Master of Science in
Business Economics: Finance
Master Thesis
“
Market reaction to the 2014 stress test
”
Did investors find valuable information in the results of the 2014 EU-wide
stress test of the European banking sector?
Author: Redian Asllanaj Student ID: 10826548
Supervisor: Dr. Tanju Yorulmazer Submitted at: 15th August, 2016
Statement of Originality
I declare that I have completed the present thesis independently, without making use of other than the specified literature and aids. Sentences or parts of sentences quoted literally are marked as quotations; identification of other references with regard to the statement and scope of the work is quoted. This thesis in this form or in any other form has not been submitted to an examination body and has not been published. This thesis has not been used, either in whole or part, for another examination achievement.
Abstract
In October 2014 the European banking authority and the European central bank conducted an Asset Quality Review and a stress test to check the resilience of the European banking sector. This stress test included 123 European banks. The results of this stress test showed that 24 banks failed and they needed to raise capital. In this paper I examine whether and how this stress test affected stock prices and credit default swaps of the banks that took part in.
Through an event study research, I find that the release of the results of this stress test affected significantly the stock prices but did not affected the CDS spreads. By separating banks in groups I find significant positive stock market reaction for banks which passed the test and negative for banks which failed. Same conclusion cannot be drawn from the results of the bond market.
- 1 -
Table of Contents
1. Introduction ... 2 -2. Literature review ... 4 -3. 2014 Stress Test ... 9 -4. Methodology ... 10 -5. Data ... 16 -6. Empirical results ... 17-6.1 Event day: Friday ... 17
-6.2 Event day: Monday ... 22
-7. Conclusion ... 24
List of References ... 25
-- 2 --
1.
Introduction
In the long term, it is vital for banks and financial institutions to be able to withstand severe unexpected losses. To resist unexpected losses, financial institutions need to have sufficient capital buffers which will serve as cushions and absorb extreme shocks. For the stability of the financial system, regulatory bodies have set specific capital requirements. Bank capital requirements are rules that force banks to maintain a minimum ratio of capital to assets. The purpose of these capital requirements is to ensure that banks can sustain significant unexpected losses in the values of assets they hold while still honoring withdrawals and other essential obligations (Calomiris, 2012).
In the recent past, stress tests have become a valuable tool in the hands of regulatory bodies in order to identify and monitor systemic risk in the banking sector. Banking stress tests are conducted firstly to ensure the effective operation of banks and secondly to restore confidence in the market and to curb bank opaqueness by helping investors distinguish better between healthy and weak institutions (European Banking Authority, 2014). Stress tests are forward-looking exercises designed to estimate losses, revenues and reserve needs for banks in the next years under two macroeconomic scenarios, one based on current macroeconomic expectations and the other more adverse (Bernanke, 2013).Since the financial crisis in 2008, it became more essential for regulatory bodies to perform stress tests.
There have been previous studies1 on whether the new information, provided from stress tests results, produces valuable new information for the market and plays a valid role in the reduction of bank opacity. All these studies conclude that stress tests provide, in different ways, related new information that investors were not able to anticipate. The purpose and findings of these studies will be discussed in the next section.
1
Morgan et al. (2010) and Neretina et al. (2015) do research on the SCAP 2009 in the USA. Cardinali and Nordmark (2011), Ellahie (2012), Alves et al. (2013) and Petrella and Resti (2013) examine the 2010 and 2011 EBA stress tets.
- 3 - On 26 October 2014 the European Banking Authority (EBA) published the results of the latest bank stress test. Out of the 123 tested banks 24 failed to meet the capital requirements in the fictional scenarios and an overall shortfall of 24.6 billion euros was found. However, these results have been criticized because regulators were looking purely at the common equity Tier 1 ratio (Baruccia et al. 2014, De Groen 2014 and Steffen 2014). This ratio is computed by the risk weighted assets of each institution. Risk weighted assets are the sum of total assets of an institution weighted accordingly to the risk they contain. Each institution calculates its risk weights by using an “internal ratings-based” (IRB) approach (Basel Committee on Banking Supervision, 2005). This leads to a different perception of risk inside the sector. Acharya et al. (2014) show that risk-weighted assets use risk measures which are cross-sectionally not correlated with market risk measures. According to the Centre for European Policy Studies (2014), 24 more banks would have failed if the European Central Bank had applied the leverage minimum ratio (De Groen 2014). Having that in mind, this paper will investigate whether the stock market and the bond market found valuable new information in the results of the 2014 EU-wide stress test or investors were able to anticipate those results.
In order to assess whether 2014 stress test provided valuable information to the market this paper is going to check the effect on the stock prices and the credit default swaps spreads of the banks which took part in the test. Although this question has been raised again during previous stress tests, the sample of stress tests is not large enough to have a clear picture of its effects. Banking stress test inside European Union is something relatively new. First stress test was conducted in EU in 2009 in which the results were not published. Since then there have been two more stress tests (2010, 2011).
EBA is conducting this stress test right before ECB takes responsibility for the supervision of euro area banks in November 2014. The ECB examines through this stress test banks’
- 4 - resilience and their balance sheets. In this stress test the ECB, for the first time, has carried out simultaneously an Asset Quality Review (AQR), which is a comprehensive audit to find out a more reliable overview of the true value of banks’ assets (Steffen, 2014). By carrying on an event study2, this paper will examine whether there was reaction in the stock market or the bond market to the announcement of the results of this stress test. The paper unfolds as follows: Section 2 summarizes previous research results on bank opaqueness and stress tests; Section 3 presents the main features of the 2014 European stress test; Section 4 describes methodology and key variables; Section 5 refers to the data sample; Section 6 shows main results; Section 7 concludes the thesis.
2.
Literature review
The EBA implemented banking stress test for two main reasons. The first reason is to evaluate the resilience of banks in an adverse macroeconomic scenario. If the results show any of the institutions have capital deficiencies, the EBA will require them to take action for their equity improvement. Second reason is to publish this information in order to provide more transparency to the market and to restore confidence (Report of 2014 EU-wide stress test results). It is argued that banks’ assets are relatively more difficult to valuate than other firms, but the topic is still controversial. There has been considerable research in the topic of opaqueness in the banking sector (Jordan et al. (2000), Morgan (2002), Flannery et al. (2004), Iannotta (2006), Haggard and Howe (2007) and Jones et al. (2012)). Due to their core activities, banks hold most of the time a portfolio of very illiquid assets with long maturities. Studies by Campbell and Kracaw (1980), Berlin and Loeys (1988), and Diamond (1991) all lead to the conclusion that bank loans are opaque.
- 5 - As a proof of bank opacity, Jordan et al. (2000), are using the stock market reaction to announcements of formal supervisory actions. They find a stock market reaction to these announcements, which means that investors were not able to anticipate the relevant announced information. Flannery et al. (2004) on the other hand, conclude that banks’ assets are no more opaque than other similar industrial firms. By assessing the market microstructure properties of banking firms’ equity they find that large bank institutions in U.S. have very similar trading properties to their matching non-banking firms. Although they find that small banks trade less frequently than their matching non-financial firms they suggest that this is happening simply because investors don’t find them interesting.
Haggard and Howe (2007), on the same level, examine the relative bank opacity and they find that banks have less firm-specific information in their equity returns than industrial matching firms. They also provide evidence that those banks have less firm-specific information in their equity returns and are therefore more prone to experience significant declines in their stock price.
Morgan (2002) approaches the topic of bank opaqueness through the deviation on banks’ evaluation by external rating agencies. He concludes that banks assets are relatively more difficult to evaluate than assets of other financial institutions. By using as a measurement the disagreements between two major external rating agencies (Moody’s and Standard and Poor), split bond rating, he finds that the probability of disagreements is significantly higher for banks rather than other financial institutions suggesting the risk of these assets is relatively hard to quantify. Using same technique as Haggard and Howe, Iannotta (2006) examines the disagreement between rating agencies (split ratings) on the European bond market. Although he finds that overall bonds issued by banks have fewer split ratings, when controlled for risk and other issue characteristics bank bonds have higher predicted probability of a split rating. Moreover, his findings suggest that subordinated bonds are subject to more disagreement
- 6 - between rating agencies, bank opaqueness increases with financial assets and decreases with bank fixed assets and bank opaqueness increases with bank size and capital ratio.
Since banking stress started being published, after the recent financial crisis in 2008, event studies have been conducted to assess whether their outcome is useful in the eye of the investors. The results of this studies not always show agreement, thus this topic remains a source for additional research.
A research by Peristiani et al (2010) on the Supervisory Capital Assessment Program (SCAP), introduced in February 2009 in the USA, illustrated that the produced information was demanded by the market. Through an event study technique, they find that although investors could predict which banks had capital gaps they were surprised by the size of the capital gaps and they used that information to reevaluate stock prices. In their results, banks with higher capital gaps experienced relatively higher abnormal returns.
Ellahie (2012) searches equity and credit market data for effects on information asymmetry and information uncertainty of the 2010 and 2011 European banks stress tests. His findings indicate that equity and bid-ask spreads were not significantly affected by stress test announcements but declined after the disclosure of stress test results. For the same stress tests, Cardinali and Nordmark (2011) find that the announcements and the clarification of the methodology of the 2010 stress test were relatively uninformative. On the contrary, they find that the disclosure of the methodology of the 2011 stress test was highly informative for all stress-tested banks. Their findings indicate that banks are opaque to a mediocre degree.
Petrella and Resti (2012) examine whether and how the 2011 European stress test affected stock prices. Their event study analysis indicates that investors found the results of the stress test informative. They find significant market reactions both on pre-results dates and upon the release of the stress test dates. Investors did not only concentrate at the historical data which
- 7 - was released, but also showed considerable importance to variables measuring each bank’s vulnerability to the simulated downturn scenario. They also find that investors were not able to anticipate the results of the test before the announcement, which is consistent with the theory of opaqueness in the banking sector.
Alves et al (2013) also address the importance of banking stress tests. They use event study methodology to examine the stock market and the CDS market reaction on the 2010 and 2011 European stress tests. They conclude that both stress tests produced valuable information on the stock market suggesting that investors could not anticipate the results, while on the CDS market the results were partially anticipated. Specifically, from the first test they observe that both, stock market and CDS market display similar patterns before and after the test. It is not observed the same from the second test implying that the ability of CDS market to anticipate and incorporate information on prices and spreads is higher, which may derive from the fact that CDS market participants are more informed.
Neretina et al. (2014) try to quantify the effects of US stress tests on stock returns, CDS spreads, systematic risk, and systemic risk during the period 2009-2013. They find that the CDS market was affected strongly by post-crisis stress tests but the impact was weak on bank stock returns. They also investigate what was the effect of the stress test on the systematic risk through the betas of the firms. They conclude that there is a decline in betas risk in some years after the stress test results suggesting that stress tests affect systemic risk.
Acharya and Steffen (2014), in their policy brief some months before the 2014 European stress test, estimate the capital shortfalls of banks that will be stress-tested under the AQR using publicly available data and a series of shortfall measures. They conclude that beyond the expected results, some core European countries such as France and Germany have large
- 8 - shortfalls. Market measures of equity imply significantly greater capital shortfalls for France and Germany than do the book measures.
Steffen (2014) states in his report that the European Central Bank faces a trade off by conducting this stress test. ECB wants to maintain its reputation as a central bank and build its standing as an effective financial regulator by comprehensively identifying the problematic assets. While on the other hand there may be looming capital shortfalls that the comprehensive assessment did not identify, with no clearly defined backstops to ameliorate the shortfalls. Since ECB will take charge of the supervision of these banks, they might have the incentives to show a better picture. He also suspects that national regulators may be faced with incentives that induce them to not to fully disclose problematic assets to the ECB.
De Groen (2014) did a study after the 2014 stress test where he finds that a large number of primarily small- and medium-sized banks still have to raise their capital levels just to comply with the ratios that will be or are very likely to be binding as a result of the implementation of the new capital requirements. The information disclosed by the ECB was insufficient to calculate the impact of the new capital regulations on the capital ratios and to determine which own funds and debt instruments would be eligible for the minimum requirement for own funds and eligible liabilities (MREL).
- 9 -
3.
2014 Stress Test
In the 2014 stress test scenarios, EBA assumes movement in some macroeconomic indices such as growth, interest rates and unemployment. Considering those movements, they examine how the equity of each institution will respond. They use two scenarios, the baseline and the adverse. Each scenario covers a period of 3 years, from 2014 until 2016. The baseline scenario assumes that the macroeconomic growth will be the same as the European Commission predicts. The adverse scenario assumes that there will be a severe worldwide recession. Risk types considered in this stress test include credit risk, market risk, sovereign risk, securitization and cost of funding.
There is a hurdle for the Common Equity Tier 1 (CET1) ratio for each of the two scenarios that the stress tested institutions need to pass. For the baseline scenario these institutions need to have a CET1 ratio above 8%. For the adverse scenario the CET1 ratio needs to be above 5.5%.
From the 1233 banks that took part in this stress test 24 failed with capital shortfalls amounting to 24 billion euro. Some banks had already taken actions to improve their capital positions and in the period January 2014 – September 2014 managed to raise Common Equity Tier 1 Capital and reduce the overall capital shortfalls to 9.5 billion euro. From these 24 failed banks 10 had already raised capital for their needs during the period January – September 2014. EBA took into consideration assets as of end 2013. Only 14 banks from the 24 failed had actual capital deficiencies when the results were announced. Banca Monte dei Paschi di Siena S.p.A. was the bank with the highest capital shortfall of this stress test and it still needs to raise capital of 2.1 billion euro (EBA, 2014). List with the results of the 24 failed banks and their capital shortfalls is in the appendix, Table 12.
- 10 -
4.
Methodology
Every new stress test provides a valuable source of information for research. From theory it is expected that investors should find valuable new information in stress test results so they should afterwards reevaluate their risk expectations on the tested banks. The risk adjustment, consequently, it is expected to cause movements in the share prices of those banks. My hypothesis would be that after the announcement of the results, share prices would experience a price adjustment for risk due to the new information.
For this research I will follow an event study methodology. The event study methodology is the standard method of measuring security price reaction to any announcement or event. This research methodology was first introduced by Fama et al. (1969). They observed the difference in stock price behavior between stock splits days and normal days. The structure of my paper is based on an event study outline as suggested by Campbell et al. (1997). There are five main steps to identify the event study: 1st event definition, 2nd selection criteria, 3rd normal and abnormal returns, 4th estimation procedure and 5th testing procedure.
1. Event definition. The first step of conducting an event study is the definition of the event of interest and the identification of the period over which the stock prices of the banks involved will be checked. The period will be the event window of my research. The day that the ECB stress test results were announced was a Sunday (26 October 2014), not a trading day. For this reason I will use two different days as the event day, Friday (24 October 2014) and Monday (27 October 2014), for my study.
As event window I will use first a three-day window, where Friday/Monday (24/27 October 2014) is the event day, following, Peristiani et al. (2010). Although these events usually occur on a single day, it is used the event window to be expanded for two main reasons. One of the reasons is that by including the day after the event you capture price effects that occur after
- 11 - the stock market closes on the announcement day. The other is that by including the day prior to the event you can investigate for any news leakages.
2. Selection criteria. The banks included in my study will be the banks that took part in the recent stress test and from them I will use the public listed banks to see the market reaction on the specific stress test. I will obtain my data from the Thomson Reuters database.
3. Normal and abnormal returns. The normal returns are defined as the expected returns if the event did not take place. Academics have two common choices for modeling the normal return: the constant-mean-return model and the market model. The first model considers that any given security has constant mean return through time, while the second considers a stable linear relation between the market return and the security return. For the credit default swap’s spreads I will use the same methodology as the following methodology of abnormal returns.
For computing the stock returns I calculate the logarithm according to:
Ri,t= ln( Pi,t Pi,t-1)
Ri,t is the logarithm-return for bank i at time t, Pi,t is the closing price of the current day and
Pi,t-1 is the closing price of the previous day.
To avoid issues with nonstationarity in the data I use continuously compounded returns. There are quite severe econometric consequences because of nonstationarity, which can lead to unreliable test statistics and predictors. From empirical evidence we know if stock returns are integrated of order 1 they become stationary, i.e. after taking the first ln-differential (Gujarati, 2003, pp. 798-810).
- 12 - In order to estimate the market model for the calculation of the normal returns, I will follow Peristiani et al. (2010) and regress the daily stock return for each individual bank, Ri,t , on the market return, Rm,t , proxied by the return on the MSCI Europe Bank Index:
Ri,t= αi+βiRm,t+ εi,t where εi,t is the zero mean disturbance term.
After having calculated the constant and the coefficient of the model for each bank I can then find what would have been the normal return for each day of the event window by using the market return of that day. When I will have the normal return, the abnormal return can be calculated. The abnormal return is the actual observed return of the security for each day of the event window minus the previously calculated normal return. Mathematically, the relation can be expressed
ARi,t= Ri,t− E(Ri,t|Xt)
Where ARi,t is theabnormal return and E(Ri,t) is the normal return for time period t. Xt is the conditioning information for the normal model.
In order to examine stock price movements around events, each bank’s returns could be analyzed separately. However, because information unrelated to the event can also cause stock price movements, separate examination is not very informative. Using the average over a significant number of banks, we can have a more informative analysis; this average should reflect the effect of that specific event since the abnormal returns are centered around that particular event. The average cancels out all other information irrelevant to the event (de Jong, 2007). To calculate the unweighted cross-sectional abnormal returns average, AARt, in period t, I use the formula:
- 13 - AARt= 1
𝑁∑ 𝐴𝑅i,t
N i=1
4. Estimation procedure. For the calculation of the normal returns it needs to be calculated the constant and the coefficient of the market model for each bank over the so-called estimation window. As estimation window, it is usually used by scholars the period exactly prior to the particular event. However, there is no certain length for the estimation window. Brown and Warner (1980) have used a 35-month period with ending point 10 months prior to the event. Fama et al. (1969) decided to use for their research an 24 months estimation window. In both of these studies monthly data were used. According to Campbell et al. (1997), the parameters of the market model for the calculation of the normal returns could be estimated over a 120 estimation window prior to the event, if daily data is used. Peristiani et al. (2010), for their estimation window, use daily data over of a period of one year.
I will follow Peristiani et al. (2010) in order to estimate the market parameters via ordinary least squares (OLS) regressions using daily data from 22 September 2013 to 22 September 2014. Thus, the estimation window ends one month before the event – the 2014 results. If we include the event period itself in the estimation window, both the normal and the abnormal returns would reflect the impact of the event. That’s why the event period should not be included in the estimation window, since the methodology is based around the assumption that the impact of the event is seized by the abnormal returns (Campbell et al., 1997).
5. Testing procedure. In event studies, statistical tests are necessary in order to identify whether or not the calculated abnormal returns are significantly different from zero at a certain significance level. In my research the null hypothesis is:
H0: E(AARt) = 0
- 14 - H1: E(AARt) ≠ 0
A simple t-test is the most common test to approve or not the null hypothesis of zero abnormal returns. It is assumed that the abnormal returns ARi,t, which their average composes the average abnormal returns, are IID (independently and identically distributed). Moreover, it is assumed that they follow normal distribution with mean zero and variance σ2. Since σ2 is unknown, I obtained from the cross-sectional variance of the abnormal returns in period t, an estimator (
s
t) of σ: st= √ 1 N − 1∑(ARit− AARt) 2 N i=1Having found the
s
t I can use the following test statistic for the average abnormal return:𝐺 = √NAARt st ~𝑡N-1
G-test follows student-t distribution with N-1 degrees of freedom. However, there is strong evidence that stock returns do not satisfy the normality assumption imposed to derive the distribution of the G test statistic. If we keep in mind that the abnormal returns are independent and have the same mean and variance, in large samples it can be shown that, G approximately follows a standard normal distribution. This is a conclusion of the central limit theorem (de Jong, 2007). Hence, I have:
G = √NAARt
st ≈ N(0,1)
Typically, in event studies, N > 30 is considered large enough. I my research I have N=53 which is much more above this limit.
- 15 - In order to examine movements over longer periods surrounding the event, it is used to calculate the cumulative abnormal returns. CARi(τ1,τ2) is defined as the cumulative abnormal return for security i from τ1 to τ2 where Τ1<τ1<τ2<Τ2 .
The cumulative abnormal return is given by:
CARi(τ1,τ2) = ARi,t1+ ⋯ + ARi,t2= ∑ ARi,t t2
t=t1
Then I calculate the cumulative average abnormal returns, CAARi(τ1,τ2), as follows:
CAAR(τ1,τ2) = 1
𝑁∑ CARi(τ1,τ2) N
i=1
As the null hypothesis it will be:
H0:CAAR(τ1,τ2) = 0 And the two-sided alternative will be:
H1: CAAR(τ1, τ2) ≠ 0
To calculate the variance of the cumulative average abnormal returns I used the formula:
Var(CAAR(τ1, τ2)) = 𝜎̅2(τ1, τ2) = 1
𝑁2∑ 𝜎2(τ1, τ2) N
i=1
To evaluate the significance of the hypothesis I regress only the dependent variable CAR for each group. The P-value on the constant from this regression will provide me the significance of the cumulative abnormal return across all banks of the group. This test is preferable to a t-test because it allows the use of robust standard errors (Princeton, 2008).
- 16 -
5.
Data
From the 123 banks which were stress tested, 554 banks are publicly listed representing 18 European countries. Stock prices of the sample were collected for the period 22/09/2013 until 31/11/2014, the reasons have been analyzed in the methodology section. As market proxy, for calculating the market model, MSCI Europe Banks Index was used. This market capitalization weighted index is consisted of representative stocks of the European banking sector.
Out of the 123 banks 24 failed to pass the test revealing capital shortfalls. Banks already took actions that mitigate the stress test capital impact and shortfall across the sample. Overall EUR 53.6BN of Common Equity Tier 1 Capital was raised or resulted from the conversion of hybrid instruments by banks in the sample between January and September 2014. The additional capital raised by banks with a shortfall in 2014 reduces the capital needs for those banks to EUR 9.5BN and the number of banks with a shortfall to 14 (Results of 2014 EU-wide stress test). From the 55 publicly listed banks of my sample, 41 belong to the banks which passed the stress test, 14 belong to the banks which failed and from these 14 failed banks 9 still need to raise additional capital.
On the Credit Default Swaps market there are 415 banks out of the 123 that publicly traded CDS. This data was again collected through DataStream with source being Thomson Reuters. For each bank of the sample, 5 year maturity CDS spreads are collected. From the 41 banks, for which there is CDS spreads data available, 7 belong to the group of 24 banks who failed the test and 6 of them need to raise additional capital. As a market proxy was used the Itraxx Europe index 5 years maturity. This index is composed with the 125 most liquid European CDS.
4 Table 8 in appendix shows all 55 banks of the stock market sample. 5 Table 9 in appendix shows all 41 banks of the CDS market sample.
- 17 -
6.
Empirical results
6.1 Event day: Friday
Stock Market
In order to investigate the stock market reaction before and after the release of the stress test results, 4 different groups were created. Group A consists of 41 banks from the sample, that passed successfully the stress test. Group B consists of 14 banks that failed to pass the test. This separation can show whether there is a different reaction in the stock market according to the result of this stress test. Furthermore, since 10 from the 24 failed banks had already raised capital when the results were released, group of failed banks (group B) was splitted into two additional groups (group C, group D). From this separation can be seen if there was different response to the new information. Group C contains 9 out of the 14 failed of the sample. These are the banks that need to raise additional capital. The rest 5, failed banks that have already raised capital before the results were announced, compose group D. Table 1 shows the composition of each group.
Table 1. Groups composition of stock data
Group Banks Group Category
A 41 Banks which passed successfully the stress test B 14 Banks which failed the stress test
C 9 Banks which failed the stress test and need to raise additional capital D 5 Banks which failed the stress test and have already raised additional capital
Table 2 shows the results of each group for the relative event window. CAAR of each group represents the average of cumulative abnormal returns of all the banks that compose the group. The significance of CAARs was calculated by running a regression in Stata with only
- 18 - the dependent variable CAR of each bank for the relative group. The significance is shown by the P-value of the constant of this regression. This test is preferable to a t-test because it allows the use of robust standard errors (Princeton, 2008). Table 2 shows results of the full sample although they are not informative for this research.
Table 2. Stock cumulative average abnormal returns (CAAR in %) – Friday as event day CAAR (-1+1) (-2+2) (-3+3) (-5+5) (-10+10) (-10,0) (0,+10) All Sample -0.27 0.68 0.99 1.19 -2 0.43 -2.46** Group A 0.75 1.64*** 2.22*** 3.40*** 1.68 0.98 0.32 Group B -3.27 -2.15 -2.61 -5.3 -12.97*** 1.19 -10.61*** Group C -6.62** -3.58 -3.79 -7.85 -18.45*** -3.76 -13.62*** Group D 2.77 0.41 -0.49 -0.72 -3.12 3.43** -5.19**
Notes: ***, ** and * denote statistical significance at 1%, 5% and 10%, respectively. The table shows the stocks’ average cumulative abnormal returns (in %), and their statistical significance, for each event window of the stress test results and by groups of banks. Results are presented for the full sample and separately.
As it can be seen from table 2, banks which passed the test, have positive statistically significant CAARs for event windows (-2+2,-3+3,-5+5). This implies that investors were surprised by the stress test result for these banks and adjusted their positions according to the positive information.
On the contrary, banks of group B experiences a negative response on their prices. This finding is in line with the theory that stocks experience a price decrease after bad news. Except from the (-10,+10) event window, there are not significant result for this group. As it is shown later from the decomposition of this group, banks experienced different reactions on their stock price according to the results of the test for additional capital.
Analyzing further the results for banks which failed the test, it can be seen that banks which need to raise more capital had a decrease in their returns. While on the other hand banks of group D, which failed the test but had already raised capital, experienced different reaction. As it can be seen in the first two event windows, banks of group D have positive CAARs which become negative as the event window is expanded. This shows that the market reacted
- 19 - to the bad news regarding banks of group C and to the good news regarding banks of group D by adjusting relatively the stock prices.
The last two event windows of the table capture differences in the market reaction 10 days prior and 10 after the announcement of the results. From this comparison can be concluded whether investors were able to obtain information before the event (news leakage, well informed) or they adjusted their positions after the information was released. From this results can be concluded that investors made use of the new information after the
announcement of the test results. This indicates that the new information produced by the stress test is valuable for the market. These results are in accord with Neretina et al. (2014) and Alves et al. (2013).
Figure 1. This plot shows the stock cumulative abnormal returns per group for the -10,+10
event window with Friday as event day. -20% -15% -10% -5% 0% 5% 10% -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 CAA R Event days Group A Group B Group C Group D All
- 20 - Figure 1 shows the CAAR results of each group for event window (-10,+10). In this graph we can see clearly the negative market reaction for banks of groups B and C, with banks of group C having the most negative reaction. The CAARs for this groups are negative during the days before the announcement and experience a sharp decrease right after. For banks that failed the test but they had already raised capital during the period January – September 2014 (group D), graph shows a positive reaction higher than all the other groups concentrated around the event day. This can be interpreted as a positive surprise for the market for these banks.
CDS Market
For the credit default swaps market 3 groups have been created. Group A consist of banks that passed the test. Group B are banks which failed the test. Group C are banks which failed the test and they need to raise additional capital. From the banks that failed the test and they had already raised capital there is only one bank (Banco Popolare - Società Cooperativa, Italy) in the data sample, so group D could not be created. The reason group C was created although it has only one bank less than group B is because of the results of Banco Popolare - Società Cooperativa. This bank had a significant decrease in its spreads after the release of stress test results. Table 3 shows the composition of each group.
Table 3. Groups composition of CDS data
Group Banks Group Category
A 41 Banks which passed successfully the stress test B 14 Banks which failed the stress test
- 21 - Table 4 shows the results for the CDS market. It can be seen that except from the results of group C, there is no significant market reaction in the bond market. This finding goes against the complete opacity (black box) hypothesis (Morgan 2010). Results of group C show that banks which failed the test and need to raise capital experienced a mild increase in their cumulative spreads which is in line with the theory that investors see those bonds as more risky investments. Table 4 shows results of the full sample although they are not informative for this research.
Table 4. CDS cumulative average abnormal returns (CAAR in bp) – Friday as event day CAAR (-1+1) (-2+2) (-3+3) (-5+5) (-10+10) (-10,0) (0,+10)
All Sample -0.13 -0.97 -1.28 0.66 9.46** 4.59** 4.64
Group A -0.4 -0.99 -1.48 -1.74 5.95 1.76 3.92
Group B 1.19 -0.86 -0.3 12.32 26.55** 18.38** 8.12
Group C 3.17* 2.52 3.55 18.80* 30.97** 19.46** 11.9
Notes: ***, ** and * denote statistical significance at 1%, 5% and 10%, respectively. The table shows the CDS spreads average cumulative abnormal returns, and their statistical significance, for each event window of the stress test results and by groups of banks. Results are presented for the full sample and separately.
Also here, the event window (-10,+10) was splitted into pre results and after results periods. From the comparison of the two event windows can be seen that the pre result event window (-10,0) contains more significant results than the post results event window (0,+10).
Compared to the same results from the stock prices can be concluded that investors in this market could adjust their positions prior to the announcement of the results. These results are in accordance with Neretina et al. (2014) findings in the Dodd-Frank Act Stress Test results.
- 22 - Figure 2. This plot shows the CDS cumulative abnormal returns per group for the -10,+10
event window with Friday as event day.
Figure 2 shows the results of each group for event window (-10,+10). In this graph can be seen an upward trend for average cumulative abnormal spreads of banks which failed the test.
6.2 Event day: Monday
Stock Market
Since the results of this stress test were announced on Sunday, not a trading day, I used also Monday as event day. The reason for that is to see whether is a different reaction around this day than previously. For group A the results remain almost the same. For groups B and C, CAAR have captured less reaction for the first event window (-1,+1) but then CAAR increase more than previously. This can be explained as the abnormal returns of Tuesday have been
0 5 10 15 20 25 30 35 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7 8 9 10 A ve rag e c u m u lativ e ab n o rm al sp re ad s i n b p Event days All Sample Grpoup A Grpoup B Grpoup C
- 23 - low while the further we move from the event day the higher abnormal returns become. By having Monday as event day I get slightly more significant results.
Table 5. Stock cumulative average abnormal returns (CAAR in %) – Monday as event day
CAAR (-1+1) (-2+2) (-3+3) (-5+5) (-10+10) (-10,0) (0,+10) All Sample 0.02 -0.92 -0.14 1.03 -2.82 0.12 -3.33** Group A 0.58 0.4 2.12*** 3.44*** 1.62 1.51** 0.25 Group B -1.63 -4.80** -6.73** -6 -15.84*** -3.97 -13.80*** Group C -3.75 -6.84** -7.88* -9.23 -20.34*** -6.57** -16.47*** Group D 2.19 -1.11 -4.66** -0.17 -7.74*** 0.69 -9.02***
Notes: ***, ** and * denote statistical significance at 1%, 5% and 10%, respectively. The table shows the stocks’ average cumulative abnormal returns (in %), and their statistical significance, for each event window of the stress test results and by groups of banks. Results are presented for the full sample and separately.
CDS Market
Table 6 shows the results of the event study for the bond market. This time Monday is considered as the event day. As can be seen in the table these results do not differ significantly than the previous results, when event day was Friday.
Table 6. CDS cumulative average abnormal returns (CAAR in bp) – Monday as event day
CAAR (-1+1) (-2+2) (-3+3) (-5+5) (-10+10) (-10,0) (0,+10)
All Sample -0.87 -0.23 0.3 0.64 9.19** 3.53* 5.58**
Group A -0.76 -0.27 0.18 0.28 5.76 0.73 4.80*
Group B -1.39 -0.06 0.9 2.35 25.87** 17.14** 9.39
Group C 1.24 3.08 4.66 6.73 30.18** 19.42** 12.71
Notes: ***, ** and * denote statistical significance at 1%, 5% and 10%, respectively. The table shows the CDS spreads average cumulative abnormal returns, and their statistical significance, for each event window of the stress test results and by groups of banks. Results are presented for the full sample and separately.
- 24 -
7.
Conclusion
In this paper was investigated, by using event study methodology, whether or not the results of the 2014 EU-wide stress test provided new information to the stock market and the credit default swap market. Moreover the results of each market are compared in order to determine if markets display similar resilience and anticipation ability before new information is revealed. Four groups were examined: banks which passed the test, banks which failed the test, banks which failed the test and had to raise additional capital and banks which failed the test but they had already raised capital when the results of the stress test were released.
After examination of the results, it can be concluded that the stock market was surprised by the stress test results while the bond market showed a mild response. There was a positive significant reaction for the stocks of banks which passed the test and a negative significant response for the stocks of banks which failed the test and they need to raise additional capital.
On the bond market there was no significant reaction from investors, except the expanded event window. This shows that investors of this market were not surprised by the results of the stress test but they were able to anticipate those results. Alves et al. (2013) gives two reasons which might explain this behavior. One reason may be the fact that CDS market participants are more informed; in the equity market, larger fraction of retail investors do exist, who are more prone to be influenced by waves of sentiment. The other reason he suggests is the fact that in the CDS market prices reflect the credit risk of the institution, or the probability of the institution going bankrupt or suffering a major restructuring process due to the likelihood of an upcoming credit event, whilst in the stock market prices reflect the value of the company.
- 25 -
List of References
Acharya, V., Engle, R., & Pierret, D. (2014). Macroprudential stress tests should not rely on regulatory risk weights. VOX: Research-based policy analysis and commentary from
leading economists.
Acharya, V. V., & Steffen, S. (2014). Falling short of expectations? Stress-testing the European banking system. Stress-Testing the European Banking System (January 15,
2014). CEPS Policy Brief, (315).Brown, S.J. and Warner, J.B., 1985. Using daily stock returns – the case of event studies. Journal of Financial Economics, 14(1), pp. 3-31.
Barucci, E., Baviera, R., & Milani, C. (2014). Is the Comprehensive Assessment really comprehensive?. Available at SSRN 2541043
Basel Committee. (2005). Compliance and the compliance function in banks.Basel
Committee.
Berlin, M., & Loeys, J. (1988). Bond covenants and delegated monitoring.The Journal of
Finance, 43(2), 397-412.
Bernanke, B. S. (2013, April). Stress Testing Banks: What Have We Learned?. InMaintaining Financial Stability: Holding a Tiger by the Tail (conference volume).
Calomiris, C. W. (2012). How To Regulate Bank Capital.National Affairs, (10), 41-57.
Campbell, J. Y., Lo, A. W., & MacKinlay, A. C. (1997). Event-study analysis.The
Econometrics of Financial Markets, 149-180.Cardinali, Andrew, and Jakob Nordmark. "How informative are bank stress tests." Bank Opacity in the European Union. Lund University (2011).
- 26 - Campbel, T. S., & Kracaw, W. A. (1980). Information production, market signalling, and the theory of financial intermediation. The Journal of Finance,35(4), 863-882.
De Groen, W. P. (2014). Was the ECB's Comprehensive Assessment up to Standard?. CEPS Policy Briefs, (325).
De Jong, F. (2007). Event studies methodology. Lecture Notes.
Diamond, D. W. (1991). Monitoring and reputation: The choice between bank loans and directly placed debt. Journal of political Economy, 689-721.
Ellahie, A. (2013). Capital market consequences of EU bank stress tests.Available at
SSRN 2157715.
European Banking Authority ( 2014). “Results of 2014 EU-wide stress test”.[online]. Available at: <https://www.eba.europa.eu/documents/10180/669262/2014+EU-wide+ST-aggregate+results.pdf>
Fama, E. F., Fisher, L., Jensen, M. C., & Roll, R. (1969). The adjustment of stock prices to new information. International economic review, 10(1), 1-21.
Flannery, M. J., Kwan, S. H., & Nimalendran, M. (2004). Market evidence on the opaqueness of banking firms’ assets. Journal of Financial Economics,71(3), 419-460. Gujarati, D. N. (2003). Basic Econometrics. 4th. New York: McGraw-Hill.
Haggard, K. S., & Howe, J. S. (2007). Are banks opaque?. International Review of
Accounting, Banking and Finance ().
Iannotta, G. (2006). Testing for opaqueness in the European banking industry: evidence from bond credit ratings. Journal of Financial Services Research, 30(3), 287-309.
- 27 - Jones, J. S., Lee, W. Y., & Yeager, T. J. (2012). Opaque banks, price discovery, and financial instability. Journal of Financial Intermediation, 21(3), 383-408.
Jordan, J. S., Peek, J., & Rosengren, E. S. (2000). The market reaction to the disclosure of supervisory actions: Implications for bank transparency.Journal of Financial
Intermediation, 9(3), 298-319.
Morgan, D. P. (2002). Rating banks: Risk and uncertainty in an opaque industry. The
American Economic Review, 92(4), 874-888..
Neretina, E., Sahin, C., & De Haan, J. (2015). Banking stress test effects on returns and risks.
Peristiani, S., Morgan, D. P., & Savino, V. (2010). The information value of the stress test and bank opacity. FRB of New York Staff Report, (460).
Petrella, G., & Resti, A. (2013). Supervisors as information producers: do stress tests reduce bank opaqueness?. Journal of Banking & Finance,37(12), 5406-5420.
Steffen, S. (2014). Robustness, validity and significance of the ECB's asset quality review and stress test exercise. Available at SSRN 2517361.
http://www.bloomberg.com/news/articles/2014-12-02/european-banks-seen-afflicted-by-82-billion-capital-gap-in-test
- 28 -
Appendix
Table 7. 123 Europeans banks that participated in the 2014 European stress test
Institution Name Country Institution Name Country
BAWAG P.S.K. Austria Banca Popolare Dell'Emilia Romagna Italy
Erste Group Bank Austria Banca Popolare Di Milano Italy
Raiffeisenlandesbank Niederösterreich Austria Banca Popolare di Sondrio Italy Raiffeisenlandesbank Oberösterreich Austria Banca Popolare di Vicenza Italy
Raiffeisen Zentralbank Austria Banca Carige. Italy
Österreichische Volksbanken Austria Credito Emiliano Italy AXA Bank Europe Belgium Banca Piccolo Credito Valtellinese Italy
Belfius Banque Belgium Iccrea Holding Italy
The Bank of New York Mellon Belgium Intesa Sanpaolo. Italy
Dexia NV Belgium Mediobanca Italy
KBC Group NV Belgium Banca Monte dei Paschi di Siena Italy Bank of Cyprus Public Company Cyprus Unione Di Banche Italiane Italy
Co-operative Central Bank Cyprus UniCredit Italy
Hellenic Bank Public Company Cyprus Veneto Banca S.C.P.A. Italy
Danske Bank Denmark ABLV Bank Latvia
Jyske Bank Denmark Banque et Caisse d'Epargne de l'Etat, Luxemburg
Nykredit Denmark Precision Capital. Luxemburg
Sydbank Denmark Bank of Valletta Malta
OP-Pohjola Group Finland ABN Amro Bank N.V. Netherlands
BNP Paribas France Bank Nederlandse Gemeenten N.V. Netherlands
Groupe BPCE France ING Bank N.V. Netherlands
BPI France France Nederlandse Waterschapsbank N.V. Netherlands Groupe Crédit Agricole France Coöperatieve Centrale Raiffeisen Netherlands
Groupe Crédit Mutuel France SNS Bank N.V. Netherlands
Caisse de Refinancement de l’Habitat France DNB Bank Group Norway
La Banque Postale France Alior Bank Poland
Banque PSA Finance France Bank BPH Poland
RCI Banque France Bank Handlowy W Warszawie Poland
Société de Financement Local France Bank Ochrony Srodowiska Poland
- 29 -
Institution Name Country Institution Name Country
Aareal Bank Germany Powszechna Kasa Oszczednosci Bank
Polski
Poland Deutsche Apotheker- und Ärztebank Germany Banco Comercial Português Portugal
Bayerische Landesbank Germany Banco BPI Portugal
Commerzbank Germany Caixa Geral de Depósitos Portugal
Deutsche Bank Germany Nordea Bank Sweden
DekaBank Deutsche Girozentrale Germany Skandinaviska Enskilda Banken Sweden DZ Bank AG Deutsche Zentral Germany Svenska Handelsbanken Sweden
HASPA Finanzholding Germany Swedbank Sweden
HSH Nordbank Germany Nova Kreditna Banka Maribor Slovenia
Münchener Hypothekenbank Germany Nova Ljubljanska banka Slovenia Hypo Real Estate Holding Germany SID - Slovenska izvozna in razvojna banka Slovenia IKB Deutsche Industriebank Germany Banco Financiero y de Ahorros Spain KfW IPEX-Bank GmbH Germany Banco Bilbao Vizcaya Argentaria Spain
Landesbank Berlin Holding Germany Bankinter Spain
Landesbank Baden-Württemberg Germany Banco Mare Nostrum Spain Landesbank Hessen-Thüringen
Girozentrale Germany Banco de Sabadell Spain
Landeskreditbank
Baden-Württemberg-Förderbank Germany Cajas Rurales Unidas Spain
Landwirtschaftliche Rentenbank Germany Catalunya Banc Spain Norddeutsche Landesbank-Girozentrale Germany Caja de Ahorros y M.P. de Zaragoza Spain
NRW.Bank Germany Kutxabank Spain
Volkswagen Financial Services Germany Caja de Ahorros y Pensiones de Barcelona Spain WGZ Bank Westdeutsche Genossenschafts Germany Liberbank Spain
Wüstenrot Bausparkasse Germany NCG Banco Spain
Wüstenrot Bank AG Pfandbriefbank Germany Banco Popular Español Spain
Alpha Bank Greece Banco Santander. Spain
Eurobank Ergasias Greece Unicaja Banco Spain
National Bank of Greece. Greece Barclays U.K.
Piraeus Bank Greece HSBC Holdings U.K.
OTP Bank Hungary Lloyds Banking Group U.K.
Allied Irish Banks Ireland Royal Bank of Scotland Group U.K. The Governor and Company of the Bank of
Ireland Ireland
Permanent tsb. Ireland
- 30 - Table 8. List of 55 Banks of the stock return sample
Institution Name Country Institution Name Country
Erste Group Bank Austria Banca Piccolo Credito Valtellinese Italy
Raiffeisen Zentralbank Austria Intesa Sanpaolo Italy
KBC Group NV Belgium Banca Monte dei Paschi di Siena italy
Dexia NV Belgium Mediobanca Italy
Hellenic Bank Public Company Cyprus UniCredit Italy
Danske Bank Denmark Bank of Valletta Malta
Jyske Bank Denmark ING Bank N.V. Netherlands
Sydbank Denmark DNB Bank Group Norway
BNP Paribas France Alior Bank Poland
Groupe Crédit Agricole France Bank BPH Poland
Société Générale France Handlowy Poland
Deutsche Bank Germany Getin Noble Bank Poland
Commerzbank Germany Banco Comercial Português Portugal
Aareal Bank Germany Banco BPI Portugal
IKB Deutsche Industriebank Germany Banco Santander Spain
Alpha Bank Greece B.B.V. Argentaria Spain
National Bank of Greece Greece Bankinter Spain
Eurobank Ergasias Greece Banco de Sabadell Spain
Piraeus Bank Greece Liberbank Spain
OTP Bank Hungary Banco Popular Español Spain
Allied Irish Banks Ireland Nordea Bank Sweeden
Permanent tsb. Ireland Svenska Handelsbanken Sweeden
Banco Popolare Italy Swedbank Sweeden
Banca Popolare Dell'Emilia Romagna Italy Barclays U. K.
Banca Popolare Di Milano Italy HSBC Holdings U. K.
Banca Popolare di Sondrio Italy Lloyds Banking Group U. K.
Banca Carige Italy Royal Bank of Scotland Group U. K.
- 31 - Table 9. List of 41 banks of the CDS spreads sample.
Bank Name Country Test Status
1 BAWAG P.S.K. Austria Passed
2 Erste Group Bank Austria Passed
3 Raiffeisen Zentralbank Österreich Austria Failed
4 AXA Bank Europe Belgium Failed
5 Dexia NV Belgium Passed
6 KBC Group NV Belgium Passed
7 Danske Bank Denmark Passed
8 BNP Paribas France Passed
9 Groupe Crédit Agricole France Passed
10 Banque PSA Finance France Passed
11 Société Générale France Passed
12 Bayerische Landesbank Germany Passed
13 Commerzbank Germany Passed
14 Deutsche Bank Germany Passed
15 HSH Nordbank Germany Passed
16 IKB Deutsche Industriebank Germany Passed
17 Alpha Bank Greece Failed
18 Eurobank Ergasias Greece Failed
19 National Bank of Greece Greece Passed
20 Allied Irish Banks Ireland Passed
21 The Governor and Company of the Bank of Ireland Ireland Failed
22 Permanent tsb Ireland Passed
23 Banco Popolare Italy Passed
24 Intesa Sanpaolo Italy Passed
25 Mediobanca Italy Failed
26 Banca Monte dei Paschi di Siena Italy Passed
27 Unione Di Banche Italiane Italy Passed
28 ING Bank N.V. Netherlands Passed
29 SNS Bank N.V. Netherlands Passed
30 DNB Bank Group Norway Passed
31 Nordea Bank Sweden Passed
32 Skandinaviska Enskilda Banken Sweden Passed
33 Svenska Handelsbanken Sweden Passed
34 Swedbank Sweden Passed
35 Bankinter Spain Passed
36 Banco de Sabadell Spain Passed
37 Banco Popular Español Spain Passed
38 Banco Santander Spain Passed
39 Barclays U.K. Passed
40 HSBC Holdings U.K. Passed
- 32 - Table 10. Bank composition of groups from the stock market sample
Group A
1 Erste Group Bank 15 OTP Bank Ltd 29 Banco Santander
2 Raiffeisen Zentralbank 16 Allied Irish Banks 30 B.B.V. Argentaria
3 KBC Group NV 17 Credito Emiliano 31 Bankinter
4 Danske Bank 18 Intesa Sanpaolo 32 Banco de Sabadell
5 Jyske Bank 19 Mediobanca 33 Liberbank
6 Sydbank 20 UniCredit 34 Banco Popular Español
7 BNP Paribas 21 Bank of Valletta 35 Nordea Bank
8 Groupe Crédit Agricole 22 ING Bank N.V. 36 Svenska Handelsbanken
9 Société Générale 23 DNB Bank Group 37 Swedbank
10 Deutsche Bank 24 Alior Bank 38 Barclays plc
11 Commerzbank 25 Bank BPH 39 HSBC Holdings
12 Aareal Bank 26 Handlowy 40 Lloyds Banking Group
13 IKB Deutsche Industriebank 27 Getin Noble Bank 41 Royal Bank of Scotland Group
14 Alpha Bank 28 Banco BPI
Group B
1 Dexia NV 6 Permanent tsb plc. 11 Banca Carige S.P.A.
2 Hellenic Bank Company 7 Banco Popolare 12 Banca Valtellinese
3 Eurobank Ergasias 8 Dell'Emilia Romagna 13 Banca Monte dei Paschi Siena 4 National Bank of Greece 9 Banca Popolare Di Milano 14 Banco Comercial Português 5 Piraeus Bank 10 Banca Popolare di Sondrio
Group C
1 Dexia NV 4 Eurobank Ergasias 7 Banca Carige S.P.A.
2 Hellenic Bank Public Company 5 Permanent tsb plc. 8 Banca Monte dei Paschi Siena 3 National Bank of Greece 6 Banca Popolare Di Milano 9 Banco Comercial Português
Group D
1 Piraeus Bank 4 Banca Popolare di Sondrio
2 Banco Popolare 5 Banca Valtellinese
- 33 - Table 11. Bank composition of groups from the stock market sample
Group A
1 BAWAG P.S.K. 18 Intesa Sanpaolo
2 Erste Group Bank 19 Mediobanca - Banca di Credito Finanziario 3 Raiffeisen Zentralbank 20 Unione Di Banche Italiane Società Cooperativa
4 KBC Group NV 21 ING Bank N.V.
5 Danske Bank 22 SNS Bank N.V.
6 BNP Paribas 23 DNB Bank Group
7 Groupe Crédit Agricole 24 Nordea Bank
8 Banque PSA Finance 25 Skandinaviska Enskilda Banken
9 Société Générale 26 Svenska Handelsbanken
10 Bayerische Landesbank 27 Swedbank
11 Commerzbank 28 Bankinter
12 Deutsche Bank 29 Banco de Sabadell
13 HSH Nordbank 30 Banco Popular Español
14 IKB Deutsche Industriebank 31 Banco Santander
15 Alpha Bank 32 Barclays
16 Allied Irish Banks 33 HSBC Holdings
17 The Governor and Company of the Bank of Ireland 34 Lloyds Banking Group
Group B Group C
1 AXA Bank Europe 1 AXA Bank Europe
2 Dexia NV 2 Dexia NV
3 Eurobank Ergasias 3 Eurobank Ergasias
4 National Bank of Greece 4 National Bank of Greece
5 Permanent tsb 5 Permanent tsb
6 Banco Popolare 6 Banca Monte dei Paschi di Siena
- 34 - Table 12. List of 24 failed banks of the 2014 stress test
CET1 ratio
Bank Name Country Baseline Adverse
Stress test shortfalls
Capital needs
1 Österreichische Volksbanken Austria 7.2% 2.1% 0.86 0.86
2 AXA Bank Europe Belgium 12.7% 3.4% 0.20 0.07
3 Dexia NV Belgium 10.8% 5.0% 0.34 0.34
4 Bank of Cyprus Public Company Cyprus 12.9% 1.5% 0.92 -
5 Co-operative Central Bank Cyprus 0.5% -8.0% 1.17 -
6 Hellenic Bank Public Company Cyprus 9.1% -0.5% 0.28 0.18
7 Münchener Hypothekenbank Germany 5.8% 2.9% 0.23 -
8 Caisse de Refinancement de l’Habitat France 5.7% 5.5% 0.00 -
9 Eurobank Ergasias Greece 2.0% -6.4% 4.63 1.76
10 National Bank of Greece Greece 5.7% -0.4% 3.43 0.93
11 Piraeus Bank Greece 9.0% 4.4% 0.66 -
12 Permanent tsb. Ireland 8.8% 1.0% 0.85 0.85
13 Banca Carige Italy 2.3% -2.4% 1.83 0.81
14 Banca Monte dei Paschi di Siena. Italy 6.4% -0.1% 4.25 2.11
15 Banca Piccolo Credito Valtellinese Italy 7.1% 3.5% 0.38 -
16 Banca Popolare Dell'Emilia Romagna Italy 8.3% 5.2% 0.13 -
17 Banca Popolare Di Milano Italy 6.9% 4.0% 0.68 0.17
18 Banca Popolare di Sondrio Italy 7.4% 4.2% 0.32 -
19 Banca Popolare di Vicenza Italy 7.7% 3.2% 0.68 0.22
20 Banco Popolare Italy 6.7% 4.7% 0.43 -
21 Veneto Banca Italy 5.9% 2.7% 0.71 -
22 Banco Comercial Português Portugal 8.8% 3.0% 1.14 1.15
23 Nova Kreditna Banka Maribor Slovenia 12.8% 4.4% 0.03 0.03
24 Nova Ljubljanska banka Slovenia 12.8% 5.0% 0.03 0.03
“Notes: This table shows the 24 banks that failed the stress test and their relative CET1 capital ratio under each scenario. Capital shortfalls are in billion Euro.”
