View PDF

The full text on this page is automatically extracted from the file linked above and may contain errors and inconsistencies.

23 July 2010

Aggregate outcome of the 2010 EU wide stress test exercise coordinated
by CEBS in cooperation with the ECB

Executive summary
The Committee of European Banking Supervisors (CEBS) was mandated by the
ECOFIN of the European Council to conduct in cooperation with the European
Central Bank (ECB), the European Commission and the EU national supervisory
authorities a second EU-wide stress test exercise.
The overall objective of the 2010 exercise is to provide policy information for
assessing the resilience of the EU banking system to possible adverse economic
developments and to assess the ability of banks in the exercise to absorb
possible shocks on credit and market risks, including sovereign risks.
The stress test has been conducted on a bank-by-bank basis and using bank’s
specific data and supervisory information.
CEBS has coordinated the exercise and conducted extensive cross-checks over
the results, which were submitted to a peer review and challenging process in
order to ensure the consistency and comparability of the results. This report
provides details on the scenarios, methodologies and aggregate results of the
stress test exercise. Results of the individual banks and comments on follow-up
actions, where needed, are provided by the banks participating in the exercise
and/or their national supervisory authorities. The results are re-published by
CEBS on its website.
National supervisory authorities routinely conduct stress testing exercises in their
respective jurisdictions, both at system-wide and individual institutions’ levels, in
order to assess potential risks facing the institutions individually and/or
collectively. The CEBS exercise is intended to complement these national

analyses by providing a coordinated assessment of European banks, using
common scenarios and methodologies.
However, as with any stress test exercise, the results are not forecasts of
expected outcomes, since the scenarios are designed as "what-if" situations
reflecting

extreme

assumptions,

which

are

therefore

not

very

likely

to

materialise. Against this background, the aggregate results discussed in this
report as well as the individual results presented by banks and/or national
supervisory authorities, aim at supporting the supervisory assessment of the
adequacy of capital of European banks, and should be interpreted with caution.
Sample of banks
The 2010 stress test exercise has been conducted on a sample of 91 European
banks 1 . In total national supervisory authorities from 20 EU Member States
participated in the exercise. In each of the 27 Member States, the sample has
been built by including banks, in descending order of size, so as to cover at least
50% of the respective national banking sector, as expressed in terms of total
assets. As the stress test has been conducted on the highest level of
consolidation for the bank in question, the exercise also covers subsidiaries and
branches of these EU banks operating in other Member States and in countries
outside Europe. As a result, for the remaining 7 Member States where more than
50% of the local market was already covered through the subsidiaries of EU
banks participating in the exercise, no further bank was added to the sample.
The 91 banks represent 65% of the total assets of the EU banking sector as a
whole.
Given the differences in size and complexity, business models, scope of
operations and risk profiles of the institutions included in the sample, it should be
borne in mind that the aggregate results presented in this report cannot be
directly applied to individual institutions, nor can be directly extrapolated to other
banks in the EU. This point is of special importance as regards the assessment of
banks’ continued reliance on government support measures, as the sample of
banks contains both institutions making use of various support measures, and
institutions which did not revert to public support.
Risk factors included in the stress test exercise
1

The sample of the 2009 exercise was composed of 22 large cross-border banks.

2

The stress test focuses mainly on credit and market risks, including the
exposures to European sovereign debt. The focus of the stress test is on capital
adequacy; liquidity risks were not directly stress tested.
The exercise has been carried out on the basis of the consolidated year-end 2009
figures and the scenarios have been applied over a period of two years - 2010
and 2011.
Scenarios used in the exercise
For the purpose of stress testing the credit risk and simulating the profit and
losses, two sets of macro-economic scenarios (benchmark and adverse) have
been developed, in close cooperation with the ECB and the EU Commission. The
benchmark scenario was based on the EU Commission Autumn 2009 forecast and
the European Commission Interim Forecast in February 2010, with several
adaptations to reflect recent macro-economic developments in a number of
countries. The adverse macro-economic scenario was based on ECB estimates.
Within the adverse scenario, the exercise also envisages a “sovereign risk
shock”, reflecting adverse conditions in financial markets.
For each macro-economic scenario, a set of key macro-economic variables
(including GDP, unemployment, interest rate assumptions) was provided for the
domestic situation for each EU Member State, the US, and the rest of the world
collectively. Some of the input parameters and assumptions have been provided
by CEBS, and by the participating supervisory authorities outside of the narrative
of the macro-economic scenarios as provided by the EU Commission and the
ECB, notably the evolution of the real estate prices.
The benchmark macro-economic scenario assumes a mild recovery from the
severe downturn of 2008-2009, whereas the adverse scenario assumes a
“double-dip” recession. For the euro area, the GDP growth under the benchmark
scenario is assumed at a level of +0.7 (2010) and +1.5% (2011), whereas under
the adverse scenario the euro area would see a decrease of GDP by -0.2% in
2010 and -0.6% in 2011. For the whole European Union (EU27) the benchmark
scenario assumes a +1.0% growth of GDP in 2010 and +1.7% in 2011, whereas
under the adverse scenario the GDP would not grow in 2010 and would decline
by -0.4% in 2011. On aggregate and over the two-year time horizon, the
adverse scenario assumes a three percentage point deviation of GDP for the EU

3

compared to the benchmark scenario. It should be noted that current macroeconomic developments remain in line with the assumptions provided in the
benchmark scenario.
In addition to a global confidence shock, that affects demand worldwide, the
adverse scenario envisages an EU-specific shock to the yield-curve, originating
from a postulated aggravation of the sovereign debt crisis. The latter impact is
differentiated across countries, taking into account their respective situation.
In particular, related to prevailing sovereign debt risks, a common upward shift
in the yield curve was applied for each country in the EU (reaching 125 basis
points for the three-month rates and 75 basis points for the 10-year rates at
end-2011), supplemented with country-specific upward shocks to long-term
government bond yields (overall amounting to 70 basis points at end-2011 for
the euro area). The rise in short-term rates reflects an assumption of tensions in
the interbank market – as was seen during earlier financial turmoil episodes. The
country-specific bond yield shock in turn accounts for differentiated fiscal
situations and related market perceptions. This results in a set of haircuts to be
applied to all EU sovereign bond holdings in the trading books of the banks in the
sample.
For the purposes of the market risk stress test, a set of stressed market
parameters was applied to the trading book positions. It should be noted that the
parameters developed for the market risk stress test are in-line with the macroeconomic scenarios, and therefore could be considered as directional, meaning
that depending upon the size and direction of their exposures, banks were able
to make gains on certain portfolios, thereby reducing the overall amount of
stress coming from the market parameters.
Key common assumptions used in the exercise
The exercise was conducted, using common templates, a common methodology
and under key common assumptions. In particular, the exercise assumes, both
for the benchmark and for the adverse scenarios, a “zero growth” assumption for
the evolution of exposures for market and credit risks over the whole stress
horizon. However, any regulatory imposed decisions (e.g. restructuring plans
agreed with the EU Commission under the State Aid reviews) as well as
management actions (e.g. capital raisings or divestment programmes) publicly

4

announced before 1 July 2010 have also been taken into account. The results do
not include any government support of recapitalisation measures taken after
1 July 2010.
In conducting the exercise, for the major cross-border banking groups, the
macro-economic scenarios were translated using internal models, internal risks
parameters and granular portfolio data, whereas for the less complex institutions
more simplified approaches were used in general (e.g. use of the reference
parameters provided by the ECB for instance).
Securitisation positions have been tested under the assumption of rigorous and
uniform reductions in the credit quality of the positions as of end 2009, which
already incorporated very material reductions in external credit ratings, as
compared to their original level. For the adverse scenario, the assumed reduction
in credit quality of the positions is equivalent to four external rating notches over
two years. The impact of such reduction has been recorded as an increase in
risk-weighted assets (the denominator of the solvency ratio) and as a direct
reduction of regulatory capital (the numerator of the solvency ratio).
Equity exposures in available for sale portfolios have been subject to a
cumulative haircut over two years of 19% in the benchmark scenario, and 36%
in the adverse scenario. Other exposures in available for sale portfolios (i.e.
bonds and loans) have been tested along with other credit exposures in the
banking book.
In light of these assumptions, the information provided for the benchmark and
forecast scenarios should in no way be construed as forecasts.
Although the exercise did not prescribe any specific restrictions to the
profitability of operations and reduction of income, especially generated in the
regions not covered directly by the macro-economic scenarios, the assumptions
and forecasts used by the banks have been challenged by the respective national
supervisory authorities and brought to the attention of CEBS.
Aggregate results
Based on the results of the calculations, the aggregate Tier 1 capital ratio, used
as a common measure of banks’ resilience to shocks, would decrease under the
adverse scenario including sovereign shock from 10.3% in 2009 to 9.2% by the

5

end of 2011. It should be noted that the aggregate Tier 1 capital ratio
incorporates approximately 169.6 bn € of government capital support provided
until 1 July 2010, which represents approximately 1.2 percentage point of the
aggregate Tier 1 capital ratio. It should be noted that the maturity of
government support measures extended to banking institutions in the sample
goes way beyond the two-year time horizon of the exercise. As such,
government support form an integral and stable part of the Tier 1 capital ratios
of the banks in question. It is not expected that any withdrawal of government
support measures could take place without appropriate substitution by private
funding sources, where relevant.
The downward pressure on capital ratios under the adverse scenario including
sovereign shock is mostly stemming from impairment losses (472.8 bn € over
the two-year period) and trading losses (25.9 bn € over the two-year horizon).
Losses associated with the additional sovereign shock would reach 67.2 bn €
over the two-year period (among which 38.9 bn € associated with valuation
losses of sovereign exposures in the trading book). In total, aggregate
impairment and trading losses under the adverse scenario including the
additional sovereign shock would amount to 565.9 bn €.
The average two-year cumulative loss rates associated with these losses are
3.0% for corporate exposures and 1.5% for retail exposures under the
benchmark scenario, and 4.4% for corporate and 2.1% for retail exposures
under the adverse scenario, compared with average loss rates of 1.5% for
corporate exposures and 0.8% for retail exposures in 2009.
As a result of the exercise, under the adverse scenario 7 banks would see their
Tier 1 capital ratios fall below 6%, with an overall shortfall of 3.5 bn € of Tier 1
own funds. The threshold of 6% is used as a benchmark solely for the purpose of
this stress test exercise. This threshold should by no means be interpreted as a
regulatory minimum (according to the CRD 2 the regulatory minimum for the
Tier 1 capital ratio is set to 4% 3 ), nor as a capital target reflecting the risk profile
of the institutions, the latter being the outcome of the supervisory review process
under Pillar 2 of the CRD.

2

Directive EC/2006/48 – Capital Requirements Directive (CRD)
The CRD regulatory minimum for the overall capital adequacy ratio is set to 8% with a
minimum Tier 1 capital adequacy ratio set to 4%. Several EU Member States have opted
for higher minimum capital adequacy ratios.

3

6

The aggregate results suggest a rather strong resilience for the EU banking
system as a whole and may appear reassuring for the banks in the exercise,
although it should be emphasized that this outcome is partly due to the
continued reliance on government support for a number of institutions. However,
given the uncertainties over the actual path of the macro-economic recovery, the
result should not be seen as a reason for complacency.
The adverse macro-economic developments seen in 2008-2009 (EU27 GDP
falling by -4.2% in 2009) led to record high loan losses reported in 2009,
whereas in early 2010 we witnessed improved macro-economic conditions which
suggest an increase of capital ratios attributed to higher retained earnings
affected by lower loan losses. In addition, it should be noted that interest
assumptions of the macro-economic scenarios, while having a minor impact on
the loan losses, may have a sizable offsetting impact on the income side, leading
to an increase of net interest income in some cases and thus positively affecting
the profitability of some banks. Last, but not least, many of the banks in the
exercise have significant operations outside the EU. Some of these countries
have weathered the crisis comparably well and continue to show strong economic
growth. Further, increased revenue streams from those economies positively
contribute to these banks overall profitability, offsetting loan losses and building
sizeable retained earnings.
Follow-up on the stress test results
Part of the mandate of CEBS is to undertake on a periodic basis these EU-wide
stress testing exercises. CEBS will continue with testing the resilience of the EU
banking sector by means of periodic EU wide and thematic risk assessments and
stress testing exercises, and will continue its work on improving convergence in
supervisory practices across Europe by addressing the topics both from a policy
and practical perspective.
CEBS supports the greater transparency of this exercise and of the results of this
stress test exercise, given the specific market circumstances under which banks
currently operate and thus welcomes the decision to publish bank individual
results, as well as detailed information on banks’ exposures to EU sovereign
debt.

7

With respect to the situation of individual institutions that fail to meet the
threshold for this stress test exercise, the competent national authorities are in
close contact with the banks in question to assess the results of the test and
their implications, in particular any potential need for recapitalisation.
The banks are expected to propose a plan to address the weaknesses that have
been revealed by the stress test. The plan will have to be implemented within an
agreed period of time, in agreement with the supervisory authority.
Details of the follow-up actions are provided at national level by the supervisory
authorities.

8

Table of contents
1.
2.
3.

Background and introduction ........................................................... 10
Objectives of the exercise ............................................................... 11
Overview and main features of the exercise ....................................... 11
3.1
Timeline .................................................................................... 11
3.2
Scope of the exercise .................................................................. 12
3.2.1
Sample of banks subject to the exercise................................... 12
3.2.2 Risk factors tested and scope of consolidation ............................... 12
3.3
Time horizon and reference date ................................................... 13
3.4
Conduct of the exercise by institutions and national supervisors ........ 13
3.4
Relation of the exercise with proposed regulatory changes ................ 14
4.
Scenarios and methodologies used in the exercise .............................. 15
4.1
General features of the macro-economic scenarios........................... 15
4.1.1
Features of the adverse macro-economic scenario ..................... 16
4.1.2
Translation of the macro-economic scenarios into reference risk
parameters and their application ......................................................... 19
4.2
General features of the trading book stress test .............................. 20
4.4
Treatment of securitisation transactions ......................................... 21
4.5
Government support measures ..................................................... 21
5. Aggregate outcome of the exercise ........................................................ 22
5.1
Evolution and dispersion of capital ratios ........................................ 22
5.2
Evolution of the main components of the aggregate Tier 1 capital ratio25
5.3
Evolution of financial results and total losses from the stress............. 27
5.4
Total loss estimates from the stress and observed loss rates ............. 29
6. Comparison of key financial indicators with realised figures..................... 32
7.
General follow-up actions ................................................................ 34
Annex 1. List of banks covered by the 2010 EU-wide stress test exercise ........ 36
Annex 2. ECB Technical note on the macro-economic scenarios and reference
risk parameters ...................................................................................... 39
Annex 3. Evolution of property prices assumed in the exercise ...................... 53
Annex 4. Backtesting of key variables of the macro-economic scenarios ......... 54
Annex 5. Set of parameters for the market risk component of the exercise ..... 55

9

1. Background and introduction
1.

Following the mandate given by the EFC in 2009, the Committee of
European Banking Supervisors (CEBS), in cooperation with the European
Central Bank (ECB), the European Commission and the national supervisory
authorities, has conducted the first EU-wide supervisory stress testing
exercise aimed at assessing the overall resilience of the financial sector in
Europe and banks ability to absorb future shocks.

2.

The results of the 2009 stress test were presented to the September 2009
Financial Stability Table of the Economic and Finance Committee of the EU
Council (EFC) and then to the October 2009 meeting of the ECOFIN. Further
to the discussion of the results of the 2009 exercise, the EFC has requested
CEBS to carry out another stress test in 2010, with a broader objective of
assessing the overall resilience of the EU banking sector and the banks’
ability to absorb further possible shocks on credit and market risks, and to
assess the current dependence on public support measures. This mandate
has been confirmed by the ECOFIN at its 2 December 2009 meeting 4 .

3.

Following the mandate, CEBS and the involved participants have started
with the preparation for the 2010 exercise in early 2010, and have carried
out a first phase of the analysis aimed at 26 major cross-border banking
groups in Europe in April-May 2010. In June 2010, it was decided to extend
the exercise to a wider share of the EU banking system also covering
relevant domestic financial institutions, so as to cover at least 50% of each
Member State’s banking sector, as expressed in terms of total assets.

4.

In the design and conduct of the 2010 exercise, CEBS took into account the
mandate given by the ECOFIN, and also addressed areas where
improvements compared to the 2009 exercise were deemed necessary as a
result of a “lessons learnt” analysis conducted by the CEBS Secretariat and
the authorities which participated in the exercise in early 2010. The exercise
has been also tailored to address the acute developments on the sovereign
debt market in Europe observed in May 2010.

5.

This aggregate report presents the process, scenarios, key assumptions,
methodology and aggregate results of the EU-wide stress testing exercise.
The report is also complemented by a set of annexes detailing the scenarios
and assumptions, and should be read in conjunction with the supplementary
information on bank specific results for each of the 91 banks included in the
sample.

6.

The EU-wide stress test was conducted, according to the Terms of
References agreed by CEBS and all involved parties, including the ECB, EU
Commission, national supervisory authorities and the Economic and Finance
Committee (EFC) of the EU Council. The exercise was conducted on a bankby-bank basis, on the highest level of consolidation, at the level of every
participating Member State. Banks’ calculations have been rigorously
reviewed and challenged by the respective national supervisors, before
being analysed, discussed and aggregated by the CEBS Secretariat, which

4

See
http://www.consilium.europa.eu/uedocs/cms_data/docs/pressdata/en/ecofin/111706.pdf

10

conducted in-depth consistency checks and challenged the results with
national supervisors.
7.

The analysis presented in this report is based on the aggregation of the
results of the stress tests done at the level of participating institutions.
Therefore the analysis should not be directly extrapolated to the EU banking
system as a whole. The reader should also bear in mind the differences in
business models, risk profiles and financial standing of the institutions in the
sample as well as in their use of government support measures.

8.

In interpreting the results of the exercise, it should be noted that a stress
testing exercise does not provide forecasts of expected outcomes, since the
scenarios are designed as "what-if" situations including, in particular for the
adverse scenario, plausible but extreme assumptions which are therefore
not very likely to materialise. Against this background, the aggregate results
discussed in this report as well as individual results presented in the
annexes should be interpreted with caution and should not be considered as
representative of the current situation nor a forecast for the future.

2. Objectives of the exercise
9.

Given the mandate received from the ECOFIN, the overall objective of the
2010 stress test is similar though broader than the one conducted in 2009,
which assessed the overall resilience of the EU banking sector and banks
ability to absorb possible shocks and overall financial stability implications
by conducting a stress test of a sample of European cross-border bank.

10. The overall objective of the exercise is to increase the level of aggregate
information among policy makers in assessing the resilience of the European
financial system as a whole. To this end, this report provides aggregate
information illustrated by dispersions of individual results.
11. However, in order to increase the transparency of the exercise and to
provide more granular information to the markets and wider audience given
the specific market circumstances under which banks currently operate, the
European Council decided to publicly disclose the bank-specific outcomes of
the exercise 5 . To this end the aggregate information presented in this report
is supported by the individual outcomes and follow-up actions disclosed by
the participating supervisory authorities and banks, where applicable.

3. Overview and main features of the exercise
3.1 Timeline
12. The preparatory phase of the exercise has started in January-February 2010
with the analysis and the “lessons learnt” from the 2009 exercise. In March
2010, agreement between all involved parties, including CEBS, national
5

EU Council conclusion of 17 June 2010, see:
http://www.consilium.europa.eu/uedocs/cms_data/docs/pressdata/en/ec/115346.pdf

11

supervisory authorities, the ECB and the EU Commission was reached on the
main features of the macro-economic scenarios, which were later
complemented by the supplementary sovereign risk scenario to reflect acute
developments in the sovereign debt markets in Europe. The exercise was
later enlarged to a wider sample of banks and finalised in July 2010.

3.2 Scope of the exercise
3.2.1

Sample of banks subject to the exercise

13. The 2010 EU-wide stress test exercise is carried out on a much wider
sample of banks compared to the 2009 exercise. In addition to the 26 major
cross-border banking groups in Europe, which are followed by CEBS in its
work on regular micro-prudential risk assessment, the stress test also
covers 65 other predominantly domestic credit institutions in Europe.
Altogether the exercise covers 91 banks in Europe, with total assets of
28 032bn € as of end of 2009, representing approximately 65% of the EU
banking system.
14. The sample of banks covers at least 50% of the national banking sectors in
each EU Member State, as expressed in terms of total consolidated assets.
Banks have been included in the exercise in descending order of their
market shares by total assets in each Member State, without any omissions.
As the exercise is conducted at the highest level of consolidation covering all
subsidiaries and branches operating in foreign countries, this effectively
means that if the market share in terms of total assets of EU banks
subsidiaries and branches in any given Member State was more than 50%,
no other bank had to be included from that Member State, unless he wished
to on a voluntary basis. As a result, 20 national authorities participated in
the exercise. The list of banks included in the exercise is attached in Annex
1 to this report.
15. The reader should be mindful that the sample of institutions is quite diverse
in terms of size, business models and risk profiles of institutions. In
particular, it contains banks which have received government support (in
the form of capital injections, asset relief measures and guarantees on
liabilities) as well as banks that have not been subject to any government
support measures. CEBS is confident that the sample is representative
enough to provide a good proxy of the overall resilience of the EU banking
sector.

3.2.2 Risk factors tested and scope of consolidation
16. With respect to the risk factors covered in the stress and similar to the 2009
exercise, the focus was put on assessing credit and market risks. Both
trading and banking book assets (including off-balance sheet exposures)
have been subject to stress testing at the highest level of consolidation for
the banking group (or banking arm of a financial conglomerate). The focus
on credit risk is fully in line with the outcomes of the regular CEBS micro-

12

prudential risk assessments, which highlighted the credit risk and associated
losses as a top source of concern for major cross-border banking groups.
17. Although the focus of the exercise remains on credit and market risks,
capital requirements for operational risk were also taken into account in the
exercise by computing a proxy of year-on-year changes in operating profit
of the participating institutions, with the actual capital charge as of year-end
of 2009 acting as a floor.

3.3 Time horizon and reference date
18. The exercise has been carried out on the basis of the consolidated year-end
2009 figures and the scenarios have been applied over a period of two years
– 2010 and 2011. The time horizon of two years is consistent with the
approach used in the 2009 exercise and most current stress testing
practices of institutions and national authorities, as well as in line with the
principles set forward in the draft CEBS’s Guidelines for stress testing 6 .
19. Government support measures received by institutions in the sample as of
end 2009 have been taken into account and subject to specific analysis (see
Section 4.5 of the report). The actual maturity of most government support
measures and instruments goes beyond the time horizon of the exercise.
20. Any regulatory imposed decisions (e.g. restructuring plans agreed with the
EU Commission) as well as management actions (e.g. capital raisings or
divestment programmes) announced before 1 July 2010 have also been
incorporated in the assessment.

3.4 Conduct of the exercise by institutions and national
supervisors
21. The exercise has been conducted on a bank-by-bank basis as a centrally
coordinated process, where the responsibility for the actual conduct of the
stress tests lies with national supervisory authorities of the banks subject to
the exercise, subject to guidelines provided by CEBS and the ECB and
agreed by all participants.
22. Given the relatively diverse sample of banks covered by the exercise both in
terms of their size and complexity, but also sophistication of risk
management techniques, the actual conduct of the exercise varied. Most of
the cross-border banking groups in the sample were tested in a bottom-up
fashion, using internal models and granular portfolio data. Less complex
institutions were subject to a simplified stress test, based on reference
parameters provided by the ECB.
23. Regardless of the way the exercise was conducted, the supervisory
authorities discussed the results of the exercise with the banks involved,
6

CEBS Guidelines on stress testing currently available
(http://www.c-ebs.org/documents/Publications/Consultationpapers/2009/CP32/CP32.aspx).

13

as

consultation

paper

and, where appropriate, challenged the results, data, parameters, business
and other key assumptions used in the exercise, before submitting them to
the CEBS Secretariat.
24. Although there are some differences in the way the macro-economic
scenarios have been translated into the risk parameters and in the actual
application of the reference parameters provided by the ECB (see discussion
on the methodology in Section 4.1.2), all participants in the exercise made
efforts to ensure the consistency of internal parameters with the ECB
reference parameters leading of the overall consistency and comparability of
individual results. The actual parameters used have been discussed between
the institutions and their respective supervisors as a part of the consistency
checks.
25. In addition, parameters and overall results were analysed and submitted to
CEBS Secretariat for a challenging process with each of the participating
authorities. In order to further increase the overall consistency of the
approaches and methodologies used, especially for the exercises run directly
by banks, a special meeting between the respective participating authorities
and CEBS Secretariat has organised (“peer-review” meeting). At this
meeting parameters used in the exercise were discussed and commonly
analysed, in a way that did not compromise the confidentiality of individual
parameters and proprietary information 7 .

3.4 Relation of the exercise with proposed regulatory
changes
26. The timeline of the exercise coincided with the Quantitative Impact Study
(QIS) exercise conducted in Europe to gauge the impact of the December
2009 proposals from the Basel Committee on Banking Supervision (BCBS) 8
and the respective proposals for the review of the Directive 2006/48/EC
(CRD IV) under consideration by the European Commission outlined in its
public consultation launched on 26 February 2010 9 , which was coordinated
by CEBS for its European part.
27. However, it should be noted that the 2010 EU-wide stress test exercise is
completely separate from the QIS exercise and is by no means aimed at
duplicating or front-running QIS outcomes. Therefore the stress test has
been conducted based on the current regulatory regime, as well as on the
regulatory changes agreed before end February 2010. Policy options put
forward by the BCBS in December 2009 and the EU Commission in its
consultation on CRD IV were not considered. In effect, the regulatory
changes still under discussion will be implemented well after the period
covered by the stress test exercise, and will in all likelihood include
grandfathering provisions and phasing-in mechanisms.
7

Such collective analysis was already conducted as a part of the “lessons learnt” exercise
post 2009 stress test.
8
See: http://www.bis.org/press/p091217.htm
9
See:
http://ec.europa.eu/internal_market/consultations/docs/2010/crd4/consultation_paper_e
n.pdf

14

28. With respect to the regulatory changes introduced by the CRD II and III
amendments, elements related to the revision of the trading book and
securitisation requirements (treatment of re-securitisations in particular)
have been taken into account in the design and conduct of the exercise.

4. Scenarios and methodologies used in the
exercise
4.1 General features of the macro-economic scenarios
29. The exercise has been conducted using two sets of macro-economic
scenarios (benchmark and adverse), including a sovereign shock scenario
developed in close cooperation with the EU Commission and the ECB and
covering the period of 2010 - 2011. The adverse scenario for GDP,
cumulated over 2010-11, is around three percentage points lower than the
benchmark one for the European Union (EU27) and for the euro area as a
whole.
30. The benchmark scenario is mainly based on European Commission forecast
numbers that were available when work on the exercise began in March
2010, i.e., the Autumn 2009 European Economic Forecast (November 2009)
and the European Commission Interim Forecast (February 2010). This was
complemented with more up-to-date information on country forecasts in
cases of significant changes. Assumptions for market interest rates as well
as for exchange rates were set in line with the methods employed by the
European Commission to construct their forecast.
31. Under the benchmark scenario, the slow recovery initiated in 2010 is
expected to gain further momentum, with e.g. GDP growth for EU27
reaching 1.7% in 2011 after 1.0% in 2010 – largely in response to the
ongoing world trade pick-up. At the same time, unemployment remains high
– even increasing in a number of countries, owing to the lagged effects of
the past activity slowdown. Consumer price inflation is assumed to be
contained and stable overall, as the upswing occurs in economies where the
degree of slack is substantial. There is however a number of countries
where inflation declines or increases significantly – reflecting their cyclical
positions or fiscal policy measures.
32. Both sets of macro-economic scenarios were commonly agreed by all
participating authorities. Specific scenarios were provided for the domestic
situation of each of the EU Member States, Norway, the US, and the rest of
the world collectively. The scenarios also came with a set of reference risk
parameters (PDs and LGDs) proposed by the ECB. Some of the input
parameters and assumptions were also provided by CEBS and the
participating authorities outside of the narrative of the macro-economic
scenarios as provided by the ECB (evolution of real estate prices for 20
different geographies, notably – see Annex 3). It should be noted that in
coordination with national authorities banks were allowed to use more
conservative macro-economic assumptions, if this would realistically reflect
their specific risk exposure.

15

33. For the purposes of the market risk stress test, a set of stressed market
parameters has been applied to the trading book positions (Annex 5). It
should be noted that the parameters developed for the market risk stress
test are in-line with the macro-economic scenarios, and therefore could be
considered as directional, meaning that depending upon the size and
direction of their exposures, banks were able to make gains on certain
portfolios, thereby reducing the overall amount of stress coming from the
market parameters.
34. The following sub-sections introduce the features of the macro-economic
and sovereign shock scenarios used in the 2010 exercise, whereas Annex 2
provides greater details on the methodologies and techniques used to
develop the scenarios, reference risk parameters and haircuts on sovereign
debt holdings.

4.1.1

Features of the adverse macro-economic scenario

35. The adverse macro-economic scenario has two main features, a global
confidence shock, that affects demand worldwide, and an EU-specific shock
to the yield-curve, also originating from a postulated aggravation of the
sovereign debt crisis. The latter impact is differentiated across countries,
taking into account their respective situation.
36. The global confidence shock occurs in a context of downgraded employment
and profit expectations worldwide. It affects both private investment and
consumption, through a lasting downward shock to these variables,
cumulating overall to some 2 percentage points of GDP, concentrated over
the second half of 2010 and the first quarter of 2011. The EU is directly
affected by this confidence shock and by the effect on exports of the implied
lower world demand.
37. In addition, a common upward shift in the yield curve was applied for each
country in the EU (reaching 125 basis points for the three-month rates and
75 basis points for the 10-year rates at end-2011), supplemented with
country-specific upward shocks to long-term government bond yields
(overall amounting to 70 basis points at end-2011 for the euro area). The
rise in short-term rates reflects an assumption of tensions in the interbank
market – as was seen during earlier financial turmoil episodes (although not
as large as in the immediate aftermath of the Lehman episode, where shortterm inter-bank market spreads temporarily increased by close to 200 basis
points). The country-specific bond yield shock in turn accounts for
differentiated fiscal situations and related market perceptions.
38. The upward shift of the long-term rates can be associated to possible
concerns about the fiscal outlook in the EU.
39. The interest rate shock was assumed to persist over the whole exercise.
40. The results for the adverse macro-economic projections were obtained by
means of simulations. More details on the different macro-economic
scenarios, including country specific parameters, are provided in Annex 2 of
the report, whereas Table 1 below provides a snapshot of key macroeconomic variables of the both sets of the macro-economic scenarios.

16

41. In the adverse scenario, the value of the haircuts for valuation losses in the
trading book and of reference probabilities of default (PDs) and loss given
defaults (LGDs) change both on account of the changes in the macroeconomic scenario and of the introduction of the sovereign shock. On the
banking book, these shocks induce a change in PDs and LGDs for the
household and corporate sector, given that higher long-term government
bond yields also imply higher borrowing costs for the private sector, which
in turn imply higher PDs and LGDs for the non-sovereign exposures (see
Section 4.1.2).
Table 1. Evolution of aggregate key macro-economic variables in the scenarios

2008
EU27
GDP (y-o-y)
Unemployment (% of labour force)
Euro area
GDP (y-o-y)
Unemployment (% of labour force)
US
GDP (y-o-y)
Unemployment (% of labour force)

Realised
2009

2010 Q1

2010 Exercise
Benchmark
Adverse
2010
2011
2010
2011

0.7%
7.0%

-4.2%
8.9%

0.2%
9.6%

1.0%
9.8%

1.7%
9.7%

0.0%
10.5%

-0.4%
11.0%

0.6%
7.5%

-4.1%
9.4%

0.2%
10.0%

0.7%
10.7%

1.5%
10.9%

-0.2%
10.8%

-0.6%
11.5%

0.4%
5.8%

-2.4%
9.3%

0.7%
9.7%

2.2%
10.0%

2.0%
10.2%

1.5%
10.2%

0.6%
11.1%

Notes: GDP changes for realised is real GDP growth rate, Q1 2010 GDP growth is
compared to Q4 2009
Source: Eurostat for realised figures, stress test scenarios

42. As can be seen from the table, the benchmark scenario assumes a mild
recovery, with GDP increasing by 1.0% in 2010 and 1.7% in 2011 in the EU
(+0.7% in 2010 and +1.5% in 2011 for the Euro area). The adverse
scenario assumes a “double dip” situation, with an unchanged GDP in 2010
(0.0%) and decrease in GDP by 0.4% in 2011 for the EU27 (-0.2% in 2010
and -0.6% in 2011 for the Euro area).
43. GDP growth is particularly affected in the adverse scenario, and is lower
than in the benchmark scenario for all countries, on average by about one
percentage point in 2010 and by close to two percentage points in 2011.
The unemployment rate is higher, especially in 2011, while inflation is
significantly lower in 2011. . Given the deterioration of the macroeconomic
environment observed in 2009, the adverse scenario appears severe enough
and substantially below available forecasts and projections, thereby
corresponding to the materialisation of downside risks to economic growth
prospects.
44. Putting the macro-economic scenarios into the historical perspective (see
Chart 1) one can clearly see under the benchmark scenario a mild recovery
from the severe downturn of 2008-2009, whereas the adverse scenario
assumes a “double-dip” recession.

17

annual rates in %

Chart 1. Real GDP growth for EU27 and euro area under the benchmark and
adverse scenarios in comparison to historical developments

5.0 %
Actual

4.0 %

Stress test 
scenarios

3.0 %
2.0 %
1.0 %
0.0 %
‐1.0 %
‐2.0 %
‐3.0 %
‐4.0 %

EU16 Benchma rk

EU16 Advers e

EU27 Benchma rk

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

‐5.0 %

EU27 Advers e

Source:  Euros ta t up to 2009, 2010‐2011 s tres s  tes t s cena ri os

45. Overall, the comparison of realised GDP figures for 2008 and 2009 with the
scenarios used in the 2009 stress test exercise shows that realised figures
were in between the benchmark and adverse scenarios. Current trends for
2010 (based on Q1 2010 GDP growth figures) suggest that the macroeconomic situation remains relatively in line with the benchmark scenario,
especially for the larger economies.
46. The main feature of the sovereign shock scenario included the series of
valuation haircuts constructed to be applied to the outstanding trading book
exposures to European sovereign debt (see Table 2 below and Annex 2,
which provides details on the actual values of the valuation haircuts and the
way they have been constructed). Modelling sovereign risk based on market
yields implies that sovereign defaults are excluded from the exercise.

18

Table 2. Valuation haircuts to sovereign debt holdings as applied in the
exercise
Country
Haircut value
Austria
5.6%
Belgium
6.9%
Cyprus
6.7%
Finland
6.1%
France
6.0%
Germany
4.7%
Greece
23.1%
Ireland
12.8%
Italy
7.4%
Luxembourg
6.9%
Malta
6.4%
The Netherlands
5.2%
Portugal
14.1%
Slovakia
5.0%
Spain
12.0%
Slovenia
4.2%
Denmark
5.2%
Sweden
6.7%
UK
10.2%
Czech Republic
11.4%
Poland
12.3%
Other non-euro area EU c
11.8%
EU average
8.5%

Source: Stress test scenarios (see Annex 2 for more details)

47. Annex 4 provides more granular comparison between the realised values
and assumptions for key macro-economic variables used in the 2009 and
2010 EU-wide stress test exercises.

4.1.2

Translation of the macro-economic scenarios into
reference risk parameters and their application

48. The macro-economic scenario assumptions were translated, for the banking
book exposures (except for securitisation exposures, which were tested with
a separate methodology, see Section 4.4), into a set of risk parameters.
Reference PD and LGD parameters were projected by the ECB over the time
horizon of 2010 to 2011, consistently with both the benchmark and adverse
macro-economic scenarios. This in turn translated into impairment loss
estimates and risk weighted assets.
49. Reference PD and LGD parameters were computed at the country level for
five main portfolios (financial institutions, sovereign, corporate, consumer
credit and retail real estate). The details regarding the way the reference
parameters provided by the ECB were calculated are provided in Annex 2.
50. The use of reference risk parameters varied depending on the approach
institutions and supervisors chose to run the stress test. As mentioned
above, generally larger cross-border institution in the sample with access to
better modelling and risk quantification techniques used predominantly

19

bottom-up approaches. In such cases, macro-economic scenarios have been
translated using institutions’ models and impact on the institutions own risk
parameters have then been computed. Those internally computed impacts
(or multipliers) have been applied to the institutions’ internal parameters as
of end of 2009 in order to calculate the impact of the stress. The actual
parameters used have been reviewed by the respective national authorities,
compared with the reference parameters provided by the ECB, as all banks
were expected to use the same valuation haircuts as provided by the ECB
for debt exposures on the trading book, and discussed between all
participants at the “peer review” meeting.
51. In the case of top-down approaches largely used for a smaller and less
complex banks in the sample, national supervisory authorities used as a
starting point their own PD and LGD levels for 2009 and applied to these
values the relative changes of the parameters provided by the ECB in 2010
and 2011 with respect to their reference values in the respective year
according to the outcomes of the ECB models.

4.2 General features of the trading book stress test
52. Apart from the sovereign risk shock modelled via the valuation haircuts, the
exercise employed a detailed and granular set of market risk parameters
provided by CEBS (see Annex 5).
53. The set of parameters, consistent with the general direction of the macroeconomic scenarios, included assumptions on interest rates and volatilities
for major currencies (EUR, GBP, USD), exchange rates and volatilities for
the abovementioned currency pairs, haircuts and changes in volatility for
major equity commodity and debt instrument indices, changes in credit
spreads for debt instruments as well as bid/ask spreads to be used for the
assessment of the impact on the market liquidity. To highlight some of the
key features, the scenario assumed a drop in the value of major equity
indices by 10% under the benchmark and 20% under the adverse scenario.
for AFS exposures, these assumptions were translated into a haircut of 19%
(benchmark scenario) and 36% (adverse scenario) of equity exposures
during the two years.
54. For the computation of the impact from the market risk shocks, the
assumption was that instantaneous shocks (both benchmark and adverse),
applied to the positions as of 31 December 2009. The different
portfolios/books were stressed using the most appropriate parameters from
the set provided. For presentation purposes, the impact of the resulting
shock was distributed evenly between 2010 and 2011 results.
55. It should be noted that the parameters were in line with the macroeconomic scenarios and therefore could be considered as directional,
allowing for some compensation between gains and losses on different
portfolios.

20

4.4 Treatment of securitisation transactions
56. The exercise applied a specific approach to the treatment of securitisation
exposures in the banking book (securitisation exposures in the trading book
were stressed along the rest of trading exposures). All exposures
(traditional and synthetic, as well as liquidity lines on securitisation
transactions) for which there was significant risk transfer (in the meaning of
the CRD) were included in the scope of the exercise.
57. It should also be noted that the exercise took into account the forthcoming
changes in the CRD, notably proposals included in the draft CRD III in
relation to the treatment of re-securitisation. For the purpose of the exercise
it was assumed that the new amendments would be in place from 31
December 2010 onward.
58. The stress was designed as a deterioration of the credit quality of
securitisation exposures, expressed in changes of their credit rating (“denotching”). The application of the stress was adapted to take into account of
the different regulatory treatments applying to the exposures, depending on
whether they are treated under the standardised or the IRB approaches.
59. For the adverse scenario, the assumed reduction in credit quality of the
positions is equivalent to 4 external rating notches over two years. The
impact of such reduction translated into an increase of risk weighted assets.

4.5 Government support measures
60. The public support measures introduced in the course of the financial crisis
to support banks in difficulties as well as to maintain funding to the real
economy can be classified into four broad categories: (i) capital increases,
through equity shares or hybrid instruments provided by governments, (ii)
guarantees of banks’ assets provided by governments, (iii) guarantees of
liabilities or funding guarantees as well as liquidity provided by
governments, and (iv) liquidity support measures introduced by central
banks.
61. The exercise takes directly into consideration only the capital support
measures and asset guarantees received by the institutions in the sample
by 1 July 2010. As of 1 July 2010:
•

34 banks in the sample benefited from capital increases with a total
injected capital amounting to 169.6 bn €, making approximately 14% of
the total Tier 1 own funds of the banks in the sample.

•

20 banks in the sample benefited from asset guarantees.

62. The analysis suggests that the overwhelming majority of the government
support measures agreed between banks and governments has a useful life
extending beyond the horizon of the exercise.

21

5. Aggregate outcome of the exercise
5.1 Evolution and dispersion of capital ratios
63. The evolution of the aggregate Tier 1 capital ratio reflects the impact of the
macro-economic scenarios on regulatory capital charges for credit, market
and operational risk, as well as the impact of an additional sovereign shock
in the adverse scenario.
64. Chart 2 below represents the evolution of the aggregate Tier 1 capital ratio
(the weighted average of the sample of 91 banks) both under the
benchmark and adverse scenarios compared to the actual capital ratio at
the end of 2009. As can be seen from the chart, under the benchmark
scenario, the aggregate Tier 1 capital ratio grows significantly over the time
horizon of the exercise, largely due to the increased level of pre-impairment
income leading to capital generation through retained earnings.
65. Under the adverse scenario before the sovereign shock, the aggregate Tier
1 capital ratio decreases from 10.3% in 2009 to 9.6% by 2011. The
additional shock on sovereign risk puts a further downward pressure on the
aggregate Tier 1 capital ratio moving it further down to 9.2% by the end of
2011.

Chart 2. Evolution of the aggregate Tier 1 capital ratio
11.5 %
11.2 %

11.0 %

10.7 %

10.5 %

10.3 %

10.0 %

10.0 %
9.9 %

9.5 %

9.6 %
9.2 %

9.0 %
8.5 %
2009
Benchmark

2010
Adverse

2011

Adverse after sovereign shock

Source: Stres s  tes t ca l cul a ti ons

66. It should be noted that a fraction of the aggregate Tier 1 capital ratio is
attributable to continued reliance of some of the institutions in the sample
on government support, which attributes to approximately 1.2 percentage
point of the aggregate Tier 1 capital ratio. The effect of the government
support on various institutions subject to the support measures is quite
different and varies between 0.1 to 11.1 percentage points of the respective
individual Tier 1 capital ratios.

22

67. As previously mentioned, the nature and features of the overwhelming
majority of government support schemes, especially capital instruments
counted as regulatory capital, suggest that government support will be in
place beyond the time horizon of the exercise. Provided institutions in the
sample do meet the conditions set for receiving such support, it can be
assumed that government paid in capital cannot be simply withdrawn.
Another factor to keep in mind is that it is likely that a potential exit from
government support would not translate into “evaporation” of the solvency
of the institutions: for instance, as was already observed for some of the
banks in the sample, in case of repayment, it is likely that the capital
subscribed by government will be replaced by capital influx from private
investors.
68. Looking at the decomposition of the effect of different component on the
Tier 1 capital ratio under the adverse scenario (see Chart 3), one can see
that the aggregate ratio is driven up by the pre-impairment income leading
to the increase of ratio by 4.5 percentage point, offset by the same
proportion by impairment charges associated with the impact of the adverse
scenario after sovereign shock. The trading losses have a marginal impact
on the composition of the aggregate Tier 1 capital ratio, driving it down by
0.2 percentage points including the impact of 24.0 bn € stemming from the
application of an haircut on European sovereign debt holdings in the trading
book (see Section 4.1.1).

Chart 3. Contribution of different components to aggregate Tier 1 ratio under
adverse scenario, after sovereign shock
16.0%
14.0%

‐0,1%
12.0%

‐0,7%

10.3%
9.2%

10.0%

4.5%

‐4,5%

8.0%

‐0,2%

6.0%
4.0%
2.0%
0.0%
2009

Other impacts

Impairments

Pre‐
impairment
income
Source: CEBS stress test calculations

RWA

2011
Trading losses

69. The following group of charts provides information on the dispersion of
individual Tier 1 capital ratios differentiating between 26 major cross-border

23

banking groups (Chart 4) and the remaining 65 banks from the sample
(Chart 5). All dispersions are presented calculated for the benchmark and,
adverse scenario before and after the sovereign shock component and
presented by means of minimum, maximum, interquartile distribution and
median values.
70. As expected, generally the Tier 1 capital ratios for the major cross-border
banks are more compactly distributed compared to the smaller banks, and
all of the 26 major cross-border banks would stay above the 6% Tier 1
capital ratio set as a benchmark threshold for the purposes of this exercise,
with a lowest being of 7.4% after the impact of the adverse scenario
including the sovereign shock.

Chart 4. Evolution of aggregate Tier 1 capital
ratio for 26 major cross-border
banking groups

Chart 5. Evolution of aggregate Tier 1 capital
ratio for 65 smaller banks
(min, max, interquartile distribution, median)
22%
20%
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%

(min, max, interquartile distribution, median)
22%
20%
18%
16%
14%
12%
10%
8%
6%
4%
2%
0%

201
0
201
0
2009

201
1

201
0

201
1

201
0

201
1

Benchmark

Benchmark

Adverse

Adverse

After
sovereign
shock

2009

201
1

201
0

201
1

201
0

201
1

Benchmark

Benchmark

Adverse

Adverse

After
sovereign
shock

After
sovereign
shock

After
sovereign
shock

Source: Stress test calculations
Source: Stress test calculations

71. Under the benchmark scenario none of the smaller banks would see their
Tier 1 capital ratios fall below 6% (it should be noted that a number of
banks in the sample have been subject to restructuring and mergers in the
first half of 2010, which explains significantly lower capital ratios in 2009
compared to 2010 and 2011).
72. The adverse scenario before and after the sovereign shock is seen having
significant impact on the individual Tier 1 ratios pushing the interquartile
distribution lower and leading to some banks starting to fall below the 6%
threshold. Thus under the adverse scenario before sovereign shock one
bank would see in 2010 its Tier 1 capital ratio below the 6% threshold,
reaching 5.5%, whereas by 2011 already 5 banks would fall below the
threshold with the lowest Tier 1 capital ratio being 4.5%, below the
threshold but still above the CRD minimum of 4.0% for the Tier 1 capital
ratio.
73. The further sovereign shock drives the distribution lower with some minor
widening towards the lower end and bringing the number of institutions
failing to meet the 6% threshold in 2011 to 7 with the lowest Tier 1 capital

24

ratio falling to 3.9%, which would be below the regulatory minimum were
this extreme situation to materialise.
74. Overall, the results can be considered rather reassuring for the banks in the
sample. At the same time, for some of the banks these results are partially
due to the continuing reliance on government support for a number of
banks.

5.2 Evolution of the main components of the aggregate
Tier 1 capital ratio
75. The movements in the aggregate Tier 1 capital ratio discussed above are
explained by the different impacts of the macro-economic scenarios, on the
components of the ratio, as well as by methodological assumptions used in
the exercise.
76. Chart 6 depicts the evolution of the aggregate Tier 1 own funds (i.e. the
numerator of the Tier 1 capital ratio) in the exercise, compared to the actual
amount of capital in 2009. As can be seen from the chart, the aggregate
amount of capital increases under the benchmark scenario, supported by
positive net results of the banks in the sample, whereas under the adverse
scenario, the aggregate amount of Tier 1 own funds remains almost
constant at the level of 2009, and as a result of the sovereign shock
component falls to its lowest level in the exercise at 1 118bn €.
Chart 6. Evolution of aggregate Tier 1 own funds

Bn EUR

1300

1277 

1250
1200

1213 
1166 

1162 

1163 

1159 

1150

1118 

1100
1050
1000
2010
2009

2011

2010

2011

Benchmark

Benchmark

Adverse

Adverse

Source: CEBS Stress test calculations

2010

2011

Adverse after Adverse after
sovereign
sovereign
shock
shock

77. As mentioned above, the aggregate Tier 1 own funds do incorporate a
significant amount government support (169.6 bn €), which remains
constant through the time horizon of the exercise.

25

78. Looking at the banks failing to meet the exercise threshold of 6% Tier 1
capital adequacy as a results of the macro-economic and sovereign shock, 7
banks have an overall shortfall of 3.4 bn €, compared to the 388.4.6 bn € of
surplus capital (above the 6% threshold) for the entire sample of 91 banks.
79. The other component of the Tier 1 ratio – risk weighted assets (RWA –the
denominator of the capital adequacy ratio-), does show rather significant
evolutions, especially in the adverse scenario. Under the benchmark
scenario, the RWA remain relatively stable compared to 2009 with an annual
increase of 0.7% in 2010 and 0.3% in 2011. However, under the adverse
scenario, the total RWA of the 91 banks in the sample increase significantly
by 7.6% in 2011 compared to 2009 (see Chart 7).

Bn EUR

Chart 7. Evolution of aggregate Risk Weighted Assets
12400
12153 

12200
12000
11800

11673 

11600
11291 

11370 

11407 

2010

11400

2011

11200
11000
10800

2009

Benchmark

2010

2011

Benchmark Adverse after Adverse after
sovereign
sovereign
shock
shock

Source: CEBS Stress test calculations

80. The increase of RWA is despite the fact that the exercise assumed a “zero
growth” assumption for the different exposures in the exercise, and also
recognises all deleveraging programmes formally accepted or announced by
1 July 2010, including restructuring plans agreed between a number of
banks and the EU Commission under the State Aid review procedures. Both
of these factors had a reducing impact on the total amount of RWA.
81. As a matter of fact, this increase in RWA is almost fully attributable to
“mechanical” effects, i.e. changes of RWA for credit portfolios subject to the
calculation of the regulatory capital using internal ratings based (IRB)
approaches (i.e. increase in expected losses), as well as to the methodology
retained for the stress of securitisation exposures. As a result, the
evolutions of RWA observed both under the benchmark and adverse
scenario, including the sovereign shock, should be interpreted with caution
and cannot be directly interpreted as the potential lending capacity of the
institutions and lending trends in general.

26

82. It should be noted that the impact of the sovereign shock component is
rather limited on the size of the risk weighted assets as it major part is
attributed to impairments estimated in the banking book and valuation
losses in the trading book.

5.3 Evolution of financial results and total losses from
the stress
83. Chart 8 provides an overview of the level of aggregate pre-impairment
income (operating income less operating costs) of the banks in the sample.
Consistent with the macro-economic scenario, under the benchmark
scenario pre-impairment income reduces in 2010 and returns in 2011 to a
slightly higher level than in 2009, reaching 277 bn €. Under the adverse
scenario, banks in the sample see their pre-impairment income decreasing
by 7.0% in 2010 compared to 2009 and rising to a level of 258 bn € at the
end of 2011, which is 4.4% lower compared to level of 2009 .

Bn EUR

Chart 8. Evolution of aggregate pre-impairment income

280
275

277 
270 

270
265

261 

258 

260
255

251 

250
245
240
235
2010
2009

2011

Benchmark

Benchmark

Source: CEBS Stress test calculations

2010

2011

Adverse
Adverse
before sovereign before sovereign
shock
shock

84. It should be noted that operating profits assumptions incorporated in the
exercise were challenged by the respective national supervisory authorities.
Following the experience of the 2009 exercise, supervisory authorities are
usually confident that operating profits estimated for the different banks are
consistent with both the environment in which the banks operate, and with
past and recent trends.
85. Notably, the aggregate stability of net interest income, and net fees and
commissions leading to somewhat positive outlook regarding the evolution

27

of pre-impairment income, could be explained by the following factors,
among others:
•

The fact that the macro scenarios assumed differentiated shocks across
countries. Therefore, depending on the geographic distribution of banks
businesses, the P&L is diversely impacted, especially for business lines
outside the EU area covered by the macro-economic scenarios,
effectively meaning that banks and/or supervisors were able to use their
own assumptions regarding the outlooks for such geographies;

•

The fact that the assumption of an increase and a flattening of the yield
curve may have immediate positive effects on earnings, especially for
institutions operating in a variable rate environment for their retail
business, which allows passing on to customers most of the increases in
interest rates;

•

The “zero growth” assumption built in the exercise, which may run
contrary to deleveraging intentions expressed by banks.

86. In that respect, Section 6 of the report provides information in terms of
backtesting of the outcomes of the stress test calculations against the
realised figures, as well as an assessment of current trends, based on the
banks in the sample that published detailed results for the first quarter of
2010.
87. As regards the major cross-border banks, CEBS will continue to monitor the
evolution of banks P&L, also attributed to massive deleveraging
programmes conducted by the banks over the last years and rising
uncertainty of the sovereign debt market outlook, in the context of its
regular micro-prudential assessments of risks and vulnerabilities facing the
EU banking sector.
88. Focusing on the aggregate level of impairments (see Chart 9), one can see a
decrease of impairment charges under the benchmark scenario compared to
the level of 2009, which is consistent with the assumption of a mild
recovery.
89. Under the adverse scenario, on the contrary, following the assumed
deterioration of the operating environment, the impairment charges rise
under the adverse scenario before sovereign shock to a total of 472.8 bn €
over 2 years (2010 and 2011). Taking into account a further impact of the
sovereign shock on impairments in the banking book (38.9 bn €), total
impairments reach a total of 511.8 bn € over 2 years.

28

Chart 9. Evolution of aggregate impairment charges, including AFS equity and
excluding trading losses

300

Bn EUR

267 
234 

250

239 

245 

2010

206 
200

177 
152 

150
100
50

2010
2009

2011

2010

2011

Benchmark

Benchmark

Adverse

Adverse

Source: CEBS Stress test calculations

2011

Adverse after Adverse after
sovereign
sovereign
shock
shock

90. The impact of the trading losses before sovereign shock can be considered
as limited, on aggregate leading to 25.9 bn € under the adverse scenario
before sovereign shock. The application of haircuts to the sovereign
exposures in the trading book (28.2 bn €) leads to further aggregate losses
under the adverse scenario, with a total of 54.1 bn € losses in the trading
book over two years under the adverse scenario after sovereign shock.

5.4 Total loss estimates from the stress and observed
loss rates
91. This section provides a summary of total loss estimates coming from the
application of macro-economic scenarios, including sovereign shock, as
opposed to the level of pre-impairment income generated by the sample of
institutions in the macro-economic environment set by the scenarios. The
discussion is further supported by the analysis of the evolution of loss rates.
92. As can be seen from Chart 10, as a result of the adverse scenario including
the sovereign shock, total estimated losses of 91 banks in 2010 would reach
258.0 bn € as opposed to 250.3 bn € of pre-impairment income estimated
under this scenario for 2010. In 2011, total losses would reach 307.8 bn €
compared to 257.7 bn € of pre-impairment income. Over the two-year time
horizon, total cumulative losses from the stress test would reach 565.9 bn €

29

Bn EUR

Chart 10. Contribution of different components to aggregate Tier 1 ratio under
adverse scenario, including sovereign shock

350
300
250

302

258
13

251

28
27

11

258

13

200
150
100

239

234

50

Impairments +
trading losses

Pre‐impairment
income

Impairments +
tradin losses

2010

Pre‐impairment
income
2011

Addi ti ona l  tra di ng l os s es  res ul ti ng from s overei gn s hock
Tra di ng l os s es
Addi ti ona l  i mpa i rments  res ul ti ng from s overei gn s hock
Impa i rments

Source: CEBS stress test calculations

93. In order to better understand the impact of the stress scenarios one needs
to assess the loss rates representing the share of impairment losses to total
corporate and retail exposures in available-for-sale, held-to-maturity and
loans and receivables portfolios.
94. Chart 12 below presents the distribution of loss rates in corporate portfolios
under the benchmark and adverse scenarios, compared to the actual loss
rates of 2009 and Chart 13 provides the same information for the retail
portfolios.

30

Chart 11. Distribution of corporate loss rates

Chart 12. Distribution of retail loss rates
(min, max, interquartile distribution, median)

(min, max, interquartile distribution, median)
14%

8%
7%
6%
5%
4%
3%
2%
1%
0%
-1%
-2%

12%
10%
8%
6%
4%
2%
0%
-2%
201
0
2009

201
1

201
0

201
1

201
0

201
1

Benchmark

Benchmark

Adverse

Adverse

After
sovereign
shock

After
sovereign
shock

201
0
2009

201
1

201
0

201
1

201
0

201
1

Benchmark

Benchmark

Adverse

Adverse

After
sovereign
shock

After
sovereign
shock

Source: Stress test calculations

Source: Stress test calculations

95. As can be noted from the chart the distribution of corporate loss rates
significantly widens under the adverse and post sovereign shock scenarios
compared to the original level of 2009.
96. As regards the loss rates on retail exposures as expected, on average they
are lower compared to the corporate related losses with a mean loss rate
varying between 0.8% in 2009 and 1.3% in 2011 after the sovereign shock
add-on.
97. Chart 13 presents the evolution of the aggregate loss rates computed for
the adverse scenario, putting them into the historical perspective, The Chart
demonstrate the severity of the exercise, when the 2010 and 2011
estimates are compared with the situation as of end 2009, where the GDP
growth was largely negative (-4.2%).

31

Chart 13. Evolution of historical average loss rates compared to loss rates
observed in the stress test

2.5 %
Actua l

Stres s  tes t outcomes

2.01%
2.0 %

1.5 %

1.63%

1.0 %

0.5 %

0.0 %
2004
Adverse

2005

2006

2007

2008

2009

2010

2011

Adverse after sovereign shock

Source: Publ i c di s cl os ures  up to 2009, 2010‐2011 Stres s  tes t ca l cul a ti ons

6. Comparison of key financial indicators with
realised figures
98. In this section CEBS provides comparisons between the results of the stress
test with realised figures observed in 2007-2009 and estimation for 2010
based on the published results for the first quarter 2010. The comparison is
provided for a sub-sample of 50 banks, for which quarterly financial
information is available in the public domain.
99. As regards pre-impairment income (see Chart 15), the realised figure in
2009 was at the same level as the annualised trend for 2010 (based on Q1
2010 data), as well as the result of the benchmark scenario for 2010.

32

Chart 14.

Comparison of pre-impairment income

bn € 

ST exercise 2010

Estimated

Realised

200 
180 
160 
140 
120 
100 
80 
60 
40 
20 
2007

2008

2009

2010

2010

2011

2010

Realised Realised Realised EstimatedBenchmark
Benchmark Adverse

2011
Adverse

Source: Stres s  tes t ca l cul a ti ons . Sa mpl e of 50 ba nks , 

100. The level of pre-impairment income under the adverse scenarios stays very
close to the realised figures for 2009, which suggests the relative resilience
of banks’ income to shocks (see the developments in paragraph 85).
101. Looking at the comparison of impairment charges made for the same subsample of 50 banks (see Chart 16) one can see that after a peak reached in
2009, the trend of impairment charges, based on first quarter figures, tends
to drop in 2010. The impairment charges estimated under the benchmark
scenario for 2010 remain higher than the realised trend observed for 2010.
Impairment charges under the adverse scenario are higher than the peak
observed in 2009.

33

Chart 15.

Comparison of impairment charges, including sovereign shocks

bn € 

Realised

Estimated

ST exercise 2010

160 
140 
120 
100 
80 
60 
40 
20 

2007

2008

2009

2010

2010

2011

2010

Realised Realised Realised EstimatedBenchmark enchmark Adverse
B

2011
Adverse

Source: Stres s  tes t ca l cul a ti ons . Sa mpl e of 50 ba nks .

102. These charts suggest no immediate concern on the aggregate level of
impairment and profit assumptions embedded in the results of the exercise.
However, given its tasks and responsibilities, CEBS will continue monitoring
the developments for the sub-set of major cross-border banking groups in
that respect, as part of its regular micro-prudential risk assessment.

7. General follow-up actions
103. The 2010 EU-wide stress test exercise coordinated by CEBS and conducted
in cooperation with the ECB, the EU commission and the national
supervisory authorities of 20 Member States is the second coordinated
stress test exercise in Europe, involving a sample of 91 major cross-border
banking groups and domestic credit institutions covering at least 50% of
total assets in every EU Member State, on a consolidated basis.
104. There are significant differences in the size, complexity and risk profiles of
the institutions in the sample, as well as in the extent of their reliance on
government support measures. Against this background, the aggregate
results cannot be necessarily directly extrapolated to the individual
conditions of institutions in the sample, nor to the general financial
conditions of all banks operating in Europe.
105. CEBS supports, in particular, the greater transparency of this exercise,
given the specific market circumstances under which banks currently
operate. We therefore welcome the publication of banks’ individual results,
particularly their respective capital positions and loss estimates under an
adverse scenario, as well as detailed information on banks’ exposures to

34

EU/EEA central and local government debt. Such disclosures ensure
transparency regarding conditions in the EU banking sector.
106. Based on the results of the EU-wide stress testing exercise, 7 institutions
would fall bellow the threshold of 6% of the Tier 1 capital adequacy ratio
under the adverse scenario including sovereign shock. The 6% threshold
has been set up exclusively for the purpose of this exercise.
107. With respect to the situation of individual institutions that fail to meet the
threshold for this stress test exercise, the competent national authorities are
in close contact with the bank in question to assess the results of the test
and their implications, in particular any potential need for recapitalization.
108. The banks are expected to propose a plan to address the weaknesses that
have been revealed by the stress test. The plan will have to be implemented
within a given period of time, in agreement with the supervisory authority.
109. More information on the individual outcomes and follow-up actions, where
necessary, is provided by banks and/or national supervisory authorities
participating in the exercise.
110. CEBS regards this exercise as a very positive development and a move
towards greater convergence of supervisory stress testing practices. Part of
the mandate of CEBS is to undertake on a periodic basis these EU-wide
stress testing exercises. CEBS will continue with testing the resilience of the
EU banking sector by means of periodic EU wide and thematic risk
assessments. It will also continue its work on improving convergence in
supervisory practices across Europe by addressing the topics both from a
policy and practical perspective.
111. It should be noted that CEBS has recently issued draft Guidelines on stress
testing 10 and risk assessment 11 , which aim at further enhancing the
convergence of supervisory approaches in both areas.

CEBS Guidelines on stress testing currently available as consultation
(http://www.c-ebs.org/documents/Publications/Consultationpapers/2009/CP32/CP32.aspx), final text will be published later in August 2010.
10

11

paper

Guidelines for the joint assessment of the elements covered by the supervisory review
and evaluation process and the joint decision regarding the capital adequacy of cross
border groups (CP39) currently available as a consultation paper (see: http://www.cebs.org/documents/Publications/Consultation-papers/2010/CP39/CP39.aspx) and will be
finalised toward the end of 2010.

35

Annex 1. List of banks covered by the 2010 EUwide stress test exercise
Country
Austria
Belgium
Cyprus

Name of the institution
ERSTE GROUP BANK AG
RAIFFEISEN ZENTRALBANK OESTERRREICH AG (RZB)
KBC BANK NV
DEXIA
MARFIN POPULAR BANK PUBLIC CO LTD
BANK OF CYPRUS PUBLIC CO LTD
DANSKE BANK

Denmark
Finland
France

Germany

Greece

Hungary
Ireland

Italy

Luxembourg
Malta

JYSKE BANK
SYDBANK
OP-POHJOLA GROUP
BNP PARIBAS
CREDIT AGRICOLE GROUP
BPCE GROUP
SOCIETE GENERALE
DEUTSCHE BANK AG
COMMERZBANK AG
HYPO REAL ESTATE HOLDING AG
LANDESBANK BADEN-WÜRTTEMBERG
BAYERISCHE LANDESBANK
DZ BANK AG DT. ZENTRAL-GENOSSENSCHAFTSBANK
NORDDEUTSCHE LANDESBANK -GZDEUTSCHE POSTBANK AG
WESTLB AG
HSH NORDBANK AG
LANDESBANK HESSEN-THÜRINGEN GZ
LANDESBANK BERLIN AG
DEKABANK DEUTSCHE GIROZENTRALE
WGZ BANK AG WESTDT. GENO. ZENTRALBK
NATIONAL BANK OF GREECE
EFG EUROBANK ERGASIAS S.A.
ALPHA BANK
PIRAEUS BANK GROUP
AGRICULTURAL BANK OF GREECE S.A. (ATEbank)
TT HELLENIC POSTBANK S.A.
OTP BANK NYRT.
FHB JELZÁLOGBANK NYILVÁNOSAN MŰKÖDŐ RT
BANK OF IRELAND
ALLIED IRISH BANKS PLC
UNICREDIT
INTESA SANPAOLO
MONTE DEI PASCHI DI SIENA
BANCO POPOLARE - S.C.
UNIONE DI BANCHE ITALIANE SCPA (UBI BANCA)
BANQUE ET CAISSE D'EPARGNE DE L'ETAT
BANQUE RAIFFEISEN
BANK OF VALLETTA (BOV)

36

Country

Netherlands

Poland

Portugal

Slovenia
Spain

Name of the institution
ING Bank
RABOBANK GROUP
ABN/ FORTIS BANK NEDERLAND (HOLDING) N.V
SNS BANK
POWSZECHNA KASA OSZCZĘDNOŚCI BANK POLSKI S.A. (PKO BANK
POLSKI)
CAIXA GERAL DE DEPÓSITOS
BANCO COMERCIAL PORTUGUÊS BANCO COMERCIAL PORTUGUÊS
S.A. (BCP OR MILLENNIUM BCP)
ESPÍRITO SANTO FINANCIAL GROUP S.A. (ESFG)
BANCO BPI
NOVA LJUBLJANSKA BANKA (NLB)
GRUPO SANTANDER
GRUPO BBVA
JUPITER:
CAJA DE AHORROS Y MONTE DE PIEDAD DE MADRID (CAJA
MADRID); CAJA DE AHORROS DE VALENCIA, CASTELLÓN Y
ALICANTE (BANCAJA); CAIXA DÉSTALVIS LAIETANA; CAJA INSULAR
DE AHORROS DE CANARIAS; CAJA DE AHORROS Y MONTE DE
PIEDAD DE AVILA; CAJA DE AHORROS Y MONTE DE PIEDAD DE
SEGOVIA; CAJA DE AHORROS DE LA RIOJA.
CAIXA:
CAJA DE AHORROS Y PENSIONES DE BARCELONA (LA CAIXA);
CAIXA DÉSTALVIS DE GIRONA.
BASE:
CAJA DE AHORROS DEL MEDITERRÁNEO (CAM); CAJA DE AHORROS
DE ASTURIAS; CAJA DE AHORROS DE SANTANDER Y CANTABRIA;
CAJA DE AHORROS Y MONTE DE PIEDAD DE EXTREMADURA.
BANCO POPULAR ESPAÑOL, S.A.
BANCO DE SABADELL, S.A.
DIADA:
CAIXA DÉSTALVIS DE CATALUNYA; CAIXA DÉSTALVIS DE
TARRAGONA: CAIXA DÉSTALVIS DE MANRESA.
BREOGAN:
CAJA DE AHORROS DE GALICIA; CAIXA DE AFORROS DE VIGO,
OURENSE E PONTEVEDRA (CAIXANOVA).
MARE NOSTRUM:
CAJA DE AHORROS DE MURCIA; CAIXA DÉSTALVIS DEL PENEDES;
CAJA DE AHORROS Y MONTE DE PIEDAD DE LAS BALEARES (SA
NOSTRA); CAJA GENERAL DE AHORROS DE GRANADA.
BANKINTER, S.A.
ESPIGA:
CAJA DE AHORROS DE SALAMANCA Y SORIA (CAJA DUERO); CAJA
DE ESPAÑA DE INVERSIONES CAJA DE AHORROS Y MONTE DE
PIEDAD (CAJA ESPAÑA).
BANCA CIVICA:
CAJA DE AHORROS Y M.P. DE NAVARRA, CAJA DE AHORROS
MUNICIPAL DE BURGOS Y CAJA GENERAL DE AHORROS DE
CANARIAS.
CAJA DE AHORROS Y MONTE DE PIEDAD DE ZARAGOZA, ARAGON Y
RIOJA (IBERCAJA).
M.P. Y C.A. DE RONDA, CADIZ, ALMERIA, MALAGA, ANTEQUERA Y
JAEN (UNICAJA)
BANCO PASTOR, S.A.
CAJA SOL:
MONTE DE PIEDAD Y CAJA DE AHORROS SAN FERNANDO DE

37

Country

Sweden

UK

Name of the institution
HUELVA, JEREZ Y SEVILLA (CAJA SOL); CAJA DE AHORRO
PROVINCIAL DE GUADALAJARA.
BILBAO BIZKAIA KUTXA,AURREZKI KUTXA ETA BAHITETXEA
UNNIM:
CAIXA DÉSTALVIS DE SABADELL; CAIXA DÉSTALVIS DE TERRASSA;
CAIXA DÉSTALVIS COMARCAL DE MANLLEU.
CAJA DE AHORROS Y MONTE DE PIEDAD DE GIPUZKOA Y SAN
SEBASTIAN (KUTXA).
CAJA3:
CAJA DE AHORROS Y MONTE DE PIEDAD DEL CÍRCULO CATÓLICO
DE OBREOS DE BURGOS (CAJA CÍRCULO); MONTE DE PIEDAD Y
CAJA GENERAL DE AHORROS DE BADAJOZ; CAJA DE AHORROS DE
LA INMACULADA DE ARAGÓN.
CAJA DE AHORROS Y MONTE DE PIEDAD DE CORDOBA (CAJASUR).
BANCA MARCH, S.A.
BANCO GUIPUZCOANO, S.A.
CAJA DE AHORROS DE VITORIA Y ALAVA (CAJA VITAL KUTXA).
CAJA DE AHORROS Y MONTE DE PIEDAD DE ONTINYENT.
COLONYA - CAIXA D'ESTALVIS DE POLLENSA.
NORDEA BANK
SKANDINAVISKA ENSKILDA BANKEN AB (SEB)
SVENSKA HANDELSBANKEN
SWEDBANK
ROYAL BANK OF SCOTLAND (RBS)
HSBC HOLDINGS PLC
BARCLAYS
LLOYDS BANKING GROUP

38

Annex 2. ECB Technical note on the macroeconomic scenarios and reference risk parameters

SEC/GenC/X/10/77.final.R - SEC/GovC/X/10/289.final.R

23 July 2010

TECHNICAL NOTE ON THE MACROECONOMIC SCENARIOS
AND REFERENCE RISK PARAMETERS

INTRODUCTION
This note presents the main technical features of the stress test exercise that has been conducted by the
CEBS and national supervisory authorities, in cooperation with the ECB. ECB staff provided the
macroeconomic scenarios (benchmark and adverse) and the corresponding key micro parameters
(probabilities of default (PDs), loss given default (LGDs), and haircuts for holdings of government
bonds in the trading book). The changes in these parameters under the adverse scenario represent a
substantial stress for the European banks.
The “benchmark” scenario is on average not very far from currently available forecasts, while the
“adverse” one, taking stock of prevailing tail risks – especially related to the sovereign debt situation –
is in turn substantially below these forecasts.
In addition, the severity of the stress arises from the combination of the increase in the haircuts and
especially from the increase in the PDs and LGDs under the adverse scenario. The reference haircuts
were computed from changes in the prices of 5-year sovereign bonds. The impact of the increase in the
haircuts on government debt in the trading book is mitigated by the fact that banks’ holdings of
government securities are primarily in the banking book, and the average maturity of these securities is
only around 5 years. On the other hand, the increase in PDs and LGDs affects all portfolios in the

39

banking book and is substantial. For instance, comparing the end-2009 values with those under the
adverse scenario in 2011, PDs of corporate assets double or triple in some countries, while for the euro
area they increase by over 61%, on average.

40

1. The Macroeconomic Scenarios
For the purpose of conducting the stress-test exercise, two macroeconomic scenarios, covering
the period 2010-11, were developed: a “benchmark” scenario (see Table 1), and an “adverse”
scenario (see Table 2), taking stock of prevailing tail risks, especially related to the sovereign
debt situation. The adverse scenario GDP, cumulated over 2010-11, is close to three percentage
points lower than the benchmark one for the European Union (EU) and for the euro area as a
whole.
The benchmark scenario is mainly based on European Commission (EC) forecast numbers
that were available when work on the CEBS exercise began in March 2010, i.e., the Autumn
2009 European Economic Forecast (November 2009) and the EC Interim Forecast (February
2010). This was complemented with more up-to-date information on country forecasts in cases
of significant changes. Assumptions for market interest rates as well as for exchange rates were
set in line with the methods employed by the EC to construct their forecast.
In this scenario, the slow recovery initiated in 2010 is expected to gain further momentum, with
e.g. GDP growth for the euro area reaching 1.5% in 2011 after 0.7% in 2010 – largely in
response to the ongoing world trade pick-up. At the same time, unemployment remains high –
even increasing in a number of countries, owing to the lagged effects of the past activity
slowdown. Consumer price inflation is assumed to be contained and stable overall, as the
upswing occurs in economies where the degree of slack is substantial. There are however a
number of countries where inflation declines or increases significantly – reflecting their
cyclical positions or fiscal policy measures.
The scenario involves somewhat more contained dynamics in 2010, while by contrast it
appears generally on the upside for 2011. On balance over the two years, differences with
currently available forecasts are limited.
The adverse scenario 12 has two main features, a global confidence shock, that affects demand
worldwide, and an EU-specific shock to the yield-curve, originating from a postulated aggravation of
the sovereign debt crisis. The latter impact is differentiated across countries, taking into account their
respective situation.

12

In all tables in this note, the adverse scenario includes the sovereign risk.

41

The global confidence shock occurs in a context of downgraded employment and profit expectations
worldwide. It affects both private investment and consumption, through a lasting downward shock to
these variables, cumulating overall to some 2 percentage points of GDP points over the horizon,
concentrated over the second half of 2010 and the first quarter of 2011. The EU is directly affected by
this confidence shock and by the effect on exports of the implied lower world demand.
In addition, related to prevailing sovereign debt risks, a common upward shift in the yield curve was
applied for each country in the EU (reaching 125 basis points for the three-month rates and 75 basis
points for the 10-year rates at end-2011), supplemented with country-specific upward shocks to longterm government bond yields (overall amounting to 70 basis points at end-2011 for the euro area). The
rise in short-term rates reflects an assumption of tensions in the interbank market – as was seen during
earlier financial turmoil episodes. The country-specific bond yield shock in turn accounts for
differentiated fiscal situations and related market perceptions.
Accordingly, the distribution of the country-specific upward shock to long-term interest rates across
countries reflects two elements. First, a widening of spreads in line with market developments since the
beginning of May 2010. Second, an additional widening of spreads reflecting an average additional
increase of 30 basis points. Its impact on each country’s long-term bond yields was determined in
proportion to the volatility of 10-year sovereign bond spreads that was observed between December
2009 and June 2010. Taken together, the country-specific shock implies an additional average increase
of 70 basis points (see Table 3). To underline the importance of the combined shocks affecting interest
rates, it is worthwhile to mention that, for example, they result in 2011 in 10-year government bond
yields of 4.7% for Germany and 14.7% for Greece (see Table 7).

The macroeconomic effects of these assumptions were calibrated using econometric models,
also taking into account trade spillovers across EU countries. GDP growth is particularly
affected in the adverse scenario, and is lower than in the benchmark scenario for all countries,
on average by about one percentage point in 2010 and by close to two percentage points in
2011. The unemployment rate is higher, especially in 2011 (e.g. by 0.6 percentage point in the
euro area), while inflation is significantly lower in 2011 (e.g. by 0.4 percentage point for the
euro area). The adverse scenario generally appears to be substantially below available forecasts
and projections, thereby corresponding to the materialisation of downside risks to economic
growth prospects.

42

Table 1: Macroeconomic scenarios – benchmark scenarios
2010 Benchmark
Austria
Belgium
Cyprus
Finland
France
Germany
Greece
Ireland
Italy
Luxembourg
Malta
Netherlands
Portugal
Slovakia
Slovenia
Spain
Euro area
Bulgaria
Czech R.
Denmark
Estonia
Hungary
Latvia
Lithuania
Poland
Romania
Sweden
UK
Rest of the EU
2011 Benchmark
Austria
Belgium
Cyprus
Finland
France
Germany
Greece
Ireland
Italy
Luxembourg
Malta
Netherlands
Portugal
Slovakia
Slovenia
Spain
Euro area
Bulgaria
Czech R.
Denmark
Estonia
Hungary
Latvia
Lithuania
Poland
Romania
Sweden
UK
Rest of the EU

GDP at constant
prices
1.1
0.6
0.1
0.9
1.2
1.2
-4.1
-1.4
0.7
1.1
0.7
0.9
0.5
1.9
1.3
-0.6
0.7
0.4
1.4
1.5
1.0
0.9
-3.3
0.5
2.9
-0.7
1.4
0.6
1.0
GDP at constant
prices
1.5
1.5
1.3
1.6
1.5
1.7
-2.6
2.6
1.4
1.8
1.6
1.6
0.2
2.6
2.0
1.0
1.5
4.0
1.8
1.8
4.0
3.2
3.9
3.1
2.4
3.6
2.1
1.9
2.2

Unemployment
6.0
9.9
6.6
10.2
10.2
9.2
11.7
14.0
8.7
7.3
7.4
5.4
11.1
12.8
8.3
20.0
10.7
8.8
8.1
5.8
16.0
11.8
20.4
17.1
10.4
8.1
10.2
8.7
9.2
Unemployment
5.7
10.3
6.7
9.9
10.0
9.3
14.1
13.2
8.7
7.7
7.3
6.0
11.9
12.6
8.5
20.5
10.9
8.0
8.5
5.6
14.5
11.9
18.2
15.9
11.5
8.8
10.1
8.0
8.9

Short-term
interest rates
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2
1.2

2.1

4.8
1.4
1.5
Short-term
interest rates
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1

2.9

5.7
2.8
3.0

Long-term
interest rates
4.0
4.0
4.7
3.5
3.8
3.5
6.8
5.1
4.4
3.8
4.5
3.8
4.7
4.1
3.9
4.4
3.5
6.9
4.7
3.8
12.1
8.4
12.7
12.1
6.3
9.4
3.6
4.3

Nominal USD
exchange rate
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
1.4
18.7
5.0
11.5
196.5
0.5
2.5
2.9
3.0
7.0
0.6

Long-term
interest rates
4.3
4.4
5.1
3.9
4.1
3.8
7.1
5.4
4.7
4.2
4.9
4.1
5.1
4.6
4.4
4.7
3.8
6.9
4.4
4.1
12.1
6.2
12.7
12.1
6.3
9.4
3.9
4.7

Nominal USD
exchange rate
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
1.5
18.8
5.0
11.6
197.2
0.5
2.6
2.9
3.1
7.0
0.6

CPI
1.3
1.3
3.1
1.6
1.2
0.7
1.4
-0.6
1.7
1.8
2.0
0.8
1.3
1.9
1.7
1.1
1.1
2.4
1.4
1.5
1.3
4.9
-3.4
0.4
1.6
4.3
1.7
2.4
2.3
CPI
1.6
1.5
2.5
1.5
1.4
1.0
2.1
1.0
2.0
1.7
2.2
1.2
1.4
2.5
2.0
2.0
1.5
2.5
1.8
1.8
1.1
3.0
0.2
1.7
1.7
2.4
1.7
1.6
1.7

Source: ECB calculations.
Note: GDP at constant prices (annual percent change (y-o-y)), Unemployment (as % of the
labour force at year-end), Short-term interest rate (Short term interest rates (3M) at year-end Euribor or Libor depending on the country), Long term interest rates (Long term interest rates
(10Y) at year-end - Treasuries), Nominal USD exchange rate (Level of nominal USD exchange
rate to the respective currency at year-end), CPI (% change from previous year (y-o-y)).

43

Table 2: Macroeconomic scenarios – adverse scenario, including sovereign risk
GDP at constant
prices
-0.1
-0.3
-0.7
-0.1
0.7
0.2
-4.6
-2.1
-0.3
-0.1
-0.8
0.0
-0.3
0.8
0.7
-1.4
-0.2
-0.7
0.9
0.8
-0.1
-0.2
-4.2
-0.9
2.1
-1.8
0.9
-0.2
0.2
GDP at constant
2011 - Adverse
prices
Austria
-1.2
Belgium
-0.6
Cyprus
-0.1
Finland
-0.6
France
-0.1
Germany
-0.6
Greece
-4.3
Ireland
1.0
Italy
-0.3
Luxembourg
-0.8
Malta
-1.2
Netherlands
-1.0
Portugal
-2.3
Slovakia
-0.6
Slovenia
0.6
Spain
-1.2
Euro area
-0.6
Bulgaria
2.8
Czech R.
0.6
Denmark
0.2
Estonia
3.0
Hungary
1.6
Latvia
2.5
Lithuania
2.4
Poland
0.5
Romania
2.1
Sweden
0.9
UK
0.1
Rest of the EU
0.5

2010 - Adverse
Austria
Belgium
Cyprus
Finland
France
Germany
Greece
Ireland
Italy
Luxembourg
Malta
Netherlands
Portugal
Slovakia
Slovenia
Spain
Euro area
Bulgaria
Czech R.
Denmark
Estonia
Hungary
Latvia
Lithuania
Poland
Romania
Sweden
UK
Rest of the EU

Unemployment
6.1
9.9
6.7
10.4
10.2
9.2
11.8
14.1
8.8
7.3
7.6
5.5
11.3
12.9
8.5
20.3
10.8
9.2
8.6
6.0
16.4
12.6
20.7
17.6
10.7
8.5
10.2
9.1
9.6
Unemployment
6.1
11.1
7.3
11.4
10.5
9.7
14.8
13.7
9.3
7.7
8.2
7.0
12.8
13.2
9.1
21.6
11.5
8.4
9.6
6.3
14.8
13.2
18.8
16.3
12.2
9.2
10.3
8.8
9.6

Short-term
interest rates
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1
2.1

3.0

5.7
2.4
2.4
Short-term
interest rates
3.3
3.3
3.3
3.3
3.3
3.3
3.3
3.3
3.3
3.3
3.3
3.3
3.3
3.3
3.3
3.3
3.3

4.1

7.0
4.1
4.2

Long-term
interest rates
4.5
4.8
5.4
4.0
4.3
4.0
11.8
6.7
5.4
4.6
5.1
4.3
7.0
4.5
4.4
5.8
4.4
8.0
5.8
4.4
13.2
9.5
13.8
13.2
7.4
10.5
4.3
5.0

Nominal USD
exchange rate
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
1.4
18.7
5.0
11.5
196.5
0.5
2.5
2.9
3.0
7.0
0.6

Long-term
interest rates
5.3
5.6
6.3
4.9
5.1
4.7
14.7
7.8
6.3
5.5
6.0
5.1
8.5
5.4
5.3
6.8
5.3
8.0
5.8
5.1
13.2
9.5
13.8
13.2
7.6
10.5
4.9
5.7

Nominal USD
exchange rate
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
0.7
1.5
18.8
5.0
11.6
197.2
0.5
2.6
2.9
3.1
7.0
0.6

CPI
1.5
1.2
3.1
1.3
1.2
0.7
1.4
-0.6
1.7
1.8
1.8
0.8
1.3
1.8
1.8
1.0
1.1
2.0
0.9
1.2
0.9
4.8
-3.9
-0.2
2.5
3.9
1.3
2.4
2.3
CPI
1.0
0.6
2.1
0.1
1.0
0.6
2.1
0.7
1.7
1.4
1.6
1.0
0.9
1.4
1.9
1.2
1.1
0.5
0.9
1.2
-1.0
2.5
-3.6
-2.3
2.3
1.2
1.2
0.6
0.9

Source: ECB calculations.
Note: GDP at constant prices (annual percent change (y-o-y)), Unemployment (as % of the labour
force at year-end), Short-term interest rate (Short term interest rates (3M) at year-end - Euribor
or Libor depending on the country), Long term interest rates (Long term interest rates (10Y) at
year-end - Treasuries), Nominal USD exchange rate (Level of nominal USD exchange rate to the
respective currency at year-end), CPI (% change from previous year (y-o-y)).

44

Table 3: Contribution of the sovereign risk shock to the five-year bond yields in the euro
area under the adverse scenario
Five-year yields

Country
Austria
Belgium
Cyprus
Finland
France
Germany
Greece
Ireland
Italy
Luxembourg
Malta
The Netherlands
Portugal
Slovakia
Spain
Slovenia
Euro area average

Benchmark 2011
3.03
3.23
4.07
3.16
2.94
2.74
6.28
3.28
3.19
3.23
4.01
2.87
3.96
3.55
3.61
3.84
3.15

Adverse 2011
4.04
4.47
5.29
4.16
3.92
3.49
13.87
5.62
4.80
4.53
5.07
3.82
7.40
4.41
5.78
4.80
4.60

Country-specific
Common
upward shift of sovereign risk
shock
the yield curves
75
25
75
49
75
47
75
25
75
24
75
0
75
685
75
158
75
86
75
55
75
31
75
20
75
268
75
10
75
142
75
21
75
70

Source: ECB calculations.
Note: Contributions are expressed in basis points. Due to insufficient data on bond yields of
Cyprus, Luxembourg, Malta, Slovakia and Slovenia, a uniform additional widening of 30 basis
points was imposed for these five countries.

2. Probabilities of Default and Loss given Default
Estimates of probabilities of default (PD) 13 and loss given default (LGD) 14 parameters were computed
at the country level for five main portfolios (financial institutions, sovereign, corporate, consumer credit
and retail real estate). For all countries in the exercise, these parameters were computed for both the
benchmark and adverse scenarios for 2010 to 2011. 15
To calculate the PDs and LGDs conditional on the different scenarios, sector-specific regression
models 16 were used to link PDs and LGDs with macroeconomic variables. These models provide

13

The PD describes the likelihood that a loan will not be repaid and that it will fall into default. To calculate
the PD for each loan category the credit history of the counterparty as well as the nature of the investment is taken
into account. All PD figures in the EU-wide stress test are constructed for non-defaulted exposures.
14
The LGD is the amount of funds that is lost by a bank or other financial institution when a borrower
defaults on a loan.
15
In particular, PDs and LGDs increase as a result of the sovereign shock included in the adverse scenario,
with the only exception of the sovereign portfolio in the banking book, for consistency with the assumption of no
government default.
16
The regression models take account of dynamic interaction between the variables that drive PDs and
LGDs. The variables considered in the models were GDP, unemployment, long-term interest rates and sectoral PD
and LGD covering the period from 1991 until end-2009.

45

estimates of sector-specific elasticities of PDs and LGDs with respect to changes in macroeconomic
variables – conditional on shocks to the system. In the models, three propagation channels for the
shocks were identified: the demand channel; the supply channel and the long-term borrowing costs
channel. To obtain country-specific PD and LGD parameters for 2010 and 2011 under the benchmark
scenario, these elasticities were multiplied by the projected changes in macroeconomic variables for
each country using the PD and LGD levels that were observed at end-2009 as a starting point. Similarly,
to obtain PDs and LGDs under the adverse scenario in 2010 and 2011, the differences between the
macroeconomic variables in the benchmark and adverse scenarios for each year were multiplied with
the elasticities implied by the sector-specific regression models. For the purposes of using these
parameters for stress-testing the balance sheets of individual financial institutions, national supervisory
authorities were encouraged to use as a starting point their own PD and LGD levels for 2009 and to
apply the changes of these parameters in 2010 and 2011 with respect to their values in the benchmark
scenario in the respective year according to the outcomes of the ECB models. For some of the largest
banks for which a full bottom-up exercise was conducted, together with supervisory authorities,
supervisors could decide to allow these banks to feed the common macroeconomic scenarios into the
banks’ own internal models for the computation of PDs and LGDs.
Regarding the data entered into the ECB models for PDs and LGDs, country-level financial sector PDs
were approximated using the Moody’s EDFs (expected default frequency) extracted from the Moody’s
KMV database. 17 Sovereign PDs were derived from CDS spreads. Retail real estate PDs, consumer
credit PDs and corporate sector PDs were obtained from the ECB Monetary and Financial Institutions
(MFI) database on write-offs, while LGDs were extracted from Moody’s LossCalc database assuming a
constant PD over time. 18
To illustrate the severity of the adverse scenario, Chart 1 plots the ranges of changes across all countries
in the PDs between the adverse scenario and the end-2009 values over 2010 and 2011, for the four
private-sector portfolios sectors considered in the credit risk part of the exercise, and Table 4 shows the
corresponding figures for 2011. Chart 2 shows the results of the same exercise as Chart 1, now for the
LGD parameters. As seen in both charts, the PDs and LGDs increase substantially across sectors and
countries under the adverse scenario compared to end-2009 in both 2010 and 2011. In this regard, it is
important to note that the stresses on long-term interest rates that result from the sovereign shock feed
through to higher PD and LGD levels.

3. Sovereign bond haircuts

17

For details see “www.moodyskmv.com”.
PD and LGD levels for 2009 were calibrated on the basis of results from data collections from national
authorities, various surveys conducted by the CEBS and the ECB, and market information.
18

46

The increase in bond yields affects the valuation of holdings of government debt in the banks’
trading books, 19 and in the exercise its impact is not offset by changes in the valuation of
derivative positions (credit derivatives, interest rate swaps, etc.) that are used to hedge the
sovereign bond exposures.
For the purposes of estimating valuation haircuts, it was agreed among participating
supervisors that a five-year maturity was representative of the approximate duration of
sovereign bond holdings held by banks in the EU. Hence, the haircuts for sovereign bonds are
computed in two steps, first by estimating
Chart 1: Changes in PDs across sectors dispersions across countries under the
adverse scenario compared to 2009

Chart 2: Changes in LGDs across sectors dispersions across countries under the
adverse scenario compared to 2009

(%, maximum, minimum, interquartile range,
median)

(%, maximum, minimum, interquartile range,
median)

400

Institutions

Corporate

350

Retail real
estate

120

Consumer
credit

Institutions

Corporate

110

Retail real
estate

Consumer
credit

100
300

90

250

80
70

200

60
150

50

100

40
30

50

20
10

0

0

-50

-10

19

2011

2010

2011

2010

2011

2010

2010

Source: ECB calculations.
Note: Box-Whisker plots show the interquartile
range of the distribution within the blue box,
which includes the median denoted by the black
bar. The upper and lower black bars at the
extremes illustrate the maximum and the
minimum of the distribution.

2011

-20

2011

2010

2011

2010

2011

2010

2011

2010

-100

Source: ECB calculations.
Note: Box-Whisker plots show the interquartile
range of the distribution within the blue box,
which includes the median denoted by the black
bar. The upper and lower black bars at the
extremes illustrate the maximum and the
minimum of the distribution.

Since no sovereign defaults are considered in the exercise, there is no impact on holdings of sovereign
bonds which are held to maturity in the banking book.

47

Table 4: Changes in PDs in 2011 across sectors
under the adverse scenario, compared to 2009
(%)
Austria
Belgium
Cyprus
Finland
France
Germany
Greece
Ireland
Italy
Luxembourg
Malta
Netherlands
Portugal
Slovenia
Slovakia
Spain
Euro area
Bulgaria
Czech Republic
Denmark
Estonia
Hungary
Latvia
Lithuania
Poland
Romania
Sweden
UK
Rest of the EU

Institutions
10.8
68.6
14.8
10.8
11.3
22.6
45.0
-0.5
10.0
11.0
11.9
66.1
31.0
0.7
-1.8
29.4
8.5
14.3
87.4
1.9
-5.4
36.2
-1.0
9.5
58.9
16.9
2.6
0.9
1.6

Corporate
47.4
112.4
69.4
46.8
31.4
57.5
364.8
21.7
41.6
71.6
54.9
88.5
147.0
23.9
7.7
113.1
61.3
12.9
61.2
26.7
5.8
35.3
13.1
6.9
56.0
19.8
32.4
22.6
25.0

Source: ECB calculations.

48

Retail real estate Consumer credit
21.9
24.9
32.0
55.4
14.5
34.8
29.2
18.4
13.0
21.4
36.2
32.1
26.5
74.2
3.6
4.9
11.2
21.4
21.8
34.6
18.5
36.0
39.0
46.9
30.3
102.3
24.9
4.2
8.0
0.8
17.1
56.3
20.8
25.8
8.5
15.2
41.6
66.7
5.6
14.7
4.5
8.6
21.5
40.8
9.7
15.9
12.6
10.8
39.7
62.3
14.9
23.4
14.5
12.3
6.2
13.9
5.5
13.7

five-year bond yields, consistent with the assumptions for ten-year yields and then, in a second
step, translating these five-year yields into their corresponding sovereign bond prices.
a. Transformation of ten-year yields to five-year yields
The transformation uses the ten-year yields prevailing in the benchmark and adverse scenarios
together with the five-year yields that were assumed to prevail in the market at the end of 2009.
The changes in five-year bond yields from 2009 to 2010 and to 2011 were set equal to the
changes (in basis points) in the ten-year yields. This method was applied for all countries, apart
from Germany, which acts as the reference sovereign issuer with the lowest yield in the euro
area. 20 The exceptions are the euro area countries where the bond markets are not liquid or
where this method would lead to a significant compression of sovereign bond yield spreads visà-vis German bonds. For those countries (Cyprus, Malta, Slovakia, and Slovenia) it was
assumed that the sovereign bond yield spreads over the German yields would remain constant
in the benchmark scenario.
In the adverse scenario, the five-year yields are constructed from the values in the benchmark
scenario using the same procedure as followed for the ten-year yields, taking into account both
the yield curve flattening and the sovereign risk components. Again Germany, being the
reference issuer, is assumed to be unaffected by the elevated sovereign risk.
b. Haircuts on sovereign debt
The haircuts were computed from changes in the prices of five-year sovereign bonds under both
scenarios. The parameters that are essential for the pricing of sovereign bonds (coupons, coupon
frequencies, coupon and maturity dates) were collected from Bloomberg. In order to eliminate potential
distortions which may arise when the bonds that are currently the most actively traded have been issued
with very high or very low coupons, 21 all bonds for which market quotes were available on Bloomberg
for each country that had a remaining maturity of 4.5 to 6.5 years were priced and the weighted average
change in their prices was used to construct the haircut. The weights in the average are based on the
outstanding amount of the bonds.
In the pricing of sovereign bonds the discounted cash-flow method was used, in which the yields to
maturity under the relevant scenario are used to construct the discount factors. This method takes into
account the actual maturity dates, coupon dates and coupon frequencies for the individual bonds.
The haircuts are applied to the market value of bonds at the end of 2009, separately for each year.
Therefore, a bond which was worth 100 at the end of 2009 and which has a haircut of 4% in 2010 and

20

For the computation of haircuts, the country-specific sovereign risk takes Germany as the reference, so
that the German yields and haircuts under the adverse scenario are not affected by the elevated sovereign risk (see
Table 3).
21
For example, a bond used in the exercise may have an original maturity of 30 years (i.e. issued in 1985)
and a remaining maturity of 5 years. The coupons on such a bond can be out of line with the prevailing yields,
thereby distorting the comparisons between the sensitivities of bond prices to changes in the yields.

49

6% in 2011 should be valued at 96 at the end of 2010 and at 94 at the end of 2010. The haircuts used in
the exercise (Table 5) are the future values of the outstanding sovereign bonds. The exercise is supposed
to provide the values of the bonds to be booked in the end-2010 and end-2011 accounts. This implies
that a 5-year bond, representative of the average maturity of this portfolio by banks, has a duration of
only 3 years at the end of 2011, when accounts are closed.

Table 5:

Five-year bonds yields and haircuts used in the exercise

Country
end-2009
Austria
2.69
Belgium
2.79
Cyprus
3.75
Finland
2.62
France
2.48
Germany
2.42
Greece
4.96
Ireland
2.91
Italy
2.80
Luxembourg
2.79
Malta
3.69
The Netherlands
2.46
Portugal
3.08
Slovakia
3.24
Spain
2.96
Slovenia
3.52
Czech Republic
3.29
Denmark
2.80
Poland
5.96
Sweden
2.41
United Kingdom
2.81
Other non-euro area EU countries
EU average

Bond yields
Benchmark
2010
2011
2.72
3.03
2.92
3.23
3.58
4.07
2.35
3.16
2.63
2.94
2.25
2.74
5.97
6.28
2.97
3.28
2.89
3.19
2.92
3.23
3.52
4.01
2.57
2.87
3.53
3.96
3.07
3.55
3.31
3.61
3.35
3.84
3.19
2.87
2.63
3.12
6.56
6.78
2.64
2.92
3.67
4.02

Adverse
2010
2011
3.29
4.04
3.66
4.47
4.30
5.29
2.91
4.16
3.18
3.92
2.81
3.49
11.03
13.87
4.50
5.62
3.90
4.80
3.72
4.53
4.13
5.07
3.08
3.82
5.83
7.40
3.46
4.41
4.74
5.78
3.84
4.80
4.35
4.22
3.63
4.29
7.72
8.13
3.32
3.97
4.34
5.07

Haircuts
Benchmark
2010
2011
1.0%
2.8%
1.4%
3.1%
0.3%
3.2%
0.0%
3.3%
1.5%
3.0%
0.1%
2.5%
3.9%
4.3%
1.6%
4.2%
1.2%
2.9%
1.4%
3.1%
0.7%
3.6%
1.1%
2.5%
2.3%
3.7%
0.1%
2.4%
1.3%
4.1%
0.0%
1.1%
0.0%
2.7%
0.0%
1.4%
2.6%
6.1%
1.3%
2.3%
5.0%
6.9%
1.3%
4.4%
1.3%
3.3%

Adverse
2010
3.1%
4.3%
3.0%
1.9%
3.7%
2.3%
20.1%
8.6%
4.9%
4.3%
2.9%
3.0%
11.1%
1.6%
6.7%
1.4%
4.6%
2.1%
6.4%
5.0%
7.7%
5.5%
5.2%

2011
5.6%
6.9%
6.7%
6.1%
6.0%
4.7%
23.1%
12.8%
7.4%
6.9%
6.4%
5.2%
14.1%
5.0%
12.0%
4.2%
11.4%
5.2%
12.3%
6.7%
10.2%
11.8%
8.5%

Source: ECB calculations.
Note: As discussed in the note, future value haircuts, relative to the market value of the bonds on
31 December 2009, are used in the exercise and are listed in this table.

The haircuts can be decomposed to reflect the three main contributing factors: the overall rise in longterm interest rates foreseen in the benchmark macroeconomic scenario, the common upward shift of the
yield curves, and the country-specific sovereign risk shock (Table 6). The decomposition illustrates that
for some non-euro area countries, the higher haircuts are driven primarily by the expected increase in
long-term interest rates, with the impact of the sovereign risk shock playing a lesser role.
For the purposes of illustration and comparison with the haircuts on five-year bonds, the same
calculations were carried out for ten-year bonds (Table 7). The yields used to calculate the haircuts on
ten-year bonds are the yields provided as part of the macroeconomic scenario that, where appropriate,
include a sovereign risk component. The haircuts on the ten-year bonds are generally higher than the
corresponding haircuts on the five-year bonds due to the higher duration. Taking Austria as an example,
the haircut on the five-year bonds under the adverse scenario is 5.6%. The corresponding haircut on tenyear bonds is 9.5%. For Greece, the respective figures are 23.1% and 42.2%.

50

Table 6:

Decomposition of the five-year adverse scenario haircuts
Common
Benchmark
upward
macroeconomic shift of the
scenario
yield curves
2.8%
2.1%
3.1%
2.3%
3.2%
2.2%
3.3%
2.1%
3.0%
2.4%
2.5%
2.2%
4.3%
2.1%
4.2%
2.9%
2.9%
2.1%
3.1%
2.3%
3.6%
2.0%
2.5%
2.1%
3.7%
2.4%
2.4%
2.2%
4.1%
2.4%
1.1%
2.4%
2.7%
2.5%
1.4%
2.4%
6.1%
2.6%
2.3%
3.0%
6.9%
2.4%
4.4%
2.6%

Countryspecific
sovereign
risk shock
0.7%
1.5%
1.4%
0.7%
0.7%
0.0%
16.8%
5.8%
2.4%
1.5%
0.8%
0.5%
7.9%
0.3%
5.5%
0.7%
6.2%
1.4%
3.5%
1.4%
0.9%
4.8%

Country
Austria
Belgium
Cyprus
Finland
France
Germany
Greece
Ireland
Italy
Luxembourg
Malta
The Netherlands
Portugal
Slovakia
Spain
Slovenia
Czech Republic
Denmark
Poland
Sweden
United Kingdom
Other non-euro area EU countries
Source: ECB calculations.
Note: the decomposition illustrates the impact of the three factors that influence the haircuts
on the sovereign bonds under the adverse scenario in 2011. These factors are: the
macroeconomic outlook in the benchmark scenario, the upward shift of the bond yields by 75
basis points that is common to all countries, and the country-specific sovereign risk shock
(see Section 1 for further details). The rows sum to the total 2011 haircut for each country’s
bonds under the adverse scenario.

51

Table 7:

An Example: Yields and future value haircuts on ten-year bonds

end-2009
Country
Austria
3.94
Belgium
3.91
Cyprus
4.60
Finland
3.73
France
3.65
Germany
3.27
Greece
5.77
Ireland
5.06
Italy
4.29
Luxembourg
3.91
Malta
4.54
The Netherlands
3.71
Portugal
4.20
Slovakia
4.72
Spain
4.03
Slovenia
4.37
Czech Republic
4.80
Denmark
3.62
Poland
6.22
Sweden
3.35
United Kingdom
3.45
Other non-euro area EU countries
EU average

Bond yields
Benchmark
2010
2011
3.97
4.28
4.05
4.36
4.66
5.08
3.46
3.88
3.81
4.11
3.50
3.80
6.79
7.09
5.12
5.43
4.37
4.68
3.80
4.23
4.49
4.92
3.81
4.12
4.66
5.08
4.14
4.57
4.38
4.68
3.93
4.36
4.70
4.39
3.77
4.05
6.82
7.05
3.58
3.85
4.32
4.66

Adverse
2010
2011
4.54
5.29
4.79
5.59
5.37
6.30
4.02
4.89
4.35
5.09
3.97
4.72
11.84
14.69
6.65
7.76
5.39
6.29
4.60
5.53
5.10
5.98
4.32
5.07
6.96
8.52
4.54
5.42
5.81
6.85
4.42
5.32
5.84
5.84
4.45
5.10
7.42
7.65
4.25
4.90
4.99
5.71

Haircuts
Benchmark
2010
2011
0.3%
2.6%
1.1%
3.2%
2.6%
5.2%
0.0%
0.6%
1.2%
3.7%
1.6%
3.5%
7.1%
8.8%
0.5%
2.7%
0.4%
2.3%
0.0%
1.4%
0.2%
3.8%
0.9%
3.0%
3.7%
6.5%
0.0%
0.0%
4.6%
8.3%
0.0%
0.0%
0.0%
0.0%
1.5%
3.7%
3.9%
4.7%
3.0%
6.0%
7.3%
9.9%
1.9%
2.4%
1.9%
3.8%

Source: ECB calculations.
Note: these haircuts were not used in the stress test exercise and are presented only for the sake of comparison.

52

Adverse
2010
4.7%
6.7%
7.5%
1.9%
5.4%
5.2%
33.3%
11.0%
7.7%
4.9%
4.5%
4.6%
19.4%
0.0%
14.6%
0.4%
7.4%
6.3%
7.8%
8.2%
11.9%
7.6%
8.3%

2011
9.5%
11.5%
12.4%
7.2%
10.4%
9.4%
42.2%
16.6%
12.3%
9.7%
10.3%
9.0%
26.6%
3.9%
21.7%
5.8%
12.2%
10.2%
13.1%
13.2%
16.2%
12.6%
13.5%

Annex 3. Evolution of property prices assumed in the exercise
UK

Benchmark Scenario

Germany

France

Netherlands

2010

2011

2010

2011

2010

2011

2010

2011

0.0%

0.0%

0.0%

0.0%

-4.5%

-2.5%

0.0%

0.0%

2.0%

1.0%

0.0%

0.0%

-4.5%

-2.5%

0.0%

2010
2011
2010
2011
2010
Adverse Scenario
Commercial Property Prices, % change from previous
-10.0% -10.0% -10.0% -10.0% -4.5%
year (y-o-y)
Residential Property Prices, % change from previous
-10.0% -10.0% -10.0% -10.0% -4.5%
year (y-o-y)

2011

2010

Commercial Property Prices, % change from previous
year (y-o-y)
Residential Property Prices, % change from previous
year (y-o-y)

Greece

Benchmark Scenario

2010

Commercial Property Prices, % change from previous
-3.0%
year (y-o-y)
Residential Property Prices, % change from previous
-3.0%
year (y-o-y)
2010
Adverse Scenario
Commercial Property Prices, % change from previous
-5.0%
year (y-o-y)
Residential Property Prices, % change from previous
-5.0%
year (y-o-y)

2011

Ireland

Italy

Belgium

Sweden

Austria

Denmark

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

-20.0% -15.0%

-0.7%

0.3%

-3.0%

-3.0%

0.0%

-2.5%

2.0%

2.7%

0.2%

2.0%

0.0%

-3.8%

-5.2%

-0.7%

0.3%

-3.0%

-3.0%

5.0%

0.0%

2.0%

2.7%

0.2%

2.0%

2011

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

-4.5% -10.0% -10.0% -35.0% -30.0%

-1.6%

-2.0% -10.0% -10.0% -12.5% -15.0%

2.0%

2.7%

-7.0%

-1.6%

-4.5% -10.0% -10.0% -8.8%

-1.6%

-2.0% -10.0% -10.0% -7.5% -12.5%

2.0%

2.7%

-7.0%

-1.6%

Cyprus

Luxembourg

2011

-15.2%

Malta

Portugal

Slovenia

Finland

Hungary

Poland

2011

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

0.0%

-14.0% -6.0%

-2.0%

2.0%

n.a.

n.a.

1.6%

2.5%

0.0%

0.0%

n.a.

n.a.

0.0%

0.0%

-5%

0%

0.0%

0.0%

0.0%

-13.0% -2.5%

-2.0%

2.0%

n.a.

n.a.

1.6%

2.5%

0.0%

0.0%

n.a.

n.a.

0.0%

0.0%

-10%

-3%

0.0%

0.0%

2011

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

2010

2011

-2.0% -17.0% -8.0%

-4%

0%

n.a.

n.a.

-2.0%

-2.0%

-5.0%

-5.0%

n.a.

n.a.

-10.0% -10.0%

-8%

-5%

0.0%

0.0%

-2.0% -17.0% -5.0%

-4%

0%

n.a.

n.a.

-2.0%

-2.0%

-5.0%

-5.0%

n.a.

n.a.

-5.0%

-13%

-8%

0.0%

0.0%

2011

2010

Spain
2010

2010

-5.0%

Notes:
Estimates of the annual changes of the property prices are provided by the national supervisory authorities participating in the exercise
and are outside the scope of the macro-economic scenarios. It should be noted that at the current time in Europe there is no consistent
reference set of property price data, which can be used for central modelling.

55

Annex 4. Backtesting of key variables of the
macro-economic scenarios
UK
DE
FR
NL
ES
IT
BE
SE
AT
DK
EL
IE
CY
LU
MT
PT
SI
SK
FI
BG
CZ
EE
HU
LT
LV
PL
RO
NO
Euro area
Rest of EEA
US
Rest of the world

GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment
GDP
unemployment

ST 2009
ST 2010
ST 2009
ST 2010
ST 2009 ST 2010
Realised Realised Realised Benchmark Adverse Benchmark Benchmark Adverse Adverse Benchmark Adverse
2008
2009
Q1 2010
2009
2009
2010
2010
2010
2010
2011
2011
‐0.1
‐4.9
0.3
‐3.8
‐5.0
0.1
0.6
‐2.9
‐0.2
1.9
0.1
5.6
7.6
7.9
8.2
8.6
9.4
8.7
10.6
9.1
8.0
8.8
1.3
‐4.9
0.2
‐5.4
‐6.8
0.3
1.2
‐3.1
0.2
1.7
‐0.6
7.3
7.5
7.3
8.6
8.6
10.4
9.2
11.1
8.2
9.3
9.7
0.2
‐2.6
0.1
‐3.0
‐3.8
‐0.2
1.2
‐1.7
0.7
1.5
‐0.1
7.8
9.5
9.9
9.6
9.6
10.7
10.2
11.3
10.2
10.0
10.5
2.0
‐4.0
0.3
‐3.5
‐5.1
‐0.4
0.9
‐3.3
0.0
1.6
‐1.0
2.8
3.4
4.2
3.9
4.2
6.2
5.4
7.8
5.5
6.0
7.0
0.9
‐3.6
0.1
‐3.2
‐4.2
‐1.0
‐0.6
‐3.9
‐1.4
1.0
‐1.2
11.3
18.0
19.5
17.3
17.7
20.5
20.0
22.4
20.3
20.5
21.6
‐1.3
‐5.0
0.4
‐4.4
‐5.7
0.1
0.7
‐1.7
‐0.3
1.4
‐0.3
6.7
7.8
8.7
8.8
8.9
9.4
8.7
10.2
8.8
8.7
9.3
1.0
‐3.0
0.1
‐3.5
‐4.8
‐0.2
0.6
‐3.3
‐0.3
1.5
‐0.6
7.0
7.9
8.5
8.5
8.6
10.3
9.9
11.4
9.9
10.3
11.1
‐0.4
‐5.1
1.4
‐4.0
‐4.7
0.8
1.4
‐1.7
0.9
2.1
0.9
6.2
8.3
8.6
8.4
8.6
10.4
10.2
11.0
10.2
10.1
10.3
2.2
‐3.9
‐0.1
‐4.0
‐5.5
‐0.1
1.1
‐2.8
‐0.1
1.5
‐1.2
3.8
4.8
4.2
6.0
6.1
7.1
6.0
7.7
6.1
5.7
6.1
‐0.9
‐4.9
0.5
n.a.
n.a.
n.a.
1.5
n.a.
0.8
1.8
0.2
3.3
6.0
7.1
n.a.
n.a.
n.a.
5.8
n.a.
6.0
5.6
6.3
2.0
‐2.0
‐1.0
n.a.
n.a.
n.a.
‐4.1
n.a.
‐4.6
‐2.6
‐4.3
7.7
9.5
11.0
n.a.
n.a.
n.a.
11.7
n.a.
11.8
14.1
14.8
‐3.0
‐7.1
2.7
n.a.
n.a.
n.a.
‐1.4
n.a.
‐2.1
2.6
1.0
6.3
11.9
12.8
n.a.
n.a.
n.a.
14.0
n.a.
14.1
13.2
13.7
3.6
‐1.7
0.1
n.a.
n.a.
n.a.
0.1
n.a.
‐0.7
1.3
‐0.1
3.6
5.3
6.8
n.a.
n.a.
n.a.
6.6
n.a.
6.7
6.7
7.3
0.0
‐4.1
n.a.
n.a.
n.a.
n.a.
1.1
n.a.
‐0.1
1.8
‐0.8
4.9
5.2
5.2
n.a.
n.a.
n.a.
7.3
n.a.
7.3
7.7
7.7
1.7
‐1.5
0.8
n.a.
n.a.
n.a.
0.7
n.a.
‐0.8
1.6
‐1.2
5.9
6.9
6.9
n.a.
n.a.
n.a.
7.4
n.a.
7.6
7.3
8.2
0.0
‐2.6
1.1
n.a.
n.a.
n.a.
0.5
n.a.
‐0.3
0.2
‐2.3
7.7
9.6
10.6
n.a.
n.a.
n.a.
11.1
n.a.
11.3
11.9
12.8
3.5
‐7.8
‐0.5
n.a.
n.a.
n.a.
1.3
n.a.
0.7
2.0
0.6
4.4
5.9
6.4
n.a.
n.a.
n.a.
8.3
n.a.
8.5
8.5
9.1
6.2
‐4.7
0.8
n.a.
n.a.
n.a.
1.9
n.a.
0.8
2.6
‐0.6
9.5
12.0
14.7
n.a.
n.a.
n.a.
12.8
n.a.
12.9
12.6
13.2
0.9
‐8.0
‐0.4
n.a.
n.a.
n.a.
0.9
n.a.
‐0.1
1.6
‐0.6
6.4
8.2
9.0
n.a.
n.a.
n.a.
10.2
n.a.
10.4
9.9
11.4
6.0
‐5.0
n.a.
n.a.
n.a.
n.a.
0.4
n.a.
‐0.7
4.0
2.8
5.6
6.8
9.5
n.a.
n.a.
n.a.
8.8
n.a.
9.2
8.0
8.4
2.5
‐4.1
0.5
n.a.
n.a.
n.a.
1.4
n.a.
0.9
1.8
0.6
4.4
6.7
7.9
n.a.
n.a.
n.a.
8.1
n.a.
8.6
8.5
9.6
‐3.6
‐14.1
‐2.0
n.a.
n.a.
n.a.
1.0
n.a.
‐0.1
4.0
3.0
5.5
13.8
19.0
n.a.
n.a.
n.a.
16.0
n.a.
16.4
14.5
14.8
0.6
‐6.3
0.9
n.a.
n.a.
n.a.
0.9
n.a.
‐0.2
3.2
1.6
7.8
10.0
11.2
n.a.
n.a.
n.a.
11.8
n.a.
12.6
11.9
13.2
2.8
‐14.8
‐3.9
n.a.
n.a.
n.a.
0.5
n.a.
‐0.9
3.1
2.4
5.8
13.7
17.4
n.a.
n.a.
n.a.
17.1
n.a.
17.6
15.9
16.3
‐4.2
‐18.0
0.3
n.a.
n.a.
n.a.
‐3.3
n.a.
‐4.2
3.9
2.5
7.5
17.1
20.0
n.a.
n.a.
n.a.
20.4
n.a.
20.7
18.2
18.8
5.0
1.7
0.5
n.a.
n.a.
n.a.
2.9
n.a.
2.1
2.4
0.5
7.1
8.2
9.9
n.a.
n.a.
n.a.
10.4
n.a.
10.7
11.5
12.2
7.3
‐7.1
‐0.3
n.a.
n.a.
n.a.
‐0.7
n.a.
‐1.8
3.6
2.1
5.8
6.9
7.4
n.a.
n.a.
n.a.
8.1
n.a.
8.5
8.8
9.2
1.8
‐1.6
‐0.1
n.a.
n.a.
n.a.
1.6
n.a.
1.2
2.1
0.9
2.5
3.1
3.5
n.a.
n.a.
n.a.
4.3
n.a.
4.3
4.1
4.3
0.6
‐4.1
0.2
‐4.0
‐5.2
‐0.1
0.7
‐2.7
‐0.2
1.5
‐0.6
7.5
9.4
10.0
9.9
10.0
11.5
10.7
12.5
10.8
10.9
11.5
n.a.
n.a.
n.a.
‐3.6
‐4.9
0.1
1.0
‐2.9
0.0
2.8
1.0
n.a.
n.a.
n.a
7.3
7.9
8.8
9.6
10.2
10.0
9.3
10.1
0.4
‐2.4
0.7
‐2.9
‐3.7
0.9
2.2
‐0.3
1.5
2.0
0.6
5.8
9.3
9.7
8.9
9.2
10.2
10.0
11.2
10.2
10.2
11.1
n.a.
n.a.
n.a.
0.2
‐0.7
3.2
4.4
1.8
3.6
4.8
3.5
n.a.
n.a.
n.a.
n.a.
n.a.
n.a.
10.2
n.a.
10.5
10.2
11.5

Notes:
GDP for realised in real GDV growth y-o-y, realised unemployment is % of unemployed as of total labour force
GDP for realised Q1 2010 is expressed in percentage change compared with Q4 2010
Source: Eurostat data for realised, stress test scenarios

55

Annex 5. Set of parameters for the market risk
component of the exercise

Interest Rates

Fx

Equity

Parameter
USD 3M
USD 2Y
USD 10Y
EUR 3M
EUR 2Y
EUR 10Y
UK 3M
UK 2Y
UK 10Y
US Volatility
EUR Volatility
UK Volatility
Spread swap / governments
EUR/USD
JPY/USD
GBP/USD
EUR/USD Volatility
JPY/USD Volatility
GBP/USD Volatility
Gold/USD
Eurostoxx50
US (S&P500)
Japan (NIKKEI)
Emerging
Eurostoxx50 Volatility
US (S&P500) Volatility
Japan (NIKKEI) Volatility
Emerging Volatility
Dividends Europe
Dividends US
Dividends Japan
Dividends Emerging

Hegde Funds
Mutual Funds
Commodities

Brent
Brent Volatility
Other commodities
Other commodities Volatility
Itraxx / CDX IG
Itraxx / CDX HY
Itraxx Levx/ LCDX
ABX AAA ( 2006 and 2007 series)
ABX lower than AAA ( 2006 and 2007
series)

Unit
bp
bp
bp
bp
bp
bp
bp
bp
bp
%
%
%
bp
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%
%

Baseline
100
90
50
100
90
50
100
90
50
30
30
30
20
-10
-10
-10
30
30
30
-7
-10
-10
-10
-10
20
25
35
25
-10
-10
-10
-10
-10
-5
-15
15
-5
30
20
20
20
40

Adverse
200
170
50
200
170
50
200
170
50
60
60
60
40
-20
-20
-20
60
60
60
-15
-20
-20
-20
-20
40
50
70
50
-20
-20
-20
-20
-20
-10
-30
30
-10
60
40
40
40
80

%

30

60

ABX AAA (other series)

30

60

%

30

CMBX AAA (all the series)

%

40
30

%

30

60

CMBS lower than AAA (Europe)

%

30

60

RMBS AAA (Europe)

%

30

60

RMBS lower than AAA (Europe)

%

20

40

Bid-ask spread

%

100

200

= % change in the spread

60

CMBS AAA (Europe)

percentage change of the
reference point (ex: 20% +
30% of 20% = 26%)
parallel move along the curve

80

%

Swap curve. For each
currency, yield curve to be
interpolated linearly up to 10Y,
constant at the level of the
10Y from that point on

60

CMBX lower than AAA (all the series)

Credit

%

ABX lower than AAA (other series)

Comments

= % change in the spread

= % change in the spread

Market Liquidity

Notes:
Reference points for all changes in parameters: 31 December 2009

55

Applies to each relevant bidask spread for a relevant
transaction


Federal Reserve Bank of St. Louis, One Federal Reserve Bank Plaza, St. Louis, MO 63102