to the public
sector, gold and other precious metals for the
old accounting standards; cash, loans and
advances to credit institutions and other loans
and advances for the IAS.
savings
Dummy = 1 if the bank is a saving bank; = 0
otherwise
CCAM
Dummy = 1 if the bank is a mutual agricultural
credit bank; = 0 otherwise
ICUE
M&A
IAS
Table 1 (continuation)
VARIABLES DESCRIPTION AND SOME DESCRIPTIVE STATISTICS
Unit
Obs.
Mean
Std. Dev.
Min
Max
11772002
3.057
2.424
1
38
11772002 1 040 303 12.8 x 106
0
4.5 x 109
Borrower characteristics
93
#rel
Number of bank relationships of the
firm
credit
The total amount of credit of the firm
age
Number of quarters that the firm has
credit
11772002
23.785
13.510
0
51
ln(1+#rel)
Logarithm of 1 plus the number of
bank relationships of the firm
11772002
1.264
0.499
0.693
3.66
ln(credit)
Logarithm of the total amount of credit
of the firm
10806094
11.139
2.763
-29.934
22.23
ln(2+age)
Logarithm of 2 plus the number of
quarters that the firm has credit
11772002
3.048
0.730
0.693
3.97
4.4 x 106
0
4.5 x 109
EUR
Loan characteristics
loan
Total credit granted by the bank to the
borrower
11772002 234 358
ln(1+loan)
Logarithm of 1 plus the total credit
granted by the bank to the borrower
11772002
8.457
4.201
0
22.23
%
10222954
48.769
39.713
0
100
Cred_LT_prop Share of long term credit on the sum of
short and long-term credit
Macro controls
GDP PT
Portuguese GDP y-o-y quarterly growth
rate
%
11772002
1.612
1.592
-1.90
5.10
π PT
Quarterly inflation rate (HICP)
%
11772002
2.926
0.702
1.90
4.40
10y PT av
Quarterly average of the 10-year
Portuguese Government bond yield
%
11772002
4.427
0.684
3.17
5.75
10y PT eoq
10-year Portuguese Government bond
yield at the end of the quarter
%
11772002
4.424
0.700
3.12
5.62
%
11772002
10.939
8.788
0.80
29.00
Robustness
NFC credit PT Quarterly growth of credit to non
financial corporations in Portugal
house p PT
Quarterly growth in house prices in
Portugal
%
11772002
2.922
2.943
0.00
9.83
GDP EA
One year ahead forecast for the euro
area GDP based on the Eurosystem MPE
%
11772002
2.204
0.464
1.21
3.40
Sources: Banco de Portugal and authors’ calculations.
Articles
Description
consider the effect of Portuguese GDP growth and inflation.4 Table 1 also includes the description of
additional variables used for robustness analysis.
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
94
4. Identification strategy
Our main objective is to test if there is a risk-taking channel in Portugal. In other words, we want to
assess whether banks grant riskier credit when policy interest rates are lower, either due to very low risk
aversion or due to search for yield strategies. Taken at face value, this would mean regressing variables
that capture bank risk-taking on the level of interest rates. However, to correctly identify the causal effect
of monetary policy on bank risk-taking, monetary policy decisions need to be exogenous. Otherwise, it
is possible that there are (omitted) variables that simultaneously affect monetary policy and bank risk-taking decisions. Our setup allows us to avoid this potentially serious endogeneity problem, as monetary
policy is fully exogenous during the period analysed. Portugal is a small open economy that joined the
euro area in 1999. The impact of macroeconomic and financial conditions specific to the Portuguese
economy on euro area interest rates should be negligible. As such, it is easy to argue that monetary
policy is exogenous, thus allowing for the correct identification of this causal effect.
This is the same argument used by Jiménez et al. (2008) and, to some extent, by Ioannidou et al. (2009)
and Geršl. et al. (2012). Indeed, this article closely replicates their empirical strategy, with the objective of
testing whether there is a risk-taking channel in the Portuguese economy. As such, our methodological
strategy lies on three main blocks, as in Jiménez et al. (2008).
First, we use discrete choice models to assess the probability of borrowers with bad credit history or no
credit history being granted loans.5 This approach allows us to test whether banks grant more loans to
riskier borrowers during periods of lower policy interest rates6. Our dependent variable takes the value
one when a new loan is granted to a borrower defined as risky (and zero when a new loan is granted
to any other borrower). It is important to note that the information in the CRC is shared between participating institutions, so that a bank is able to know whether a firm is currently defaulting on any loan,
as well as whether the firm has any other outstanding loans.
Second, we explore the results of within borrower regressions, to test whether smaller banks are more
prone to risk-taking when policy interest rates are lower. We are able to do this because most borrowers
have more than one bank relationship, thus allowing us to test the behaviour of different banks towards
the same firm.
Third, we conduct a survival analysis to assess the impact of monetary policy rates on the time until a
firm defaults. Whereas in the first two parts we examine the probability of granting a loan to borrowers
that show recent evidence of acute financial distress, materialized in a default, or for which there is
no credit history available, in this part we examine the ex-post credit quality of the borrowers. In other
words, we examine whether loans granted in periods of lower policy interest rates display higher future
default probabilities.
4 We also computed a country risk measure (the spread between 10-year Portuguese and German government
bond yields), but this variable was not included in the results presented in this article due to its high correlation
with GDP growth.
5 Granting loans to borrowers with limited historical data increases the expected profitability of banks, while
fostering innovation, as shown by Thakor (2013). However, it also increases the risk held by banks.
6 It should be noted that this analysis may be somewhat biased due to selection issues, as the data includes only
approved loans. However, our dataset does not allow us to overcome this problem.
5. Results
5.1 Granting loans to (ex-ante) risky borrowers
We evaluate the probability of a new loan being granted given that the borrower has a recent bad
credit history or that the borrower has no credit history. We consider that there is a recent bad credit
history when the borrower has some credit overdue in the current and in the previous quarter. Since
borrowers’ credit situation can be verified by any bank through the CRC, we consider that there is bad
credit history when the firm is defaulting on any bank loan, i.e., not only on the bank offering the new
loan. We are interested in studying how monetary policy rates in the quarter prior to loan origination
influences the probability of granting loans to these higher risk borrowers. To more accurately identify
this effect, we control for several bank, borrower and loan characteristics and also for macro variables
(defined in detail in Table 1).
Table 2 presents the results of the estimation using as dependent variable the dummy bad_hist, which
equals one when the borrower has credit overdue in the current and previous quarter. We find that
lower short-term interest rates increase the probability of banks granting a loan to a borrower with
recent episodes of default on loans. This result is quite robust to different specifications, namely if one
considers either the ECB main reference rate, in end-of-quarter (columns I and II), average quarter values
(column III), or the quarterly average of the EONIA rate (column IV). This impact is slightly higher than
the one found by Jiménez et al. (2008). We find consistent evidence that banks increase lending to firms
that were riskier in the recent past when the level of monetary policy rates is lower. If this corresponds
truly to bank risk-taking, then, from a prudential viewpoint, it suggests that loose monetary policy may
contribute to the increase of risks in banks’ balance sheets, thus sowing the seeds for a potential future
deterioration of banks’ asset quality. However, it is possible that this result does not necessarily imply
a risk-taking channel, but may rather be evidence of a credit channel. Indeed, these results may simply
imply that banks increase overall lending when interest rates are lower, including also to firms with a
higher net worth, under a low interest rate environment. This result is in line with previous evidence
obtained for Portugal by Farinha and Marques (2003) on the credit channel.
There seems to be a negative relationship between bank size, measured by the log of assets, and the
probability of granting a loan to a riskier borrower. Indeed, the effect of monetary policy on this probability
is more pronounced when we include an interaction term between the short-term interest rate and the
size of the bank (column II). Indeed, the coefficient of this term is slightly positive and the coefficient for
the policy variable decreases further, meaning that the probability of granting a loan to a risky borrower
is higher for smaller banks, i.e., these banks tend to take on more risk when monetary policy rates are
lower. Under this specification, the negative coefficient obtained from the log of assets is also reinforced.
Regarding banks’ balance sheets, we find that more capitalised banks have a larger probability of granting loans to riskier borrowers. This result is somewhat counterintuitive, but it may suggest that these
banks may have a greater leeway for taking on more risk. We also observe that banks with a higher
liquidity ratio tend to take less risk on granting loans. Moreover, banks with a relatively higher share of
non-performing loans also tend to be more careful. In what concerns the type of banks, we see that
mutual agricultural credit banks are relatively more prudent in their lending decisions.
Regarding borrower characteristics, the results are broadly in line with Jiménez et al. (2008): borrowers
with more outstanding credit, more bank relationships and a longer credit history have a higher probability
of being granted a new loan when they have a recent bad credit history.7 Regarding loan characteristics,
7 This result is also consistent with Bonfim et al., (2012), who find that, after default, banks are willing to extend
credit faster to larger and older firms, as well as those with more bank relationships.
95
Articles
In this section, our analysis is based on the estimation of discrete choice models for new bank loans.
Table 2
RESULTS OF THE PROBIT ESTIMATION
II
Dependent variable: bad_hist
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
96
Dependent variable:
new_rel
I
II
III
IV
V
VI
VII
VIII
Coef.
Coef.
Coef.
Coef.
Coef.
Coef.
Coef.
Coef.
S.e.
S.e.
S.e.
S.e.
i ECB eoq t-1
Dependent variable:
default
-0.043***
-0.208***
0.003
0.032
S.e.
S.e.
-0.029***
S.e.
-0.146***
S.e.
0.087***
-0.871***
0.003
0.027
0.002
0.014
-0.033***
i ECB av t-1
0.004
-0.031***
i EONIA av t-1
0.003
i*ln(assets) t-1
0.007***
0.005***
0.001
0.001
0.041***
0.001
-0.042***
-0.064***
0.002
0.005
0.002
0.002
0.002
0.004
0.001
0.002
-0.007***
-0.007***
-0.007***
-0.007***
-0.006***
-0.006***
-0.015***
-0.015***
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
-0.008**
-0.010**
-0.008**
-0.008**
-0.004
-0.005
-0.029***
-0.037***
0.002
0.002
0.002
0.002
0.001
0.001
0.001
0.001
0.021***
0.022***
0.021***
0.021***
0.022***
0.023***
0.020***
0.023***
0.001
0.001
0.001
0.001
0.001
0.001
0.000
0.000
0.296***
0.299***
0.296***
0.296***
0.250***
0.252***
-0.130***
-0.115***
0.014
0.014
0.014
0.014
0.013
0.013
0.006
0.006
-0.070***
-0.071***
-0.070***
-0.070***
-0.086***
-0.087***
-0.093***
-0.108***
0.021
0.021
0.021
0.021
0.017
0.017
0.008
0.008
-0.057
-0.051
-0.057
-0.057
-0.191***
-0.186***
-0.296***
-0.275***
0.029
0.029
0.029
0.029
0.025
0.025
0.008
0.008
-0.075***
-0.071***
-0.075***
-0.075***
-0.082***
-0.079***
0.022***
0.040***
0.005
0.005
0.005
0.005
0.005
0.005
0.004
0.004
-0.055***
-0.053***
-0.052***
-0.051***
-0.016***
-0.015***
0.398***
0.409***
0.007
0.007
0.007
0.007
0.006
0.006
0.005
0.005
0.021***
0.021***
0.021***
0.021***
0.021***
0.021***
0.002
0.002
0.002
0.002
0.002
0.002
0.357***
0.357***
0.356***
0.356***
0.380***
0.380***
0.010
0.010
0.010
0.010
0.009
0.009
0.111***
0.112***
0.112***
0.112***
0.035***
0.035***
0.004
0.004
0.004
0.004
0.004
0.004
-0.082***
-0.082***
-0.082***
-0.082***
-0.076***
-0.076***
-0.169***
-0.170***
0.001
0.001
0.001
0.001
0.001
0.001
0.000
0.000
0.006***
0.006***
0.006***
0.006***
0.006***
0.006***
0.002***
0.002***
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
-0.019***
-0.019***
-0.023***
-0.024***
-0.011***
-0.012***
-0.005***
-0.004***
0.002
0.002
0.002
0.002
0.001
0.001
0.001
0.001
0.078***
0.077***
0.072***
0.071***
0.105***
0.104***
0.026***
0.017***
0.004
0.004
0.004
0.004
0.003
0.003
0.002
0.002
-0.014***
-0.014***
-0.013***
-0.012***
-0.011***
-0.010***
-0.018***
-0.015***
0.002
0.002
0.002
0.002
0.001
0.001
0.001
0.001
0.000***
0.000***
0.000***
0.000***
0.000***
0.000***
0.000***
0.000***
0.000
0.000
0.000
0.000
0.000
0.000
0.000
0.000
-0.929***
-0.432***
0.939***
-0.940***
-0.474***
-0.117***
3.804***
6.818***
0.058
0.119
0.058
0.058
0.050
0.101
0.023
0.052
Nº obs.
2 655 604
2 655 604
2 655 604
2 655 604
2 655 604
2 655 604
3 320 469
3 320 469
Log
pseudolikel.
-660 740
-660 710
-660 807
-660 810
-859 858
-859 839
-1 342 552
-1 339 995
Prob > chi2
0
0
0
0
0
0
ln(assets) t-1
liq ratio t-1
rel npl/assets t-1
capital/assets t-1
savings t-1
CCAM t-1
ICUE t-1
M&A t
IAS t
ln(credit) t-1
ln(1+#rel) t-1
ln(2+age) t-1
ln(1+loan) t
share LT credit t
GDP PT t-1
π PT t
trend
trend2
constant
-0.042***
-0.042***
-0.054***
-0.069***
-0.132***
0
-0.262***
0
Sources: Banco de Portugal and authors’ calculations.
Notes: * significance at 10 per cent; ** significance at 5 per cent; *** significance at 1 per cent. All variables defined in Table 1.
riskier borrowers are more likely to be given a new loan when the amount of the loan is smaller and
when they have a larger share of long-term credit.
In case one considers the probability of granting credit to firms defaulting in the current quarter instead
theless, the effect of the monetary policy variable on risk-taking is slightly smaller.
97
However, when assessing the probability of a new firm-bank relationship being established (columns
Articles
of in the current and previous quarter (columns V and VI), the results remain broadly unchanged. None-
VII and VIII), the results differ slightly. When estimating solely with the ECB interest rate in the previous
quarter (column VII), there seems to be no risk-taking channel operating. However, when we include
the interaction term of the interest rate with bank size (column VIII), the coefficient on the ECB interest
rate becomes negative and much higher (in absolute terms) than in the regression with bad history as
a dependent variable. The coefficients for the log of assets and for the interaction term are also higher
(in absolute terms). This may suggest that mostly smaller banks take on more risk on granting loans to
new borrowers when monetary policy rates are lower.
The GDP growth coefficient is negative. When economic activity is stronger, there should be a larger
pool of “good” borrowers. As such, banks can increase lending volumes mainly through these higher
quality borrowers, reducing the overall likelihood of granting loans to riskier borrowers. On the contrary,
when inflation is higher, one might expect that the increased costs of debt leads to a higher proportion
of riskier borrowers, thus increasing the probability of granting a loan to a riskier borrower. Finally, the
trend coefficient is negative, meaning that over time banks tend to grant fewer loans to riskier borrowers.
For robustness purposes, we considered another empirical test of the risk-taking channel.8 Instead of
focusing on the probability of granting loans to borrowers with weaker credit quality, we focused on the
determinants of loan growth, at the firm level. We found that loan growth is higher when interest rates
are lower and when the firm has a good track record in terms of credit quality, as expected. However, the
interaction between these two variables provides some evidence in favour of a risk-taking channel, i.e.,
when interest rates are lower bad quality borrowers face less discrimination in terms of access to credit.9
All in all, the results of the discrete choice models do not reject the hypothesis of a risk-taking channel
in Portugal, as there is an increased lending activity to ex-ante riskier borrowers in periods during which
monetary policy rates are lower.
5.2 Within borrower comparison
Following the empirical strategy of Jiménez et al. (2008), we also conduct a within borrower comparison
in order to test whether smaller banks tend to have a riskier behaviour, as also suggested by the results
of Buch et al. (2011). Given that many firms borrow from more than one bank, we are able to explore
changes in lending behaviour by small and large banks, when banks are lending to the same borrower.
In this approach, the dependent variable is the quarterly change in the difference between the percentages of loans from small and large banks.10 In case the firm’s funding needs changes, ceteris paribus,
there is no reason to expect a change in the share of credit obtained from large or small banks. Thus,
this change is expected to be null in case only borrowers’ demand changes. Otherwise, we would have
evidence of a group of banks with a clear incentive to increase risk.
Table 3 presents the results of the panel data estimations with fixed effects at the borrower level and
robust standard errors. The table presents two specifications, one including all borrowers with multiple
8 The results are not reported, but are available upon request.
9 The coefficients on the policy interest rate, on the bad credit history and on the interaction term between the
two previous are -0.008, -0.045 and 0.009, respectively.
10 We define a small/large bank as being below/above the median asset size in each quarter.
Table 3
RESULTS OF THE WITHIN BORROWER COMPARISON ANALYSIS
II
All borrowers with multiple bank
relationships
Borrowers with small and large
banks
Coef.
Coef.
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
98
S.e.
i ECB eoq t-1
i*bad_hist t-1
bad_hist t-1
ln(credit) t-1
GDP PT t-1
trend
trend2
S.e.
-0.001***
-0.005***
0.000
0.001
0.001***
-0.004
0.000
0.002
0.000
0.025***
0.001
0.007
0.003***
0.045***
0.000
0.002
-0.003***
0.010***
0.000
0.001
-0.002***
0.001
0.000
0.001
0.000***
0.000***
0.000
0.000
constant
-0.014***
-0.440***
0.002
0.028
Nº obs.
3 035 927
390 103
Log pseudolikel.
0.0004
0.0006
Prob > chi2
0
0
Sources: Banco de Portugal and authors’ calculations.
Notes: *significance at 10 per cent; ** significance at 5 per cent; *** significance at 1 per cent. The dependent variable is the
quarterly change in the difference between the percentages of borrowing from small and large banks. The first column regression
includes all borrowers with multiple bank relationships; the second column includes only borrowers with relationships with at least
one large and one small bank.
bank relationships (first column with results) and another including only borrowers with relationships
with at least one large and one small bank (second column). The coefficient on the ECB interest rate is
negative and significant, but very low, thus suggesting that there is a slight increase in loan supply by
small banks to all borrowers following an expansion in monetary policy (first column). This effect is more
relevant for firms that have loans outstanding with both small and large banks (second column). The
coefficient for the bad_hist dummy goes along the same lines: it is only significant for firms with loans
from both small and large banks and it is positive, suggesting that small banks take more risk than large
banks. The interaction term between the interest rate and the recent bad credit history does not reinforce the risk-taking effect when monetary policy rates are lower. It is only significant for the regression
including all borrowers with multiple relationships and it has a positive coefficient, thus mitigating the
risk-taking effect (first column).
All in all, there is some evidence of a more aggressive behaviour of small banks on loan granting activity,
which tends to amplify slightly in periods of lower monetary policy rates.
5.3 Granting loans to (ex-post) risky borrowers
Whereas in the previous two subsections we explored the existence of a risk-taking channel of monetary
policy in Portugal by assessing how the likelihood of banks granting loans to new borrowers or to borrowers with a recent bad credit history is affected by a low interest rate environment, in this subsection
we look for another dimension of bank risk-taking, relating to the probability of granting loans when
policy interest rates are lower to borrowers that eventually default in the future. To assess this we use
survival analysis, modelling the hazard rate of the loans granted to the firms, where the failure event
is the occurrence of default. The hazard function is defined as the instantaneous probability of a firm
defaulting on the bank conditional on having no default up to time t.
We consider that a new loan is granted whenever the credit outstanding increases or a new firm-bank
relationship is established. A default occurs when the bank classifies a loan as being overdue or in litigation. The time at risk is defined as the time elapsed between these two events. However, it is important
same bank. We consider that the relevant unit of analysis is the firm-bank relationship instead of the
individual loan, given that default in a loan, under certain conditions, may represent also a credit event
at the borrower level, from the banks’ risk management and provisioning perspectives.
Following Jiménez et al. (2008), we estimate a parametric model with a Weibull distribution, which allows
for a monotonic hazard function, i.e., the hazard rate either increases or decreases over time according
to the Weibull distribution parameter. The Weibull hazard function is given by
h(t )  pt p 1
γ
where
is parameterized as i  exp( xi  ) . In case p>1 (p<1), the hazard function is monotonically
increasing (decreasing). For robustness, we also estimated a Cox proportional hazard model.
Even though we observe the beginning of the time at risk for all firm-bank relationships (i.e., when a
new loan is granted), there is naturally a lot of right censoring, as the majority of loans do not record
any default during the sample period. This was taken into account in the estimations.
Table 4 presents the results of the survival estimation. Columns I to IV present the specifications with time
invariant covariates.11 We also estimated the equations with time varying covariates (columns V and VI)
and only with macro variables (GDP and inflation) varying over time (columns VII and VIII).
The most striking observation is that lower policy interest rates prior to loan concession decrease the
hazard rate, in most specifications (the only exceptions are columns V and VI, where all the variables
vary over time). The effect is more pronounced when only macro controls are included (column I) than
when we include bank, borrower and loan characteristics (columns II, III and IV). The regressions with
the variables fixed for the moment of loan concession and including bank, borrower and loan characteristics do not show a significant effect of the policy interest rate, either taking into account borrower
heterogeneity or not (columns III and IV). When we include inflation varying through the life of the loan,
the coefficient on the interest rate turns statistically significant (column VII). In sum, the survival analysis
results do not support the hypothesis of a risk-taking channel in Portugal, as loans granted during
periods of lower interest rates do not show higher default probabilities over time. As mentioned above,
the only exception to these results comes from the specifications with time-varying covariates (columns
V and VI). However, in this specification we are explicitly considering the role of changing firm, bank
and macro characteristics over the life of the loan. As these changes could not be fully anticipated by
the bank when deciding to grant a loan, it is not reasonable to argue that banks were taking more risk
based solely on these two specifications.
Even though the survival analysis is not generally supportive of the existence of a credit risk-taking
channel in Portugal, it is important to note that these results are not necessarily in contradiction with
the results from the probit models. In the first part of our analysis, we used discrete choice models to
assess how monetary conditions influence loan concession to observable ex-ante riskier borrowers. In
this section, we are evaluating how monetary policy rates at loan concession affects borrowers ex-post
probability of defaulting, increasing the credit risk implicit in banks’ balance sheet. As banks do not have
11 The size of the sample decreases substantially when we include bank, loan and borrower characteristics fixed
at the moment prior to the loan concession (e.g., columns I and II compared to columns III and IV) mainly due
to two reasons: (i) we do not have the lagged data in the beginning of the sample or (ii) there are some periods
for which we do not have data on banks’ capital.
99
Articles
to note that it is possible that the default occurs with respect to another loan previously granted by the
Table 4 (continue)
SURVIVAL ANALYSIS RESULTS
II
Non-time varying
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
100
i ECB eoq
(loan) t-1
Time-varying
Time-varying
GDP and π
I
II
III
IV
V
VI
VII
VIII
Coef.
Coef.
Coef.
Coef.
Coef.
Coef.
Coef.
Coef.
S.e.
S.e.
S.e.
S.e.
S.e.
S.e.
S.e.
S.e.
0.126***
0.071***
0.016
0.019
-0.013***
-0.055***
0.062***
0.056***
0.005
0.005
0.017
0.022
0.005
0.005
0.018
0.018
-0.034***
0.024**
0.080***
-0.057***
-0.032***
0.084***
0.092***
ln(assets) t-1
liq ratio t-1
rel npl/assets t-1
capital/assets t-1
savings t-1
CCAM t-1
ICUE t-1
M&A t
IAS t
0.003
0.011
0.016
0.004
0.003
0.015
0.015
-0.011***
-0.007***
-0.004***
-0.011***
-0.014***
-0.004***
-0.004***
0.000
0.001
0.001
0.001
0.001
0.001
0.001
0.041***
0.095***
0.235***
0.038***
0.036***
0.222***
0.213***
0.001
0.010
0.016
0.002
0.002
0.016
0.015
0.035***
0.067***
0.106***
0.034***
0.045***
0.112***
0.117***
0.001
0.007
0.009
0.001
0.001
0.009
0.008
0.175***
0.332***
0.231***
0.001*
-0.019
0.267***
0.289***
0.017
0.052
0.074
0.023
0.023
0.074
0.072
-0.357***
-0.108
0.098
-0.304***
-0.199***
0.140
0.157
0.025
0.105
0.118
0.032
0.032
0.117
0.114
-0.048**
0.247***
0.678***
0.203***
0.306***
0.636***
0.653***
0.021
0.093
0.128
0.025
0.024
0.127
0.124
0.054***
0.003
-0.048
0.013
-0.056***
-0.116*
-0.088
0.015
0.061
0.079
0.018
0.018
0.078
0.076
0.043**
0.451**
0.446*
0.038*
-0.006
0.370
0.559**
0.020
0.208
0.250
0.021
0.020
0.250
0.244
-0.002***
0.004***
-0.082***
-0.082***
0.006***
0.004***
ln(credit) t-1
ln(1+#rel) t-1
bad_hist t-1
0.005
0.009
0.002
0.002
0.009
0.009
0.271***
0.464***
0.667***
0.641***
0.458***
0.445***
0.025
0.039
0.013
0.013
0.039
0.038
1.821***
2.350***
1.483***
1.515***
2.344***
2.368***
0.036
0.059
0.013
0.013
0.058
0.057
ln(2+age) t-1
-0.471***
-0.735***
-0.350***
-0.329***
-0.731***
-0.666***
0.025
0.038
0.007
0.007
0.038
0.036
ln(1+loan) t
0.023
0.063***
0.175***
0.175***
0.064***
0.061***
0.005
0.007
0.002
0.002
0.007
0.007
0.002***
0.002***
0.006***
0.006***
0.002***
0.002***
0.000
0.000
0.000
0.000
0.000
0.000
-0.073***
share LT credit t
GDP PT (loan)t-1
GDP PT t
π PT t
trend
trend2
constant
p
θ
-0.075***
-0.058***
-0.051***
-0.033***
-0.056***
0.003
0.003
0.010
0.012
0.011
0.010
0.004
-0.002
-0.039***
-0.049***
-0.035***
0.011***
-0.003
0.013***
0.003
0.003
0.008
0.008
0.004
0.003
0.008
0.006
-0.070***
-0.036***
0.095***
0.130***
0.171***
0.101***
0.149***
0.046***
0.006
0.014
0.013
0.006
0.006
0.020
0.027
0.007
-0.012***
-0.015***
-0.063***
-0.059***
-0.023***
0.003
0.003
0.007
0.007
0.003
0.007
0.000***
0.000***
0.001***
0.002***
0.001***
0.002***
-0.072***
0.000
0.000
0.000
0.000
0.000
-4.166***
-3.395***
-4.489***
-6.199***
-5.333***
-5.721***
-6.456***
0.000
0.103
0.069
0.291
0.395
0.096
0.089
0.394
0.381
1.037***
1.042***
1.173***
1.350***
1.334***
1.375***
1.325***
1.370***
0.003
0.003
0.013
0.012
0.005
0.005
0.012
0.120
0.435***
3.956***
2.126***
2.035***
3.905***
3.430***
0.094
0.142
0.031
0.030
0.140
0.108
-7.004***
Table 4 (continuation)
SURVIVAL ANALYSIS RESULTS
Time-varying
Time-varying
GDP and π
I
II
III
IV
V
VI
VII
VIII
Coef.
Coef.
Coef.
Coef.
Coef.
Coef.
Coef.
Coef.
S.e.
S.e.
S.e.
S.e.
S.e.
S.e.
S.e.
S.e.
shared frailty
bank
yes
no
no
no
no
no
no
no
shared frailty
NFC
no
no
no
yes
yes
yes
yes
yes
Nº obs.
7 193 128
7 087 951
1 384 696
1 384 696
5 833 210
5 833 210
1 384 696
1 384 696
Log
pseudolikel.
-363 163
-358 391
-46 713
-44 823
-224 247
-224 696
-44 780
-44 937
Prob> chi 2
0
0
0
0
0
0
0
0
Sources: Banco de Portugal and authors’ calculations.
Notes: * significance at 10 per cent; ** significance at 5 per cent; *** significance at 1 per cent. t refers to the moment when the
loan is granted. i ECB eoq (loan) and GDP PT (loan) are fixed to the moment prior to the loan concession. All variables defined in
Table 1.
perfect foresight on borrower quality, the risk-taking behaviour on these two situations is quite different: whereas in the former banks were granting loans to borrowers which clearly had poor quality, the
decision might not have been so clear in the latter case. Therefore, results are not entirely contradictory.
A possible interpretation is that even though Portuguese banks grant credit to riskier borrowers when
monetary policy rates are lower, they are not necessarily increasing the overall risk of their loan portfolio,
but instead they consider that these ex-ante riskier borrowers become more attractive as their net worth
increases, for instance (balance sheet channel). Furthermore, it is important to note that there is a volume
effect associated with the bank lending channel: as banks grant more loans, there is necessarily more
heterogeneity in borrower quality. Given these arguments, we could have evidence more in favour of a
bank lending and balance sheet channel (already documented in Farinha and Marques, 2003) than of a
risk-taking channel operating in Portugal.
Higher GDP growth in Portugal, both at the moment of the loan concession and during the life of the loan,
decreases the hazard rate, in the generality of the specifications. This is broadly in line with the literature
and previous evidence found for Portugal (Bonfim, 2009). When we only include macro and/or bank
characteristics, GDP growth over the life of the loan is not found to be statistically significant (column I).
The expected coefficient on inflation is not clear. One could consider that higher inflation could reduce
the probability of default because it reduces the real value of debt. Alternatively, higher inflation is usually
associated with higher nominal interest rates, increasing the nominal cost of debt and thus may increase
the probability of default. We find that, when taking into account firm and loan characteristics, inflation
both at the moment of the loan concession and during the life of the loan has a positive coefficient,
i.e., higher inflation increases the hazard rate. Regarding bank characteristics, it is worth referring to the
coefficients on capital and on non-performing loans ratio, which are significant and consistent across
specifications. In line with the probit analysis, more capitalized banks tend to grant loans with a higher
hazard rate. However, in contrast with the probit results, banks with more non-performing loans relative
to the whole sector also tend to take more risk when granting loans. In turn, banks with a higher liquidity
ratio tend to be more prudent as they seem to be exposed to loans with lower hazard rates. The results
on bank size are not very stable across specifications.
Turning to firm and loan characteristics, we find that recent bad credit history is a highly relevant borrower
characteristic for the loan hazard rate. The coefficient turns out high and highly significant regardless
the specification. Thus, firms that have defaulted on loans in the recent past are also much more likely
to default in the future, as shown by Bonfim et al. (2012). We observe that borrowers with more bank
101
Articles
Non-time varying
relationships (usually larger firms) tend to be riskier, what may be somewhat counterintuitive. Instead,
in line with previous evidence firms with a longer credit history tend to be less risky. Finally, it is worth
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
102
mentioning that, in line with the probit analysis, firms with a higher share of long-term credit tend also
to present loans with a higher hazard rate.
The Weibull distribution parameter p is greater than one and, therefore, the hazard function is monotonically increasing in all specifications. This means that, after controlling for bank, borrower and loan
characteristics, macro variables and policy interest rate level, the probability of the firm defaulting on
the loan increases over time.
We also performed additional robustness tests which we do not report, since the main conclusions are
not significantly affected. When we include interaction terms between the policy rate and some bank
or firms’ characteristics (bad history, age as borrower, banks’ assets, liquidity rate, relative NPL and type)
the conclusions are broadly the same. The effect of the policy rate is no longer relevant in any of these
specifications. Only the interaction term with the liquidity ratio shows up as relevant. We also controlled
for other macro variables, namely credit growth, house prices growth, euro area GDP forecasts and
long-term rates, but results are similar to the reported ones. Given that Portuguese banks can observe
in the CRC the current credit status of borrowers in their outstanding loans and that we do not follow
exactly each loan but a borrower-bank relationship, we also conducted the survival analysis considering
as the failure event a default of the firm with any bank. The coefficient on the interest rate turns out
even higher. The results of the Cox regression do not provide any relevant addition to our results.
6. Concluding remarks
In this paper, we tested whether Portuguese banks take more risk in their balance sheets when monetary
policy interest rates are lower. The analysis was based on three major blocks: (i) discrete choice models to
assess the probability of borrowers with bad credit history or no credit history being granted loans, (ii) a
regression to test whether smaller banks are more prone to risk-taking when policy interest rates are lower
and (iii) a survival analysis to assess the impact of monetary policy rates on the time until a firm defaults.
The results of the discrete choice models show that lower short-term interest rates increase the probability
of banks granting a loan to a borrower with recent bad credit history and this result is quite robust to
different specifications. Smaller banks tend to grant more loans to ex-ante riskier borrowers than larger
banks when monetary policy is looser. When we look only at new firm-bank relationships, we also
conclude that it is mostly the smaller banks that take more risk on granting loans to new borrowers when
monetary policy rates are lower. These results support the hypothesis of the existence of a risk-taking
channel in Portugal. However, they are not entirely conclusive, since under low interest rate environments
banks may increase credit to riskier firms because both of a volume effect and of an increase in firms’
net worth. Thus, these results may simply support the existence of a credit channel.
We find some evidence of a more aggressive behaviour of small banks on loan granting activity, which
tends to amplify slightly in periods of lower policy rates. There seems to be a slight increase in loan
supply by small banks to all borrowers following an expansion in monetary policy, which is consistent
with evidence obtained by Jiménez et al. (2008) or Buch et al. (2011).
While the discrete choice models suggest an increase in the ex-ante risk taken by banks in their loan
activity when policy rates are lower, the survival models do not confirm this increase in risk-taking ex-post,
i.e., over the life of the loan. When bank, borrower and loan characteristics are fixed at the moment of
the loan concession, lower policy interest rates decrease the hazard rate of the loans. The only exception
to this result occurs when we consider time-varying covariates. However, these latter results are not
sufficient to support the existence of a risk-taking channel, as the banks’ decisions when granting loans
could not perfectly foresee the future evolution of firm, bank and macro conditions.
In sum, we find consistent evidence that in periods of lower policy interest rates banks are more likely
to grant loans to borrowers with worse credit quality (namely borrowers with recent defaults or without
credit history). However, despite this increased risk-taking, the entire portfolio of loans granted during
such periods does not show higher default probabilities through time. As such, our results do not support
supportive of a credit channel.
References
Acharya, V. and H. Naqvi (2012), “The seeds of a crisis: A theory of bank liquidity and risk-taking over the
business cycle”, Journal of Financial Economics, 106(2), 349-366.
Adrian, T. and H.S. Shin (2010), “Liquidity and leverage”, Journal of Financial Intermediation, 19(3),
418-437.
Adrian, T. and H.S. Shin (2008), “Financial intermediaries, financial stability and monetary policy”, Federal Reserve Bank of New York Staff Reports, nº 346.
Altunbas, Y., L. Gambacorta and D. Marquez-Ibanez (2010), “Does monetary policy affect bank risktaking?”, BIS Working Paper No 298.
Angeloni, I., E. Faia and M. Lo Duca (2010), “Monetary policy and risk-taking”, Bruegel Working paper
2010/00.
Bernanke, B. and A. Blinder (1988), “Credit, Money, and Aggregate Demand,” American Economic
Review 78 (May), 435—439.
Bernanke, B. and M. Gertler (1989), “Agency Costs, Net Worth, and Business Fluctuations”, American
Economic Review, vol. 79(1), pages 14-31, March.
Bernanke, B. and M. Gertler (1995), “Inside the Black Box: The Credit Channel of Monetary Policy Transmission”, Journal of Economic Perspectives, vol. 9(4), pages 27-48, Fall.
Bonfim, D. (2009), “Credit risk drivers: evaluating the contribution of firm level information and macroeconomic dynamics”, Journal of Banking and Finance, 33(2), 281-299.
Bonfim, D., D.A. Dias and C. Richmond (2012), “What happens after default? Stylized facts on access to
credit”, Journal of Banking and Finance, 36(7), 2007-2025.
Borio, C. and H. Zhu (2012), “Capital regulation, risk-taking and monetary policy: a missing link in the
transmission mechanism”, Journal of Financial Stability, 8, 236-251.
Brissimis, S.N., and M.D. Delis (2010), “Bank heterogeneity and monetary policy transmission”, ECB
Working Paper nº 1233.
Brunnermeier, M.K. (2001), “Asset Pricing under Asymmetric Information-Bubbles, Crashes, Technical
Analysis and Herding”, Oxford, Oxford University Press.
Buch, C., S. Eickmeier and E. Prieto (2011), “In search for yield? Survey-based evidence on bank risktaking”, Deutsche Bundesbank Discussion Paper No 10/2011.
De Nicolò, G., G. Dell’Ariccia, L. Laeven and F. Valencia (2010), “Monetary policy and bank risk-taking”,
IMF Staff Position Note 10/09.
Delis, M.D. and G.P. Kouretas (2011), “Interest rates and bank risk-taking”, Journal of Banking and
Finance, 35, 840-855.
Dell’Ariccia, G.D., L. Laeven and R. Marquez (2011), “Monetary policy, leverage and bank risk-taking”,
CEPR Discussion Paper nº 8199.
103
Articles
the existence of a fully-fledged risk-taking channel in Portugal. Instead, they seem to be generally more
Diamond, D. and P. Dybvig (1983), “Bank runs, deposit insurance, and liquidity”, Journal of Political
Economy, 91(3), 401-419.
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
104
Diamond, D.W. and R.G. Rajan (2012), “Illiquid banks, financial stability and interest rate policy”, Journal
of Political Economy, 120(3), 552-591.
Disyatat, P. (2011), “The bank lending channel revisited”, Journal of Money, Credit and Banking, 43(4),
711-734.
Farinha, L. and C.R. Marques (2003), “The bank lending channel of monetary policy: Identification and
estimation using Portuguese micro bank data”, in Monetary Policy Transmission in the Euro Area, I
Angeloni, A Kashyap and B Mojon (eds), Cambridge University Press, Part 3, Chap. 22, pp 359-371.
Freixas, X. and J. C. Rochet (2008), “Microeconomics of banking”, MIT Press, second edition.
Gaggl, P. And M.T. Valderrama (2011), “Do banks take more risk in extended periods of expansive monetary policy? Evidence from a natural experiment”, unpublished.
Gambacorta, L. (2009), “Monetary policy and the risk-taking channel”, BIS Quarterly Review, December
2009.
Gameiro, I.M., C. Soares and J. Sousa (2011), “Monetary policy and financial stability: Na open debate”,
Issue for discussion, Economic Bulletin – Spring 2011, Banco de Portugal.
Geršl, A., P. Jakubík, D. Kowalczyk, S. Ongena and J.L. Peydró (2012), “Monetary Conditions and Banks´
Behaviour in the Czech Republic”, Czech National Bank Working Paper 2 2012.
Ioannidou, V., S. Ongena and J.L. Peydró (2009), “Monetary policy, risk-taking and pricing: Evidence
from a quasi-natural experiment”, unpublished.
Jiménez, G., S. Ongena, J.L. Peydró and J. Saurina (2008), “Hazardous times for monetary policy: What
do twenty-three million bank loans say about the effects of monetary policy on credit risk-taking”,
Banco de España working paper nº 0833.
Maddaloni, A. and J.L. Peydró (2011), “Bank risk-taking, securitization, supervision and low interest
rates – Evidence from the Euro area and the US lending standards”, Review of Financial Studies,
24, 2121-2165.
Paligorova, T. And J. A. C. Santos (2012), “When is it less costly for risky firms to borrow? Evidence from
the bank risk-taking channel of monetary policy”, Bank of Canada Working Paper 2012-10.
Rajan, R.G. (2006), “Has Financial Development Made the World Riskier?”, European Financial Management, 12(4), 499-533.
Thakor, A. (2013), “Incentives to innovate and financial crises”, Journal of Financial Economics, 103(1),
130-148.
Valencia, F. (2011), “Monetary policy, bank leverage and financial stability”, IMF working paper 11/244.
INVESTMENT DECISIONS AND FINANCIAL STANDING OF
PORTUGUESE FIRMS – RECENT EVIDENCE*
105
Luisa Farinha** | Pedro Prego**
Articles
Abstract
The analysis of firms’ investment decisions and the firm’s financial standing is
particularly relevant under a scenario of (i) the high indebtedness levels of Portuguese
firms, (ii) the reduction in profitability of these firms, which reduces the amount of
internally available funds thus increasing the demand for external financing, and (iii)
the ongoing Financial and Economic Crisis that considerably changed the conditions
and access to the credit markets. In this article, yearly balance sheet and financial
statements data from the Central Balance Sheet since 2006 until 2011 is used. The
results obtained indicate that firms’ financial standing is indeed relevant in explaining
corporate investment decisions, where the burden of servicing debt, the cost of capital,
and the firm’s indebtedness all have a negative relationship with firm’s investment rate.
As for profitability the results suggest a strongly and positive relationship with firms’
investment rate. Nonetheless, these results are predominantly seen for smaller firms
where large firms investment rate only seem to be affected by the profitability levels.
Moreover, there is evidence suggesting that the impact of firms’ financial standing
became more relevant during the period of the sovereign debt crisis in the euro area.
1. Introduction
The analysis of firms´ investment decisions is particularly relevant when assessing and projecting economic
activity. In the context of financial frictions that can significantly affect firms´ demand of productive factors
and hence future economic output capacity, the financial accelerator literature states that corporate
investment is highly volatile and strongly concentrated in certain periods followed by sharp declines
(Bernanke et al., 1999). In this context, it is argued that the presence of financial frictions exacerbates
business cycles. Therefore, examining the relationship between firms’ financial health and their investment decisions is an important matter.
Despite the relevance of this topic, only a few papers in the empirical literature use data for the Portuguese economy. Farinha (1995) used a firm level dataset ranging from 1986 to 1992 and concluded that
the availability of internally generated funds affects investment decisions of firms, except in the case of
the largest firms. Barbosa et al. (2007) focused on the period from 1995-2005 and found a negative
relation between several measures of firms financial pressure and their investment flows. Oliveira and
Fortunato (2006) used balance sheet data from 1990 to 2001 and found that smaller and younger firms
have higher growth-cash flow sensitivities than larger and more mature firms.
*
The authors would like to acknowledge the valuable comments and suggestions of Pedro Portugal as well as
all the support provided by Lucena Vieira. The opinions expressed in the article are those of the authors and
do not necessarily coincide with those of Banco de Portugal or the Eurosystem. Any errors and omissions are
the sole responsibility of the authors.
** Banco de Portugal, Economics and Research Department.
The present economic and financial crisis is leading to an unprecedented drop in investment of Portuguese corporations. The collapse of investment has been determined by the drop in demand and the
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
106
high uncertainty concerning its recovery. Moreover, the narrowing of firms’ internal financing capacity
has been going together with tight credit supply. In such a context is likely that firms’ financial position
is playing a more relevant role in explaining investment rates at the firm level. The aim of this article is
to give further insight into the relationship between private sector investment decisions and its financial
standing. More specifically, in the line of Barbosa et al. (2007) this paper analyses how firm’s financial
position (proxied by Indebtedness level, Interest Burden, Financing Cost, and Profitability) relates to and
potentially limits corporate investment decisions.
Using a very comprehensive dataset composed of all non-financial private sector firms’ balance sheet
and income statements covering the period from 2006 to 2011, we are able to present a more detailed
analysis of this relationship than previous studies. Furthermore, the study also focuses on two additional
aspects concerning this relationship. On the one hand, we test if the effect of financial conditions changes
according to the size of the firm. On the other hand, we test if the relationship between financial conditions and corporate investment changed during this period.
The rest of the article is structured as follows. Section 2 reviews some of the existing literature on firm’s
investment decisions and financial factors. Section 3 describes the data used in the article and provides
a descriptive analysis on the relationship between investment rates and firm’s financial position. Section
4 presents the baseline specification and estimation method as well as the main results, closing with
some robustness checks. Section 5 concludes by summarizing the main results and presenting some
lines for future research.
2. Literature Review
The neoclassical theory of capital accumulation established that under perfect capital markets the firm’s
capital structure is irrelevant (Modigliani and Miller, 1958). With that regard, several theories have been
proposed, a common stepping stone being the existence of asymmetric information and non-frictionless
capital markets as the main factors explaining how firms may be rationed out of credit markets and how
different origins of capital have different costs for firms and thus different attractiveness.
The work of Stiglitz and Weiss (1981) and Myers and Majluf (1984) are references when studying the
problem of asymmetric information. Stiglitz and Weiss (1981) developed a model of credit rationing
where asymmetry of information creates a problem of adverse selection. This happens because when
interest rates rise, and given that lenders cannot differentiate borrowers’ quality (or the probability of
a borrower to repay its loan), the borrowers to initially leave the market for credit are the ones with
the highest quality (or more likely to re-pay). This process is then reinforced where a larger proportion
of “bad” borrowers remain in the market for credit thus reducing the expected payoff for the lenders,
making these raise interest rates in order to limit the supply of credit.
In the same line, Myers and Majluf (1984) show that relying on internally generated capital to fund
investment opportunities is preferred to all other options of external finance given that managers have
complete information not only on the firm’s current state but also on future investments under consideration. Moreover, Myers and Majluf (1984) show that if firms do have to raise capital externally, when
running out of internally generated funds, doing it through debt rather than through equity is optimal
because debt is information insensitive. These arguments have come to be known as the “pecking order”
theory stating that internally generated funds are the most preferred (cheaper) way of raising capital,
followed by debt and only as a last resort do firms issue new equity.
Different branches have been developed in the domain of imperfections in credit markets. The balance
sheet channel hypothesis developed by Kiyotaki and Moore (1997; 2002) states that firms might become
financially constrained, even in cases where they were not directly exposed to a negative shock, given
reductions in assets prices in other sectors that were being used as collateral, which leads to a higher
cost of financing thus increasing the likelihood of default. Yet, Kiyotaki and Moore (1997; 2002) also
mention that the aforementioned process is only valid when occurring among credit constrained firms.
In that sense, the chain effect (or the balance sheet channel) stops as soon as one of the firms in the
credit chain is not credit constrained, limiting the contagion effect among credit constrained firms.1
the quantity level rather than from prices [in the line of Stiglitz and Weiss (1981)]. The model in Holmstrom
and Tirole (1997) works through a reduction in the supply of available resources to the credit market
(be it a credit crunch, a collateral squeeze or a savings squeeze), showing that poorly capitalized firms
suffer disproportionally. A particular feature of this model is that credit rationing might happen at the
same time as the price of credit decreases.2
From the seminal work of Fazzari et al. (1988) a vast empirical literature followed the work of Stiglitz
and Weiss (1981) and Myers and Majluf (1984). In this respect, Fazzari et al. (1988) used an approach
based on the Tobin’s Q model of Investment that has the advantage of capturing the market’s valuation
of the firm’s investment opportunities.3
In addition, the authors followed an approach where they used firm’s cash flow information to proxy
the existence of internally generated funds, showing that firm’s investment decisions are significantly
affected by internally generated funds. Nonetheless, some authors pointed that including a measure for
cash flow in an investment equation cannot, by itself, be a proof of financing constraints. In particular,
Kaplan and Zingales (1997) argued that the results previously found are highly dependent on the definition of financially constrained firms and the higher sensitivity found may be a result of precautionary
savings or risk-aversion. In that regard, Kaplan and Zingales (1997) classify firms as unconstrained,
possibly constrained and constrained and found that the latter are less sensitive to cash-flow. More
recently, Alti (2003) showed that even in the case of frictionless credit markets investment decisions are
sensitive to cash-flow because, due to uncertainties around investments’ outcomes, firms make future
investment dependent on cash-flows realizations making investment highly sensitive to cash-flow realizations. Nonetheless, Alti (2003) showed that, even after performing a correction for firm’s investment
opportunities (proxied by the Q value), firms do seem to be sensitive to cash-flows. In the same spirit,
Oliveira and Fortunato (2006), using a data set of Portuguese manufacturing firms from 1990 to 2001,
argued that the higher sensitivity to cash-flows found for smaller and younger firms might in part be
due to the fact that cash-flows realization is particularly important for these firms, and not necessarily
indicating the existence of financing constraints.
Recent studies, such as Martínez-Carrascal and Ferrando (2008), have followed an approach of using
several “financial pressure” indicators where they showed that investment decisions are significantly
constrained by firm’s financial position (proxied by Indebtedness level, Interest Burden and Profitability).
Additionally, Nickell and Nicolitsas (1999) studied how firm’s financial pressure (proxied by the firm’s
Interest Burden) related to firm level employment finding a negative relationship between the two.
Likewise, Benito and Hernando (2007) found not only a negative relationship between financial pressure
and employment but also to inventories and dividend policies. In addition, Benito and Hernando (2007)
presented evidence of the existence of non-linearities in the relationship between financial position
1 In an empirical analysis, Martínez-Carrascal and Ferrando (2008) argued that monetary policy will have important consequences not only through the common interest rate channel but also through this balance sheet
channel, given that higher interest rates reduce discounted cash flows and collateral values. Thus, the balance
sheet channel might be especially relevant given the deleveraging process that Portugal (and other highly credit
constrained countries in the euro area) is currently undertaking.
2 Ivashina and Scharfstein (2010) gave recent empirical support to the case of a “credit crunch” for the US economy after the failure of Lehman Brothers in 2008 where banks more exposed to credit-line drawbacks cut
lending to a greater extent.
3 In this article, the Tobin’s Q model approach was not used because it is rather difficult to apply this methodology
for an economy as Portugal given the very small number of listed companies, thus reducing the possibility of
having a good measure of Tobin’s Q.
107
Articles
Holmstrom and Tirole (1997) developed a model of “credit crunch” where the market for credit clears at
and investment decisions, a result that is also referred in the work of Martínez-Carrascal and Ferrando
(2008) and Hernando and Martínez-Carrascal (2008). In this respect, Marchica and Mura (2010) found
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
108
that financial flexibility, defined as firms with debt levels ‘permanently’ below what would be expected
ex-ante, allowed firms to take advantage of unexpected investment opportunities. In addition, they also
found that these “financially flexible” firms invested more heavily and with higher levels of profitability
than firms that lacked such flexibility, which might had to pass on profitable investment opportunities.
3. Data and Descriptive Analysis
The analysis in this article uses Banco de Portugal annual Central Balance Sheet database, which is based
on Informação Empresarial Simplificada (IES). IES collects balance sheet and financial statements from
virtually all Portuguese corporate firms in the Portuguese Economy, both private and state owned firms
(the latter having been removed from the sample) since 2006 until 2011 (which is the most recent data
available).
One of the main benefits of using IES is the ability to focus the analysis at the micro level. In this article
only private non-financial indebted firms (from now on it will only be referred to as firms for simplicity)
were considered. Morevoer, observations that did not have strictly positive values of financial debt and
interest paid were removed from the database, as well as self-employed individuals. It was also necessary
to eliminate from the database firms that reported incomplete or incoherent data, such as observations
with negative total assets or negative business turnover. Furthermore, firms that did not appear in the
dataset for a minimum of three consecutive years were also excluded from the analysis. For the purpose
of the econometric analysis only firms with positive gross operating income were considered.4
Especial attention had also to be paid to extreme firm assets variations. In that sense, firms that had
an increase in fixed assets of more than 500% or a decrease bigger than 75% were removed from
the sample (which corresponded to around 5% of initial sample observations). Finally, in order to deal
with spurious outlier observations, observations below (and above) the 1st (and 99th) percentile of the
relevant variables were winsorised. After applying all the above criteria, the data used in this study is
an unbalanced panel of 189.266 observations, corresponding to 97.761 firms observed in the period
between 2006 and 2011.
The objective of the study is to analyze firms’ investment decisions focusing particularly on the effect
of several factors that are related with firms’ financial pressure. The investment rate is defined as the
amount of Total Investment made by the firm in a given year divided by the Stock of Capital in the year
immediately before. Both Total Investment and Total Stock of Capital include tangible and intangible
assets. A yearly fixed depreciation rate of 10% was assumed. In line with previous empirical studies, the
firm-level proxies for “financial pressure” were the following: a measure for the Interest Burden defined
as the ratio between Paid Interest and Gross Operating Income; firm’s Indebtedness level defined as
Financial Debt to Total Assets; the Financing Cost defined as the ratio between Paid Interest and Financial
Debt; firm’s Profitability defined as Gross Operating Income to Total Assets.
The charts below present the simple percentiles of order 10, 25, 50, 75 and 90 of the distribution of the
variables of interest for the different years under analysis. This approach makes it possible to evaluate
the evolution of the typical firm (i.e. the median firm every year), as well as the evolution of firms in
several points of the distributions.
Chart 1 shows how Portuguese firms in general reduced their investment ratio during the period under
analysis, except in 2010. Additionally, the data shows that firms that were located in the top percentiles
4 This condition is necessary in order to preserve the monotonicity of the relation between the interest burden
ratio and firms’ financial standing. In fact, the interest burden resulting from a negative operative income with
a large absolute value is lower than the interest burden resulting from a negative operating income with a small
absolute value.
decreased their investment rate more strongly, squeezing the right hand side of the distribution. With
respect to the median firm in each size category there are not considerable differences (see Chart 2).
These results are in line with the ones obtained in the previous study made by Barbosa et al. (2007)
where a reduction in the investment rate had been already occurring since 1999.
significantly in 2010 and decreased again in 2011 (yet, still staying above the level observed in 2009).
Indebtedness levels for Medium and Large sized firms were considerably higher than those for Micro
and Small sized firms from 2006 until 2009 (Chart 4). In 2010, indebtedness levels of Micro and Small
firms increased considerably, reaching the same indebtedness level of Medium sized firms, decreasing
slightly in 2011 for the former. On the other hand, indebtedness levels for Large firms decreased in 2010
and remained constant in 2011.
With respect to the evolution of firms’ Financing Cost during the period under analysis two different
patterns emerged, with a downward trend from 2006 until 2010, being especially significant since
2008, followed by a slight increase in 2011. There was a significant compression of the distribution in
2010, which occurred mostly from the right-hand side of the distribution (see Chart 5). From Chart 6 it
is interesting to note that the pattern described seems to be the same across firm sizes.
Following the same pattern of the Financing Cost indicator, there is a significant reduction in the dispersion
of the Interest Burden distribution that occurred mostly in 2010 (see Chart 7). It is also worth pointing to
the fact that Small and Medium sized firms present a similar level for the median interest burden across
the time span of this study, which stood above the level recorded for Micro and Large Firms (see Chart 8).
Chart 9 shows that the overall profitability of firms has been slightly declining over the entire period
under analysis. This result is particularly relevant given that, as documented by Barbosa et al. (2007),
the profitability of the representative firm in Portugal has already been in decline since 1995. Our results
also show how the dispersion of the distribution increased slightly for the left-hand size, suggesting that
less profitable firms were particularly affected. When looking at the different size categories (Chart 10),
the profitability level of the representative firm for each firm-size category increases slightly, with the
exception of the Micro sized firm category.5
Chart 1
Chart 2
EVOLUTION OF INVESTMENT RATE
EVOLUTION OF MEDIAN INVESTMENT RATE BY
FIRM SIZE
160
25
140
20
Per cent
Per cent
120
100
80
60
15
10
40
5
20
0
2007
2008
2009
2010
Percentile 10 Percentile 25 Percentile 50
Percentile 75 Percentile 90
Source: Banco de Portugal (Central Balance Sheet).
2011
0
2007
2008
Micro
2009
Small
Medium
2010
2011
Large
Source: Banco de Portugal (Central Balance Sheet).
5 The results from Figure 9 and Figure 10 might seem contradictory given that the median Profitability for the
overall distribution is slightly declining and when splitting according to size the results show an upward Profitability level with the exception of Micro sized firms. The reason for this difference relies upon the fact that the
distribution in Figure 9 is dominated by Micro sized firms (which represent 83% of the sample).
109
Articles
Indebtedness of Portuguese firms (Chart 3) increased slightly (but steadily) from 2006 until 2009, increased
Chart 3
Chart 4
EVOLUTION OF INDEBTEDNESS LEVEL
EVOLUTION OF MEDIAN INDEBTEDNESS LEVEL
BY FIRM SIZE
II
80
30
70
20
50
Per cent
Per cent
60
40
30
10
20
10
0
2006
2007
Percentile 10
Percentile 75
2008
2009
Percentile 25
Percentile 90
2010
2011
0
2006
Percentile 50
2007
Micro
Source: Banco de Portugal (Central Balance Sheet).
2008
Small
2009
Medium
2010
2011
Large
Source: Banco de Portugal (Central Balance Sheet).
In order to assess the relationship between the investment rate and several indicators of the firms’ financial situation, it is useful to perform a bivariate analysis, which provides an initial understanding of the
relationship between the investment rate and the main variables of interest. The chosen variables are
those presented before, that is, indebtedness, the financing cost, the interest burden and profitability.
Charts 11, 12, 13 and 14 compare the median investment rate in different corporate groupings defined
on the basis of the variables that are expected to influence the investment rate. Each chart presents the
median cash holding ratio for firms with high, medium and low levels, respectively, of that indicator. The
median decile (which includes firms between percentiles 45 and 55) can be regarded as representative of
the behaviour of the typical firm of that size in terms of the respective variable, while the top (bottom)
decile includes the 10% of firms with the highest (lowest) value of the variable. Moreover, Charts 15, 16,
Chart 5
Chart 6
EVOLUTION OF FINANCING COST
EVOLUTION OF MEDIAN FINANCING COST BY
FIRM SIZE
35
10
30
8
Per cent
25
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
110
20
15
6
4
10
2
5
0
2006
2007
2008
2009
2010
Percentile 10 Percentile 25 Percentile 50
Percentile 75 Percentile 90
Source: Banco de Portugal (Central Balance Sheet).
2011
0
2006
2007
Micro
2008
Small
2009
Medium
2010
Large
Source: Banco de Portugal (Central Balance Sheet).
2011
Chart 7
Chart 8
EVOLUTION OF INTEREST BURDEN
EVOLUTION OF MEDIAN INTEREST BURDEN BY
FIRM SIZE
250
40
111
Articles
200
Per cent
Per cent
30
150
20
100
10
50
0
2006
2007
2008
2009
2010
Percentile 10 Percentile 25 Percentile 50
Percentile 75 Percentile 90
0
2011
2006
2007
Micro
Source: Banco de Portugal (Central Balance Sheet).
2008
Small
2009
Medium
2010
2011
Large
Source: Banco de Portugal (Central Balance Sheet).
17and 18 (in Appendix – Chart-Analysis part]) depict the median investment rate for firms belonging to the
top, median and bottom deciles of the indicators for micro, small, medium and large firms, respectively.
Chart 11 shows that firms with high levels of indebtedness show substantially lower investment rates
than firms with intermediate or low levels of this variable. This pattern is broadly seen in all size groups
(Chart 15). One interesting feature is the fact that, for the group of Large Firms with a lower indebtedness
level (and less for the group of firms in the mid decile), the investment rate in 2011 increased moderately, which might indicate a decision for these less-constrained firms to take advantage of potential
investments opportunities.
Chart 9
Chart 10
EVOLUTION OF PROFITABILITY
EVOLUTION OF MEDIAN PROFITABILITY BY FIRM
SIZE
30
5
20
4
0
2006
2007
2008
2009
2010
-10
2011
Per cent
Per cent
10
3
2
-20
1
-30
-40
0
Percentile 10 Percentile 25 Percentile 50
Percentile 75 Percentile 90
Source: Banco de Portugal (Central Balance Sheet).
2006
2007
Micro
2008
2009
Small
Medium
2010
Large
Source: Banco de Portugal (Central Balance Sheet).
2011
When looking at the relation between investment and the Financing Cost variable (Chart 12), firms in
the last decile of the Financing Cost distribution present an investment rate that is higher but close to
that under a bivariate analysis there are factors affecting investment that are not controlled. For instance,
small and young firms are usually assumed to be riskier thus facing a higher financing cost but still having
a greater growth potential. This may explain the higher investment demand. This pattern is also generally
observed for firms in the different size categories (Chart 16).
As for the Interest Burden indicator (Chart 13) a clearer cut relationship is obtained. In fact a higher
Interest Burden ratio might indicate a more financially pressured firm. In that sense, the results show
that firms with a higher Interest Burden present the smaller rates of investment. The same pattern is
generally seen for the different size categories (Chart 17).
With respect to the relationship between investment and the Profitability indicator (Chart 14) the results
show two different patterns. On the one hand, firms in the upper decile of Profitability present higher
Chart 11
Chart 12
EVOLUTION OF INVESTMENT RATE ACCORDING
TO DIFFERENT INDEBTEDNESS LEVELS
EVOLUTION OF INVESTMENT RATE ACCORDING
TO DIFFERENT FINANCING COST LEVELS
20
20
16
16
Per cent
Per cent
12
12
8
8
4
4
0
2007
2008
Low Indebtedness
2009
2010
Medium Indebtedness
0
2011
2007
2008
Low Financing Cost
High Indebtedness
2009
Medium Financing Cost
2010
2011
High Financing Cost
Source: Banco de Portugal (Central Balance Sheet).
Source: Banco de Portugal (Central Balance Sheet).
Chart 13
Chart 14
EVOLUTION OF INVESTMENT RATE ACCORDING
TO DIFFERENT INTEREST BURDEN LEVELS
EVOLUTION OF INVESTMENT RATE ACCORDING
TO DIFFERENT PROFITABILITY LEVELS
30
50
25
40
20
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
112
the ones with a median Financing Cost. This apparently unexpected result could be related to the fact
Per cent
II
15
20
10
10
5
0
30
2007
2008
2009
2010
2011
Low Interest Burden Medium Interest Burden High Interest Burden
Source: Banco de Portugal (Central Balance Sheet).
0
2007
2008
Low Profitability
2009
Medium Profitability
2010
2011
High Profitability
Source: Banco de Portugal (Central Balance Sheet).
investment rates. On the other hand, and somehow unexpected as firms with low profitability levels (in
the first decile) have a higher investment rate than firms with median profitability. However, this pattern
seems to break down when the analysis is performed according to firm sizes (Chart 18) with the exception
being the Large Firms group in 2010, where firms located in mid-decile of the Profitability distribution
present a lower investment rate than firms located in the lower profitability decile.
Articles
113
4. Econometric Analysis and Main Results
4.1 Methodology and model specification
Previous section presents a simple bivariate analysis of firms’ investment rate according to the various
measures of their financial standing. These measures are expected to be correlated with each other, as
well as with other firms’ characteristics what makes the interpretation of the results based on that analysis
particularly difficult. In order to overcome this limitation, this section presents the results of multivariate
regressions for firms’ investment rate. The objective of the econometric analysis is not to obtain a causal
relationship between firms’ financial standing and investment but merely to test the sign and significance
of the correlation between them. The analysis is based on the estimation of the following equation:
Invi ,t   i   Invi ,t 1   X i ,t 1   Salesi ,t 1   j Size j ,t 1  t  Si   i ,t
(1)
where i indexes firms i  1,2, , N and t indexes year t  1,2, , T and j indexes each firm size
category where j  1,2, 3, 4 corresponding to Micro, Small, Medium and Large Firms (with the latter
being the omitted category) respectively. As for the variables used,
Invi,t
refers to the firm’s i investment
rate at time t , Xi,t 1 represents the vector of financial variables of interest (interest burden, indebtedness,
financing cost, and profitability), Sales are the log of real sales or business turnover, t are time effects
controlling for macroeconomic influences, Si are fixed industry effects, and finally i,t is the error term.
With the exception of the investment rate variable, all continuous variables are presented as logarithms
and as deviations from the sample mean in order to facilitate the reading of the results.6
Besides the above specification and with the aim of evaluating how the impact of the financial variables
of interest change according to firm size, equation 1 was re-estimated using instead the financial variable
under review interacted with the four firm-size categories. In addition, a test was made to a differentiated
effect for the period before and after crisis (defined as a dummy variable equal to 1 for 2010 and 2011).
The estimation method consists of using the GMM-System estimator proposed by Arellano and Bover
(1995) and further developed by Blundell and Bond (1998). These models are particularly indicated when
the number of years is small and the number of firms is large; there are fixed individual effects, autoregressive variables that show high persistence; and independent variables that are not strictly exogenous.
By using this methodology, equations are estimated in levels and in differences and the instruments for
these are the lagged values of the non-strictly exogenous variables. For all estimations, the Hansen test
(at the conventional 5% level) for the validity of instruments and the Arellano and Bond (1991) test of
non-existence of first and second order serial correlation in the first-differenced residuals (labeled AR 1
and AR 2) are presented.
6
The specification in question corresponds to considering for the interest burden (B), for instance, the variable
where ln(B ) is the sample mean.
 
bi,t  ln Bi,t  ln(B )
4.2 Main results
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
114
The baseline scenario
The main results are presented in Table 1 and are structured as follows: Column 1 presents the results
of estimating the base model that does not include any of the financial variables; Columns 2 to 5 report
the estimates of the basic specification augmented with one financial variable at a time. As referred in
section 3, these results were obtained with a large sample of private, indebted non-financial corporations. Moreover, in order to preserve the monotonicity of the relation between the interest burden ratio
and firms’ financial standing only observations with a positive gross operating income were considered.7
The results for the base model suggest that there is persistency in company level investment as the estimated coefficient for the 1-period lagged investment rate assumes a positive and statistically significant
value (in line with the work of Barbosa et al., 2007). As expected, the effect of Sales, which proxies firms’
growth potential, is positive and significant. In line with the prior that larger firms are more mature in
their respective life cycle, which reduces the need to make significant investments in capital accumulation,
the magnitude and significance of the size dummies in column 1 show that the investment rate of small
and micro firms is significantly higher than the investment rate of large firms (the omitted category).
The coefficients associated to the year dummies show that the investment rates in 2009 and particularly
in 2011 were lower than in 2008 (the omitted year) but were higher in 2010. The above results are
globally confirmed by the models that include firms’ financial variables with a few exceptions such as
the negative and statistically coefficient (yet only at the 10% level) for the Sales variable in column 3
and the negative but not statistically significant coefficient for the 1-period lag investment in column 4.
Regarding the effect of firms’ financial variables, which is the main focus of the analysis in this article,
the results show that these effects have the expected sign and are statistically significant at the usual
levels of significance as found in related work (Martínez-Carrascal and Ferrando, 2008; Bond et al.,
2003; Benito and Hernando, 2007; Hernando and Martínez-Carrascal, 2008; Barbosa et al., 2007). More
specifically, the model in column 2 includes the variable measuring firm’s indebtedness level, which shows
a negative and significant relationship with the investment rate suggesting that higher levels of debt
might restrain firms from future investments. In column 3, Profitability also has the expected positive
and significant coefficient indicating that firms with high levels of profitability tend to invest more in the
subsequent year. With respect to the Interest Burden variable (column 4) the results showed a negative
and highly significant coefficient suggesting that this measure appears to be relevant in the analysis of
the Portuguese corporate investment. A similar result was found for the impact of firms’ Financing Cost
(column 5) that showed a negative and highly significant coefficient indicating that an increase in the
cost of capital makes it less likely for future investments to become worthwhile pursuing. In general,
these results support the hypothesis that financial pressure faced by firms is relevant to explain corporate
investment decisions, as the interest burden, indebtedness, financing cost and profitability are found to
be significant when included in investment equations.8
7 The models in columns 1, 2, 3 and 5 were also estimated without imposing this last restriction. The results of
these estimations do not change the conclusions.
8 As in Barbosa et al. (2007) and Benito and Hernando (2007) a specification that included three of the financial
variables under analysis (Indebtedness, Financing Cost and Profitability) simultaneously was also tested. Under
this specification, only Financing Cost and Profitability are statistically significant and have the expected negative
and positive coefficient respectively. On the other hand, the Indebtedness variable had a positive coefficient yet
not statistically significant. This result is similar to the ones found in Barbosa et al. (2007) where the signs of the
coefficients were the same but only Profitability and Indebtedness were statistically significant.
Table 1
ECONOMETRIC RESULTS FOR THE INVESTMENT RATE (INVI,T)
Inv i,t-1
Sales i,t-1
Micro firms i,t-1
Small firms i,t-1
Medium firmsi,t-1
Year 2009
Year 2010
Year 2011
Indebtedness
Profitability
0.0145***
0.0135***
0.0138***
-0.00971
0.00735*
(4.717)
(3.904)
(4.016)
(-1.263)
(1.789)
0.00244*
0.00264*
-0.00243*
-0.00119
0.00384***
(1.779)
(1.919)
(-1.771)
(-0.683)
(2.655)
0.0918***
0.0936***
0.0843***
0.116***
0.104***
(7.966)
(8.073)
(7.336)
(7.982)
(8.550)
0.0403***
0.0445***
0.0450***
0.0844***
0.0518***
(3.709)
(4.022)
(4.154)
(4.728)
(4.520)
0.000262
0.00766
0.0102
0.0449**
0.00310
(0.0228)
(0.642)
(0.891)
(2.478)
(0.268)
-0.0451***
-0.0407***
-0.0456***
-0.0366***
-0.0373***
(-14.14)
(-12.45)
(-13.59)
(-7.216)
(-11.01)
0.0533***
0.0572***
0.0515***
0.0357***
0.0385***
(15.31)
(15.27)
(14.16)
(7.432)
(6.446)
-0.0760***
-0.0540***
-0.100***
-0.163***
-0.132***
(-24.89)
(-6.954)
(-31.76)
(-6.821)
(-7.006)
-0.0295**
0.0460***
-0.0688***
-0.0410***
(-2.464)
(42.29)
(-3.186)
(-3.238)
Xi,t-1
Interest Burden
Financing Cost
Observations
188,852
188,852
188,852
188,852
188,852
No of firms
97,499
97,499
97,499
97,499
97,499
Hansen (p-value)
0.152
0.220
0.453
0.380
0.117
AR 1 (p-value)
0
0
0
0
0
AR 2 (p-value)
0.881
0.800
0.771
0.983
0.970
Source: Banco de Portugal (Central Balance Sheet).
Notes: Estimation by GMM system estimator, using the routine xtabond2, developed by Roodman (2005). The variable Xi,t-1 corresponds to the financial variables under analysis presented on top of each column. Lags 1 to 4 in levels of investment rate were used
for columns 1, 2, 3 and 5 while in column 4 it was only used lags 2 to 4. Lag 2 of investment rate was also used as first-differencing
instrument in column 1 and 4. Were used as instrument in column 2 the indebtedness lag 2 and 3 in levels; in column 4 the interest
burden lag 3 and 4 in level; in column 5 the financing cost lag 4 in first-differencing; and Profitability was taken as strictly exogenous
in column 3.The variables Sales, the Size indicator dummies, as well as the year and sectoral dummies were used as regular instrument, even though Sales was used only in levels. T-statistics in parentheses. *** p-value<0.01, ** p-value<0.05, * p-value<0.1.
The effect of firm size
Financial frictions in the access to credit motivated mainly by asymmetries of information are expected to
affect smaller firms more significantly. Larger firms will be more capable of reducing information asymmetries vis-à-vis potential lenders as they are able to better report their financial conditions. Furthermore,
larger firms also have easier access to alternative financing sources than smaller firms, such as equity
markets. Hence, it is expected that larger firms will be less affected by their own financial conditions and
thus less financially constrained then smaller firms. Accordingly, the base model presented in Equation
1, was re-estimated considering that the coefficient of the financial variable under review is different
115
Articles
Base Model
for each of the four categories of firm size9, which corresponds to estimate the following specification:
Invi,t   i   Invi ,t 1   k X i ,t 1Sizekj ,t 1   Salesi ,t 1   j Size j ,t 1  t  Si   i ,t
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
116
(2)
where k  1,, , 4 indexes the four firm-sizes categories and where the financial variable under review
is interacted with each of the firm size dummies (ranging from Micro to Large Firms) and the results are
reported in Table 2.
In this specification, the results show that the 1-period lagged investment rate variable is positive and
highly significant when using the Indebtedness and Profitability financial variables, which reinforces the
previous result for the existence of persistence in this variable. Nevertheless, it is no longer statistically
significant in column 4 (Financing Cost) and is negative and statistically significant in column 3 (Interest
Burden).10
Overall, the signs and statistical significance of the coefficients associated with the financial variables are
in line with the ones previously obtained. As expected in the case of Large Firms, the effect of some of
the financial variables are non-significant, which might indicate that these firms are less severely affected
by their financial conditions. More specifically, this is the case for the Interest Burden, which is negative
and significant for all firm size categories with the exception of large firms. In addition, the magnitude
of the coefficient increases for smaller firms, which supports the suggestion that smaller firms are particularly affected by a high ratio of the cost of debt to income. In the same line, the interaction between
Financing Cost and the Large Firm size indicator is also not statistically significant. On the other hand, and
given the proximity of the size of the coefficients and the fact that all are highly statistically significant,
the measure that seems to affect all firm sizes in the same manner is Profitability.11
The effect of the Financial Crisis
The relationship between firm’s financial conditions and its investment decisions may have changed
during the period of financial and economic crisis where access to bank finance and the credit markets in
general deteriorated considerably.12 To test this hypothesis equation 1 was re-estimated but now adding
interaction terms, combining the financial crisis dummy with the financial variables of interest, i.e.,:
Invi ,t   i   Invi ,t 1  1 X i ,t 1   2 X i ,t 1Crisisi ,t 1   Salesi ,t 1   j Size j ,t 1  t  Si   i ,t
(3)
where Crisis is a dummy variable that is equal to 1 for the years of 2010 and 2011 and zero otherwise.
The estimation results from this analysis (reported in Table 3) show that the several financial variables used
had different impacts on corporate investment during the financial crisis period. The results concerning
Profitability (column 2) and Interest Burden (column 3) suggest that the effect of these variables was
amplified during the most recent period. More specifically, the results show that the interaction term
between the Crisis dummy and Profitability, the crisis interaction term is positive and highly significant
9 A specification with a continuous measure for firm size (based on the logarithm of total assets at constant prices)
with a quadratic term was tested, but it was not statistically significant, even though the linear term was negative and significant, representative of the same negative relationship between firm size and investment rate.
10 It is worth noting the lack of consensus in the literature regarding the expected sign of the 1-period lag investment variable given that both positive (Martínez-Carrascal and Ferrando, 2008; Barbosa et al., 2007), insignificant (Benito and Hernando, 2007; Bond et al., 2003; Hernando and Martínez-Carrascal, 2008), as well as
statistically significant negative results (Martínez-Carrascal and Ferrando, 2008) have been found.
11 The above analysis was re-estimated by running regressions solely on the specific firm-size categories instead of
using a dummy variable approach. The results from these estimations were in line with the ones found when
using firm-size dummy interactions with the financial variable of interest. More specifically, the financial variables stop being statistically significant as firms get larger with the exception being the firm’s profitability, which
is statistically significant regardless of firm size.
12 The tightening of credit standards being determined by banks difficulties in financing in the international wholesale debt markets, reflecting the increase in sovereign risk premium and a general rise of risk aversion.
Table 2
ECONOMETRIC RESULTS FOR THE INVESTMENT RATE (INVI,T) BY SIZE
Inv i,t-1
Profitability
Interest Burden
Financing Cost
0.0128***
0.0137***
-0.0108**
0.00382
(3.714)
(4.003)
(-2.433)
(0.973)
0.00254*
-0.00244*
-0.00326**
0.00427***
(1.846)
(-1.775)
(-2.104)
(2.818)
0.101***
0.0852***
0.152***
0.0996***
(7.878)
(7.374)
(8.009)
(6.924)
0.0508***
0.0451***
0.120***
0.0460***
(4.163)
(4.127)
(6.409)
(3.338)
0.0262*
0.0102
0.0651***
-0.00961
(1.924)
(0.881)
(3.037)
(-0.651)
-0.0398***
-0.0456***
-0.0290***
-0.0355***
(-12.21)
(-13.60)
(-7.737)
(-10.55)
0.0568***
0.0515***
0.0311***
0.0295***
(15.71)
(14.16)
(8.098)
(5.645)
-0.0554***
-0.100***
-0.206***
-0.167***
(-11.41)
(-31.75)
(-20.03)
(-11.03)
-0.0288***
0.0470***
-0.117***
-0.0635***
(-3.031)
(32.53)
(-10.26)
(-5.828)
-0.0135*
0.0446***
-0.103***
-0.0542***
(-1.769)
(25.69)
(-11.17)
(-5.826)
-0.101***
0.0443***
-0.0856***
-0.153**
(-4.091)
(13.34)
(-6.310)
(-2.542)
-0.134**
0.0384***
-0.0545
0.0689
(-2.092)
(3.006)
(-1.289)
(0.823)
Observations
188,852
188,852
188,852
188,852
No of firms
97,499
97,499
97,499
97,499
Hansen (p-value)
0.0754
0.457
0.302
0.0729
AR 1 (p-value)
0
0
0
0
AR 2 (p-value)
0.833
0.773
0.676
0.990
Sales i,t-1
Micro firms i,t-1
Small firms i,t-1
Medium firmsi,t-1
Year 2009
Year 2010
Year 2011
Xi,t-1 * Micro firmsi,t-1
Xi,t-1 *Small firmsi,t-1
Xi,t-1* Medium firmsi,t-1
Xi,t-1 * Big firmsi,t-1
Source: Banco de Portugal (Central Balance Sheet).
Notes: The variable Xi,t-1 corresponds to the financial variables under analysis interacted with the respective firm size indicator
variable. Lags 1 to 4 in levels of investment rate were used for all regressions. In column 1, it was used as first-differencing instruments lag 2 of Indebtedness*Micro and Indebtedness*Medium, and lag 2 and 3 in first-differencing of Indebtedness*Small and
Indebtedness*Big. It was also used lag 1 to 3 in levels of Indebtedness*Small. In column 3, it was used lags 3 both in levels and
first-differencing for all interest burden interactions. In column 4, it was used as first-differencing instruments lag 3 to 4 of Financing
Cost*Micro, Financing Cost*Small and Financing Cost*Big, and in levels lag 3 of Financing Cost*Small, Financing Cost*Medium and
Financing Cost*Big. In column 2, Profitability interactions were once again used as strictly exogenous. The variables Sales, the Size
indicator dummies, as well as the year and sectoral dummies were used as regular instrument, even though Sales was used only in
levels. T-statistics in parentheses. *** p-value<0.01, ** p-value<0.05, * p-value<0.1.
117
Articles
Indebtedness
Table 3
ECONOMETRIC RESULTS FOR THE INVESTMENT RATE (INVII,T) INTERACTED WITH CRISIS YEAR DUMMY
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
118
Indebtedness
Profitability
Interest Burden
Financing Cost
0.0138***
0.0137***
0.288*
0.00540
(3.992)
(3.985)
(1.754)
(1.487)
0.00172
-0.00243*
-0.00936***
0.00450***
(1.230)
(-1.772)
(-2.621)
(3.212)
0.0934***
0.0841***
0.0912***
0.113***
(8.085)
(7.321)
(3.659)
(9.543)
0.0426***
0.0448***
0.0791***
0.0598***
(3.892)
(4.134)
(3.721)
(5.373)
0.00179
0.00998
0.0559***
0.00677
(0.155)
(0.871)
(3.482)
(0.582)
-0.0406***
-0.0459***
-0.0119
-0.0351***
(-12.40)
(-13.68)
(-1.242)
(-10.56)
0.0512***
0.0537***
0.0845***
0.0258***
(13.20)
(13.99)
(3.025)
(5.873)
-0.0674***
-0.0992***
-0.186***
-0.151***
(-14.08)
(-30.92)
(-12.91)
(-17.06)
0.0192*
0.0431***
-0.103***
-0.0683***
(1.684)
(30.29)
(-13.10)
(-11.26)
-0.0353***
0.00619***
-0.00863*
0.0214***
(-2.689)
(2.972)
(-1.710)
(3.772)
Observations
188,852
188,852
188,852
188,852
No of firms
97,499
97,499
97,499
97,499
Hansen (p-value)
Inv i,t-1
Sales i,t-1
Micro firms i,t-1
Small firms i,t-1
Medium firmsi,t-1
Year 2009
Year 2010
Year 2011
Xi,t-1
Xi,t-1* Crisisi,t-1
0.211
0.454
0.134
0.239
AR 1 (p-value)
0
0
6.07e-05
0
AR 2 (p-value)
0.848
0.772
0.0472
0.998
Source: Banco de Portugal (Central Balance Sheet).
Notes: The variable Xi,t-1 corresponds to the financial variables under analysis interacted with a dummy indicator equal to 1 in years
2010 and 2011. Lag 1 to 3 in levels of investment rate were used for all regressions with the exception of column 3 where it was only
used lag 3. In column 1, it was used lag 1 to 3 of indebtedness in levels and lag 2 in levels and first-differencing for the interaction
variable. In column 3, it was used as instrument lag 2 to 4 of interest burden in first-differencing, and lag 3 to 4 of the interaction in
levels. In column 4, it was used lag 2 to 4 of Financing Cost in first-differencing and lag 2 and 3 of the interaction in levels. In column
2, Profitability interactions were once again used as strictly exogenous. The variables Sales, the Size indicator dummies, as well as the
year and sectoral dummies were used as regular instrument, even though Sales was used only in levels. T-statistics in parentheses.
*** p-value<0.01, ** p-value<0.05, * p-value<0.1.
suggesting that internally generated funds became increasingly important during the period of harsher
access to financial markets. As for Profitability, the crisis interaction term and Interest Burden is negative
and statistically significant (yet, only at the 10% level) indicating that firms with higher debt burdens
suffered the most during this period.
The results also show that, in 2010 and 2011, the effect of Indebtedness on corporate investment was
negative, more than offsetting the positive effect that is estimated for the period prior to 2010 (though
this positive effect is significant only at the 10 per cent level of significance). This result suggests that a
high level of indebtedness can become particularly harmful to firms in a period of economic and financial
stress. Moreover, and despite the fact that no explicit test for the existence of non-linear effects between
financial conditions and corporate investment rates was made (as in Hernando and Martínez-Carrascal,
is non-linear and also depends on external financial conditions.
With respect to the Financing Cost variable, the results show that firms’ investment has been less sensitive
to their cost of financing in the more recent period, a result that is consistent with the prevalence of credit
rationing in the amount of credit available rather than through its price (as suggested by Holmstrom
and Tirole (1997)). Even though Financing Cost remained significant in explaining corporate investment
decisions its impact on investment decisions seemed to decrease during the financial crisis period.
5. Concluding remarks
The aim of this analysis was to study how corporate investment decisions relate to the financial standing
of Portuguese indebted firms, proxied by Indebtedness levels, Profitability, Interest Burden and Financing
Cost. The analysis of the link between corporate financial standing and investment decisions is particularly
relevant given (i) the high indebtedness levels of Portuguese firms, (ii) the reduction in profitability of the
Portuguese firms, which reduces the amount of internally available funds thus increasing the demand
for external financing, and (iii) the ongoing Financial and Economic Crisis that considerably changed the
conditions and access to the credit markets.
The results obtained indicate that firms’ financial standing is indeed relevant in explaining corporate
investment decisions, as the financial variables of interest are all statistically significant in the estimated
investment equations and have the expected signs. More specifically, the results show that the burden
of servicing debt, the cost of capital, and the firm’s indebtedness all have a negative relationship with
firm’s investment rate. Furthermore, the results strongly suggest a positive (and statistically significant)
relationship between the firm’s profitability and its investment decisions. Nonetheless, this sensitivity is
not uniform across firms and depends on some of their specific characteristics. In particular, we studied
potential differences between smaller and larger firms in the relationship between financial standing and
investment rates. From this analysis, we found some evidence that larger firms seem to be less sensitive
to financial pressure, as the significance of some of the financial conditions indicators (Interest Burden
and Financing Cost) were either not-significant in explaining investment decisions for larger firms or had
coefficients with smaller magnitude. In addition, there is evidence suggesting that the impact of firms’
financial standing became more relevant during the period of the sovereign debt crisis in the euro area.
In particular, we found that the magnitude of Profitability and Interest Burden was amplified during this
period. The impact of firm’s indebtedness level moved from positive to negative and the impact of the
Financing Cost seemed to have become less pronounced in the most recent period.
Several avenues for future research may be explored such as the existence of potential non-linear relationships between the financial variables considered and corporate investment decisions. More specifically,
and following the work of Benito and Hernando (2007) and Hernando and Martínez-Carrascal (2008)
we intend to perform quantile regressions to test the hypothesis that by being in a different point of the
distribution of the financial variable relates differently to the level of corporate investment.
119
Articles
2008), these results may also suggest that the impact of financial conditions on firm’s investment demand
Appendix – Investment and Financial Standing according to Firm Size
II
Chart 15
EVOLUTION OF INVESTMENT RATE ACCORDING TO DIFFERENT INDEBTEDNESS LEVELS BY FIRM SIZE |
SEGUNDO A DIMENSÃO DA EMPRESA
Micro firms
Small firms
30
20
20
Per cent
Per cent
30
10
10
0
0
2007
2008
2009
2010
2011
2007
2008
Medium firms
2009
2010
2011
2010
2011
Big firms
30
20
20
Per cent
30
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
120
10
10
0
0
2007
2008
2009
2010
Low Indebtedness
Source: Banco de Portugal (Central Balance Sheet).
2011
Medium Indebtedness
2007
2008
High Indebtedness
2009
Chart 16
EVOLUTION OF INVESTMENT RATE ACCORDING TO DIFFERENT FINANCING COST LEVELS AND FIRM
SIZES | SEGUNDO A DIMENSÃO DA EMPRESA
Micro firms
Small firms
30
20
20
121
Per cent
Per cent
Articles
30
10
0
10
2007
2008
2009
2010
0
2011
2007
2008
Medium firms
2009
2010
2011
2010
2011
Big firms
30
20
20
Per cent
Per cent
30
10
10
0
0
2007
2008
2009
2010
Low Financing Cost
Source: Banco de Portugal (Central Balance Sheet).
2011
Medium Financing Cost
2007
2008
High Financing Cost
2009
Chart 17
EVOLUTION OF INVESTMENT RATE ACCORDING TO DIFFERENT INTEREST BURDEN LEVELS AND FIRM
SIZES | SEGUNDO A DIMENSÃO DA EMPRESA
Micro firms
30
20
20
Per cent
Per cent
30
10
0
10
2007
2008
2009
2010
0
2011
2007
2008
Medium firms
2009
2010
2011
2010
2011
Big firms
30
30
20
20
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
122
Small firms
Per cent
II
10
10
0
0
2007
2008
2009
2010
Low Interest Burden
2011
Medium Interest Burden
Source: Banco de Portugal (Central Balance Sheet).
2007
2008
High Interest Burden
2009
Chart 18
EVOLUTION OF INVESTMENT RATE ACCORDING TO DIFFERENT PROFITABILITY LEVELS AND FIRM SIZES
| SEGUNDO A DIMENSÃO DA EMPRESA
Small firms
50
40
40
30
30
20
20
10
10
0
0
2007
2008
2009
2010
2011
123
Articles
50
Per cent
Per cent
Micro firms
2007
2008
50
50
40
40
30
20
10
10
0
2007
2008
2009
2011
2010
2011
30
20
0
2010
Big firms
Per cent
Per cent
Medium firms
2009
2010
2011
Low Profitability
Source: Banco de Portugal (Central de Balanços anual).
Medium Profitability
2007
2008
High Profitability
2009
References
Alti, A. (2003), “How Sensitive is Investment to Cash Flow when Financing is Frictionless?”, Journal of
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
124
Finance , 58, 707-722.
Arellano, M., & Bond, S. (1991), “Some Tests of Specificaiton for Panel Data: Monte Carlo Evidence and
an Application to Employment Equations”, Review of Economic Studies , 58, 277-297.
Arellano, M., & Bover, O. (1995), “Another Look at the Instrumental-Variable Estimation of Error-Components Models”, Journal of Econometrics, 69, 29-52.
Barbosa, L., Lacerda, A., & Ribeiro, N. (2007), “Investment Decision and Financial Standing of Portuguese
Firms”, Banco de Portugal, Economic Bulletin-Winter.
Benito, A., & Hernando, I. (2007),”Firm Behaviour and Financial Pressure: Evidence from Spanish Panel
Data”, Bulletin of Economic Research, 53 (4), 283-311.
Bernanke, B., Gertler, M., & Gilchrist, S. (1999), “The Financial Accelerator in a Quantitative Business Cycle Framework”, In Handbook of Macroeconomics (Vol. 1, pp. 1341-1393). Amsterdam: Elsevier.
Blundell, R., & Bond, S. (1998), “Initial Conditions and Moment Restrictions in Dynamic Panel Data Models”, Journal of Econometrics , 87, 115-143.
Bond, S., Elston, J., Mairesse, J., & Mulkay, B. (2003), “Financial Factors and Investment in Belgium,
France, Germany and the United Kingdom: a Comparison using Company Panel Data”, The Review
of Economics and Statistics , 85 (1), 153-165.
Farinha, L. (1995), “Investimento, Restrições de Liquidez e Dimensão das Empresas: uma Aplicação ao
Caso Português”, Banco de Portugal, Economic Bulletin -December.
Fazzari, S., Hubbard, R., & Petersen, B. (1988), “Financing Constraints and Corporate Investment”,
Brookings Papers on Economic Activity , 1988 (1), 141-206.
Hernando, I., & Martínez-Carrascal, C. (2008), “The impact of financial variables on firm’s real decisions:
Evidence from Spanish firm-level data”, Journal of Macroeconomics , 30, 543-561.
Holmstrom, B., & Tirole, J. (1997), “Financial intermediation, loanable funds, and the real sector”, Quarterly Journal of Economics , 112, 663-691.
Ivashina, V., & Scharfstein, D. (2010), “Bank Lending during the financial crisis of 2008”, Journal of Financial Economics, 97, 319-338.
Kaplan, S. N., & Zingales, L. (1997), “Do investment-cash flow sensitivities provides useful measures of
financing contraints?”, The Quartely Journal of Economics, 112 (1), 169-215.
Kiyotaki, N., & Moore, J. (2002), “Balance-Sheet Contagion. Papers and Proceedings of The One Hundred Fourteenth Annual Meeting of the American Economic Association. 92”, pp. 46-50, The
American Economic Review.
Kiyotaki, N., & Moore, J. (1997), “Credit Cycles”, Journal of Political Economy, 105 (2), 211-248.
Marchica, M., & Mura, R. (2010), “Financial Flexibility, Investment Ability, and Firm Value: Evidence from
Firms with Spare Debt Capacity”, Financial Management-Winter, 1339-1365.
Martínez-Carrascal, C., & Ferrando, A. (2008), “The Impact of Financial Position on Investment: An
Analysis for Non-Financial Corporations in the Euro Area”, Documentos de Trabajo, 0820, 6-40.
Modigliani, F., & Miller, M. (1958), “The Cost of Capital, Corporation Finance, and the Theory of Investment”, American Economic Review, 48 (3), 261-297.
Myers, S., & Majluf, N. (1984), “Corporate Financing and Investment Decisions When Firms Have Information That Investors Do Not Have”, Journal of Financial Economics, 13, 187-221.
Nickell, S., & Nicolitsas, D. (1999), “How Does Financial Pressure Affect Firms?”, European Economic
Review, 43, 1435-1456.
dynamic analysis”, Small Business Economics, 27 (2), 139-156.
Stiglitz, J., & Weiss, A. (1971), “Credit Rationing in MArkets with Imperfect Information”, American
Economic Review , 71, 393-410.
125
Articles
Oliveira, B., & Fortunato, A. (2006), “Investment decisions and financial standing of portuguese firms: A
Carlos Santos**
Abstract
This note aims at contributing to the assessment of the empirical relevance of a set of
determining factors for bank interest rates in Portugal. For such purpose, an innovative
dataset is used, considering micro-information on most new loan operations to nonfinancial corporations in the period June 2012 to February 2013. The results obtained
point to the existence of a set of factors that induce discrimination in the setting of
interest rates by the different customers. These factors include the risk attached to the
customer, the size of the loan and the customer, their private or public nature, and the
fact of having or not exporting activity.
1. Introduction
For most advanced economies, monetary policy is normally conducted through some form of interest rates
targeting. However, the monetary policy transmission mechanism is rather complex, with several channels
operating simultaneously, some upon the financial intermediaries, other depending on the structural and
current characteristics of the non-financial agents. Thus, it is important for central banks to consider an
extensive range of information, allowing for a thorough analysis in their decision making process. Taking
into account the central role performed by banks in the financial intermediation process, it is crucial
an understanding of how banks set their retail interest rates, both in deposit and lending operations.
In tandem, the assessment of banking interest rate levels is also of particular interest from a financial
stability perspective. Taking into account the existence of a relationship between risk and compensation,
embodied in the risk premium, it is important to assess if lending rate levels are adequate for the risks
assumed. However, this assessment is clearly a multidimensional equation, which may not be feasible
with a few aggregate parameters. This is clearly a field where micro-datasets may prove quite useful.
In this context, Banco de Portugal has established in June 2012 a new statistical requirement, aiming at
obtaining a representative set of data concerning micro data on new euro denominated loan operations
to euro area resident non-financial corporations. This note presents the initial results obtained from the
data available so far, with the variable of interest being the level of interest rate set in each operation.
The potential richness of this innovative dataset will obviously be further explored in future work.
* The author thanks the comments and suggestions of Ana Cristina Leal, Nuno Alves, Paula Casimiro and Rita Lameira. The opinions expressed in the article are those of the author and do not necessarily coincide with those of
Banco de Portugal or the Eurosystem. Any errors and omissions are his sole responsibility.
** Banco de Portugal, Economics and Research Department.
127
Articles
BANK INTEREST RATES ON NEW LOANS TO NON-FINANCIAL
CORPORATIONS– ONE FIRST LOOK AT A NEW SET OF MICRO
DATA*
The remaining of the note is structured as follows: Section 2 describes the statistical data request. Section
3 provides initial insights on the data, highlighting some major composition effects assessed as affecting
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
128
both the level and evolution of the aggregate loan rates on the non-financial corporations segment.
Section 4 resorts to an econometric approach to the assessment of loan interest rate determinants for
the considered period. Section 5 concludes.
2. The dataset
In accordance with its Organic Law, Banco de Portugal (BdP) shall ensure the collection and compilation
of the monetary, financial, foreign exchange and balance of payments statistics, particularly within the
scope of its co-operation with the ECB. This function is also included in the BdP’s contribution to the
national statistical framework, but is also justified on the grounds of the BdP’s need to perform its own
assessment of the Portuguese economy, in general, and of the financial system in particular. Under the
scope of this last objective, BdP is entitled to adjust the definition of its statistical requirements in response
to relevant developments observed in these fields. BdP may demand any entity, private or public, the
necessary information to perform its duties.
In this context, BdP has issued Instruction No. 20/2012, through which it has set additional statistical
requirements, amending Instruction No. 12/2010. The new requirements include individual information
on the banks’ interest rate on new loans to Non-Financial Corporations (NFC). Only euro denominated
operations and loans to euro area resident entities are considered. The new requirement is not exhaustive,
to the extent that it only applies to the institutions that grant in each month at least 50 million euro in
new loan operations with non-financial corporations.1
The new statistical requirement allows for the characterization of each operation in a number of relevant
dimensions. Directly, to the extent that the requirement includes the date of the operation, contractual
maturity, initial rate fixation period, amount, annualized interest rate, the existence or not of collateral, the
relationship nature of the loan (completely new, renegotiation of contract terms with active involvement
of clients or renegotiation of contract terms without the active involvement of the client, i.e., automatic
renewal), and residence in Portugal or in other euro area country. Indirectly, through the possibility of
obtaining additional customer information by means of the linkage to another databases through the tax
identification number. In this note, this has allowed for the characterization of customers according to the
following dimensions: private/public nature, exporting activity, size and NACE classification of the NFC.2
At this stage, it is important to be aware of some characteristics of the dataset which limit its analytical
potential. First, it is available only for a short time period, from June 2012 to February 2013. This limits
the longitudinal analysis of the data, to the extent that some variables have limited volatility in the period.
1 The concept of new operation is defined under the scope of ECB regulation ECB/2001/18 (with the changes
introduced by Regulation ECB/2009/7), concerning statistics on interest rates applied by monetary financial
institutions to deposits and to loans to households and non-financial corporations. This concept excludes the
operations associated with credit restructuring and debt consolidation (grouping of several credits into a new
contract) when there is non-performing situations. Therefore, it may occur that some banks which, on a monthly
basis, engage in credit restructuring operations, may not be subject to the requirement set up by Instruction no.
20/2012, to the extent that those are operations are not taken into account in the above referred eligibility criteria. Nevertheless, the available information clearly signals that the data collected under the new requirement
corresponds to almost all of the operations considered as new operations for the universe of all institutions, to
which add some other operations not included in the new operation concept, such as the already mentioned
credit restructuring operations.
2 NFCs will be classified as micro, small, medium and large, according to the following criteria: Micro corporations: number of employees below 10 and turnover and/or annual balance-sheet total not above 2 million euros.
Small corporations: number of employees below 50 and turnover and/or annual balance-sheet total does not
exceed 10 million euros. Medium-sized corporations: number of employees below 250 and annual turnover not
exceeding 50 million euros and/or annual balance-sheet total not exceeding 43 million euros. Large corporations: remaining cases.
Then, the initial assessment of the data has allowed for the identification of some variables which, in
cooperation with BdP, reporting institutions must improve their answers. This is particularly clear for
the relationship nature of the loan, concept that shall be redefined, in a more precise manner, so that
consistency in the submissions of the different banks is assured. Finally, some initial submissions were
need to adjust the IT solutions. All these aspects point to the need of considering both the data and the
corresponding analysis as having a preliminary nature.
3. Data description (June 2012 – February 2013)
As mentioned, the new statistical requirement allows for the characterization of the new loans to NFCs.
This characterization illustrates the importance of composition changes in the monthly flows of loans
to the determination of the aggregate interest rate. This analysis is viable for operations with resident
entities, for which the tax identification number allows for the inclusion of additional information in the
analysis. The concept of new operation under analysis does not include overdrafts.
Charts 3.1 and 3.2 illustrate the monthly evolution of the number and amount of new loan operations
with NFCs, separating public, private non-exporting and exporting corporations.3
It can be seen that the loans are mostly directed towards private non exporting corporations, both in
terms of number of operations and, to a lesser extent, in terms of amount. It is also noteworthy that
loans to state owned corporations, while clearly limited in terms of number, stand for a non-negligible
share of the total amounts for some months. This is a relevant issue given that the aggregate figure
for the whole NFCs sector is obtained as a weighted average (by amount) of each operation and it is
expectable that the determinants for price setting in loans to state owned corporations may differ from
those for private NFCs. In fact, this seems to be the case, as for most months under analysis, the interest
rates charged on loans to state owned corporations were below those of private corporations (Chart
3.3). In turn, the rates charged on exporting private corporations are typically above those of the other
segments (in average terms, not controlling for the characteristics of the customers and operations).
A closer look at the data suggests that other composition effects may be playing a relevant role in this
result. In fact, for most periods, the minimum rates are charged on loans to large private non-exporting
corporations and to medium and large state owned corporations (Chart 3.4). These subsets stand for a
relatively small fraction of total operations, while being relevant in terms of amounts (Charts 3.5 and 3.6).
Other decompositions of the aggregate figures may be established on the basis of the size of the loan
and the contractual maturity. In a consistent way, visual inspection points to lower interest rates on loans
with higher amounts and on loans with contractual maturity exceeding 1 year (Charts 3.7 and 3.8).
This evidence should be taken into account as signaling the importance of composition effects for the
determination of average aggregate figures for the interest rate on loans to NFCs. These features may
potentially include characteristics (i) of the operation, such as the amount, the contractual maturity,
the initial rate fixation period and the existence of collateral, (ii) of the lending institution, such as its
capital position, liquidity and funding cost, and (iii) of the borrower, such as its size, branch of activity,
overall risk profile (in this note by means of the z-score variable4 ), exporting/non-exporting activity and
state owned/private nature. To some extent, all these features may contribute to justify the differences
3 Private-owned exporting companies are defined as a) companies that export more than 50% of the turnover;
or b) companies that export more than 10% of the turnover and the total amount exceeds 150 thousand euro.
In turn, public NFCs include entities controlled by the public administrations which are not included in that institutional sector.
4 This variable was calculated for 2011, according to the methods and specifications presented in “A scoring
model for Portuguese non financial enterprises”, Ricardo Martinho and António Antunes (2012).
129
Articles
conditioned by the difficulty in achieving the precision in the collection of some variables, due to the
NUMBER OF OPERATIONS
AMOUNT
60 000
5 000
Public
Private non-exporting
Private exporting
Total
50 000
Public
Private non-exporting
Private exporting
Total
4 500
4 000
3 500
40 000
Million EUR
3 000
30 000
20 000
2 500
2 000
1 500
1 000
10 000
0
500
45
66
Jun-12
45
83
Aug-12
54
Oct-12
46
52
Dec-12
57
0
34
Jun-12
Feb-13
Source: Banco de Portugal.
Aug-12
Oct-12
Dec-12
Feb-13
Source: Banco de Portugal.
between aggregate loan rates for different countries. However, due to comparable data unavailability
for other countries, this exercise cannot be further developed for the time being.
However illustrative, the analysis presented above does not take into account the simultaneous relevance
of the different factors in determining interest rates. This will be addressed in the next section, resorting
to an econometric approach.
Chart 3.3
Chart 3.4
INTEREST RATE
INTEREST RATE
10
7.0
9
6.5
8
7
6.0
6
Per cent
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
130
Chart 3.2
Per cent
II
Chart 3.1
5.5
5
4
5.0
3
2
4.5
1
4.0
Jun-12
Aug-12
Total
Public
Private non-exporting
Private exporting
Source: Banco de Portugal.
Oct-12
Dec-12
Feb-13
0
Jun-12
Aug-12
Oct-12
Public micro
Public small
Public medium
Public large
Private non-exporting micro
Private non-exporting small
Private non-exporting medium
Private non-exporting large
Private exporting micro
Private exporting small
Private exporting medium
Private exporting large
Source: Banco de Portugal.
Dec-12
Feb-13
Chart 3.5
Chart 3.6
NUMBER OF OPERATIONS
AMOUNT
1 600
12 000
1 400
10 000
1 200
8 000
1 000
6 000
4 000
131
Articles
1 800
14 000
Million EUR
16 000
800
600
400
2 000
200
large
Public
Jun 12
large
small
medium
large
micro
medium
Private non-exporting
Private exporting
Feb 13
Source: Banco de Portugal.
Source: Banco de Portugal.
Chart 3.7
Chart 3.8
INTEREST RATE
INTEREST RATE
9
9
8
8
7
7
6
6
5
5
Per cent
Per cent
small
micro
large
small
Private exporting
medium
0
micro
small
medium
large
Private non-exporting
micro
small
medium
large
Public
micro
small
medium
micro
0
4
4
3
3
2
2
1
1
0
0
Jun-12
Aug-12
Oct-12
Public < EUR 0.25 M
Public < EUR 1 M
Public > EUR 1 M
Private non-exporting < EUR 0.25 M
Private non-exporting < EUR 1 M
Private non-exporting > EUR 1 M
Private exporting < EUR 0.25 M
Private exporting < EUR 1 M
Private exporting > EUR 1 M
Source: Banco de Portugal.
Dec-12
Feb-13
Jun-12
Aug-12
Oct-12
Public ≤ 31 dias
Public ≤ 91 dias
Public ≤ 365 dias
Public > 365 dias
Private non-exporting ≤ 31 dias
Private non-exporting ≤ 91 dias
Private non-exporting ≤ 365 dias
Private non-exporting > 365 dias
Private exporting ≤ 31 dias
Private exporting ≤ 91 dias
Private exporting ≤ 365 dias
Private exporting > 365 dias
Source: Banco de Portugal.
Dec-12
Feb-13
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
132
4. Econometric assessment
The variable under study is the level of the interest rate set in each euro denominated new loan operation
with a resident NFC. This variable will be assessed in terms of an array of theoretical determinants. This
assessment will be based on the evidence obtained through a set of alternative econometric specifications, on the basis of the sets of variables presented in the preceding sections, which will allow for an
indication of the robustness of the results.5
A priori, a number of factors can influence the determination of interest rates in NFCs loans: the cost
of funds and its volatility, operating expenses (transaction costs, risk management costs, geographic
dispersion), provisioning costs (the cost of risk and the regulatory framework), tax expenses, level of
competition, credit rating of the customer (risk premium), inflation levels, management competency
(product adequacy), financial literacy of customers.
At this stage, we will present results based on the full information obtained from Instruction no. 20/2012
and from a still limited set of data concerning customer and bank information.6 Future work will benefit
from the gradual inclusion of other potentially significant data, and from the progressive lengthening
of the time dimension of the dataset. This lengthening will favor increased volatility for some variables,
for which the sample now considered is rather limited – with the variables associated with the bank’s
characteristics being the main example at the moment.
The following table presents the results from a set of econometric specifications, applied to the data
covering the period from June 2012 to February 2013. The results should be read as an econometric
systematization of the behavior of banks in setting the interest rates during the short period under
analysis, and not necessarily reflecting any structural pricing models.
The reading of the table suggests the following regularities:
•
The interest rate level exhibited a negative relation with the maturity of the operations. This may
reflect the fact that longer term operations are typically associated with investment operations. The
purpose of the loan is not available in the information set. Further, banks shall be granting longer
term loans to customers for whom the associated prospective risk is smaller, which shall benefit
from lower interest rates;
•
In the same vein, there is a negative relation between the amount of the loan and the level of the interest
rate. This result calls for future additional analysis, as it is expectable that the relevant variable should be the
total amount of loans granted to each customer and not necessarily the amount of each operation;
•
NFC risk (as proxied by the zscore) is clearly significant, signaling the importance of NFCs’ financial situation
in the determination of the cost of funding. However, it should be noticed that this variable comes as lagged,
as it is based on 2011 information, with banks having, typically, a more updated and forward looking
approach to risk assessment;
5 The robustness of the results was also assessed by means of the estimation of the presented specifications for
different samples, namely through the exclusion of the operations with public NFCs and of repeated operations,
i.e., those for which all relevant variables are repeated, situation that takes place in the most visible manner
when there is a sequence of successive renewals of short-term operations, so that the only change in the observations concerns the settlement date.
6 The only variable excluded from the empirical analysis was the relationship nature of the loan, to the extent that
the concept is not yet fully harmonized among the reporting institutions. Additional restrictions to the samples
come as the result of the unavailability of information for some variables in some of the months of the sample.
Table 1
ECONOMETRIC SPECIFICATIONS
I
II
III
IV
V
VI
VII
VIII
Log (maturity, days)
-0.83
-0.81
-0.86
-0.84
-0.86
-0.98
-0.98
-0.98
Log (amount, EUR million)
-0.13
-0.15
-0.12
-0.15
-0.13
-0.19
-0.19
-0.19
0.25
0.20
0.27
0.22
0.27
0.61
0.61
0.61
-
-
-
-
-
-
-
-
0.03
0.02
0.03
0.02
0.03
0.03
0.03
0.03
-
-
-
-
-
-
-
-
Small
-1.03
-1.52
-1.05
-1.53
-1.05
-1.26
-1.27
-1.27
Medium
-2.00
-2.60
-2.03
-2.61
-2.02
-2.42
-2.42
-2.45
Large
-2.35
-2.51
-2.39
-2.52
-2.38
-2.91
-2.90
-2.91
Dummy (public corporation)
-0.51
-1.15
-0.42
-1.07
-0.42
-0.32
-0.34
-0.34
Dummy (exporting)
-0.45
-0.43
-0.46
-0.44
-0.46
-0.56
-0.55
-0.56
Dummy (economic activity sector)
-
-
-
-
-
-
-
-
Dummy (bank)
-
-
-
-
-
Core Tier 1 ratio
0.27
0.24
0.30
Credit at risk ratio - NFC
0.29
0.30
0.28
Characteristics
of the
operation
Dummy (colateral)
Dummy (relationship nature of
the loan)
Z-score (%)
Dummy (dimension)
Characteristics
of the
corporation
Characteristics
of the bank
Interest rate on deposits outstanding for NFPS
-0.35
Dummy (domestic banks)
Cross-effects
0.44
Cross dummy (bank * z-score)
-
-
Cross dummy (bank * dimension
of corporation)
-
-
Euribor
Dummy (month)
Constant
No. observations
R-squared
-
-
10.8
11.1
0.79
0.84
10.4
10.7
-0.29
0.20
0.53
0.39
-
-
-
11.9
5.7
6.7
4.6
372 217 372 217 338 130 338 130 338 097 301 775 301 775 301 775
43%
44%
43%
44%
43%
37%
37%
36%
Source: Banco de Portugal.
Notes: The gray areas correspond to the variable (in line) that was not included in the specification (in column). In turn, the dashes
signal that the variable was included in the specification, even if, for sake of parsimony, the corresponding coefficients are not presented; NFC: Non-financial corporation; NFPS: Non-financial private sector; Credit at risk ratio: please refer to definition in “Section
4 Credit risk”, of this Report.
.
•
NFC size revealed a significant negative relation with the interest rate level;
•
State owned NFCs benefited from a premium in the interest rates they pay, i.e., they tend to pay
lower interest rates;
•
Likewise, exporting NFCs benefited from a premium in their loan interest rates, in the range of
40-50 basis points;
•
The cost of funding, proxied either by money market rates (euribor) or interest rates on deposits,
played a relevant role in determining the cost of loans;
•
Operations for which collateral was reported to exist recorded higher interest rates. This is a standard result found in the research carried for Portugal, signaling that those borrowers could not
access bank loans without collateral and that the collateral posted is not sufficient to compensate
the higher risk of these borrowers. The coefficient shall be interpreted as reflecting a regularity and
not a causal relation.
133
Articles
Variable / Specification
5. Conclusions
This note shows the first systematization of the information Banco de Portugal has started collecting
II
BANCO DE PORTUGAL | FINANCIAL STABILITY REPORT • May 2013
134
under the scope of statistical requirement defined in Instruction no. 20/2012, concerning micro data
on the banks’ interest rate on new euro denominated loans to resident Non-Financial Corporations.
The focus of the analysis was on the level of the loans interest rates and on the empirical assessment of
its determinants. For the reasons put forward throughout the note, this systematization shall be read as
preliminary, reflecting ongoing work.
The evidence presented highlights the importance of composition effects in the determination of the
aggregate interest rate on loans to NFCs. Composition effects may concern size, sector, state owned
or private nature of the counterparty, term of the operation or associated credit risk. This suggests the
need to use of caution in the short term analysis of the evolution of aggregate interest rates, and on
the comparison of interest rates recorded in Portugal and in other euro area countries for this segment.
At another level, the econometric analysis confirms the importance of a set of factors considered as relevant in the determination of bank’s interest rates, such as the cost of funding and credit risk. In fact, the
coefficients associated with either money market or deposit interest rates and with the z-score variable,
used as a proxy for credit risk, appear as significant and positive. In this context, it should be mentioned
that the obtained results illustrate the importance of promoting a sounder financial structure for NFCs,
as a means to promote a reduction in the cost of bank loans.
Public NFCs tended to benefit from lower interest rates. While corresponding to a small number of customers, the operations with public NFCs tended to assume relevant total amounts, therefore significantly
affecting the aggregate level of interest rates.
The size of the loans and of the NFCs presents a negative relation with the cost of funding through bank
loans. Longer term operations, higher amounts and larger customers tended to benefit, in the sample
period, of lower interest rates. Finally, exporting firms benefited from a premium in their funding through
bank loans (close to 50 basis points).
References
Martinho, R. and Antunes, A., (2012), “A Scoring Model for Portuguese Non-Financial Enterprises”,
Banco de Portugal, Financial Stability Report - November.
RELATÓRIO DE ESTABILIDADE FINANCEIRA Maio 2013
RELATÓRIO DE
ESTABILIDADE FINANCEIRA
Maio 2013
EUROSISTEMA