Capital ratios and the influence on Banks' stock performances; an empirical research on the difference between the weighted and unweighted capital ratios.

Capital ratios and the influence on Banks' stock performances; an

empirical research on the difference between the weighted and

unweighted capital ratios.

In this research an in-depth analysis is provided on how capital ratios, based on Basel III regulation, influence the stock performances of European banks. Using panel data between 2010 and 2015 the perception on weighted and unweighted ratios by investors is measured. The findings show that higher quality equity (Tier 1) has a significant positive influence on the stock performances. Furthermore, no evidence is found that investors value weighted and unweighted ratios different. Lastly, Risk Weighted Assets (RWA) density is introduced and shows that for unweighted Tier 1 capital ratio, low dense banks perform better than average dense banks. The findings can be used for decision-making on future capital products and provides, for the first time, quantitative research on RWA density.

Keywords: Capital Ratios, Basel III, CRD IV, CRR, RWA density, Financial institutions

JEL Classification: E58 – G21 – G28

1 University of Groningen, Faculty of Economics and Business Student number: 1989286

Table of Contents

Appendix A: Basel III outline 40

Pillar 1 42

Pillar 2 46

Pillar 3 46

Appendix B: Key concerns of multiple stakeholders on RWA. Cited from Le Le Leslé and Avramova

(2012, p. 7-8) 47

Abbreviations

A-IRB: Advanced Internal Rate Based BIS: Bank for International Settlement CET 1: Common Equity Tier 1

CRD: Capital Requirements Directive CRR: Capital Requirements Regulation EBA: European Banking Authority EU: European Union

F-IRB: Fundamental Internal Rate Based LCR: Liquidity Coverage Ratio

LGD: Lost Given Default MBS: Market Backed Securities NFA: New Financial Architecture NSFR: Net Stable Funding Ratio PD: Probability of Default RWA: Risk Weighted Assets SA: Standardized Approach SIV: Secured Investment Vehicle

Preface

1. Introduction

With the collapse of Lehman Brothers in 2008, the full extent of the Global Financial Crisis was uncovered. The crisis exposed the fragility of the financial sector. This fragility was partly caused by what is referred to as “New Financial Architecture” (NFA), a global and deeply interwoven system of large banks and a shadow banking systems formed by hedge funds, Special Investment Vehicles (SIV) and investment banks. The NFA provided incentives for excessive risk taking. Furthermore, there was a lack of regulation, and excessive risk was build up by the trade of Mortgage Backed Securities (MBS). Eventually, this led to the crash and insolvability of many banks (Crotty, 2009).

can be described as a global regulatory framework, which tries to achieve a system with more resilient banks and banking systems (BIS, 2011).

In 2010, the Basel Committee on banking supervision published their first version of Basel III. The reforms focus on improving the ability of the banking sector to absorb shocks, arising from financial and economic stress. By improving these shock absorbing capacities, the Basel Committee wants to reduce the risk of spill-over from the financial sector to the real economy (BIS, 2011). The regulation aims to improve the risk management and governance of financial institutions and to increase the transparency and disclosure. The reforms proposed in Basel III target on two levels, a micro prudential level and a macro prudential level. The first means on bank level and is focused on increasing the resilience of individual banking institutions. The latter is on a system wide level and targets the risk across the banking sector and the pro-cyclical amplification of these risks over time. The Basel III framework is based on three pillars as a basis for the regulations, the first pillar focuses on the minimum capital requirements, the second on the supervisory review process and the last on market discipline. A more in-depth analysis of the Basel III framework can be found in Appendix A, in which the differences between the three pillars are explained and the most important implications are highlighted.

Part of the Basel framework are the capital ratios, these capital ratios can be split into two groups; weighted and unweighted. Weighted capital ratios use the risk weighted assets (RWA) of a bank as denominator and unweighted capital ratios do not adjust the assets of a bank for the risk it is exposed to, so for those ratios the total assets are used as denominator. The nominator for both ratios is the equity of the bank. This equity is defined into multiple groups, depending on the robustness of it. Those groups are Common Equity Tier 1 (CET 1), Tier 1 and Tier 2, in which the CET 1 contains the highest quality of equity.

Following Basel II, the focus was solely on the weighted capital ratios. However, the reforms proposed in Basel III shifted more towards the unweighted capital ratios. The underlying idea is that this can be used as a backstop measure in order to increase the resilience of the banks (BIS, 2011). In 1999 the Basel Committee already noticed that banks were finding ways to avoid the limitation that fixed capital requirements put on the amount of risk they can take for the corresponding capital levels. They stated that “this is undoubtedly starting the comparability and even the meaningfulness of the capital ratios maintained.”

(BIS,1999). Since the financial crisis, there is an increased focus on the capital ratios of banks and more data about those ratios has become available.

Furthermore, investors are starting to doubt the robustness of the weighted capital ratios due to their complexity and opacity. Research by Barclays (2014) has shown that the trust of investors in the RWA is decreasing. They find that 63% of the investors found the RWA less trusted than before. For 33% it stayed the same and only 4% indicated that their trust has risen. Investors find risk weight calculations particularly difficult to scrutinize and appear to be losing confidence in the accuracy of the RWA as a result (Bank Of England, 2012). Banks can increase their weighted capital ratios in two ways. By increasing the numerator, which is the capital, or by decreasing the denominator, which is the RWA. After the crisis most banks were able to increase their weighted but not their unweighted capital ratios (Das & Sy, 2012).

(Demirguc-Kunt, Detragiache and Merrouche, 2012; Hogan, 2015; Yang and Tsatsaronis, 2012; Miles, Yang and Marcheggiano, 2012). Those studies are mainly focused on the capital ratios as implied in Basel II, which are risk based, and use data up to 2012, which is before the start of the phase-in of Basel III. Based on this, the following research question is tried to be answered in this paper.

To which extend are the stock performances of banks impacted by their capital ratios?

To answer the research, question multiple sub-questions are formulated which will be used to capture the differences between ratios based on weighted assets and unweighted assets. Due to the large difference in legal and regulatory frameworks between countries, the focus of this study will be on Europe, which is defined as the countries that have implemented Capital Requirements Directive (CRD IV) and the Capital Requirements Regulation (CRR). This is the translation of the Basel III accords into European Union (EU) regulation. Since the financial crisis of 2008, there is an increased focus on capital ratios and in 2010 the outline of Basel III was presented. Based on this, the period of this research will be between 2010 and 2015.

extent the weighted and unweighted ratio influences the stock performances, the following sub-question will be researched:

To which extent does the risk weighted and unweighted capital ratios influence the stock

performances?

From here on, the focus will be on the difference between the weighted and unweighted capital ratios. From the investors side there is a distrust in the correctness of the weighted capital ratios due to the optimization of the RWA (Mariathasan & Merrouche, 2014). In order to see whether investors put more value in the unweighted ratio, the following sub-question is formed:

Is the impact of unweighted capital ratios larger than the weighted capital ratios on the stock

performances of banks?

To incorporate the difference between a bank’s weighted and its unweighted capital ratios, the RWA density is incorporated in the model. As explained before, the RWA density shows the riskiness of the bank’s assets. In order to research this, groups of banks will be

formed. Based on the RWA density of the bank in a certain quarter, the banks are grouped in either high, normal or low dense, for which the density compared with the other banks is used as benchmark. The upper and lower boundary of the normal group are calculated by the mean plus and minus half a standard deviation. Banks that fall outside are grouped in either the low or high group. This is all done to answer the last sub-question:

The findings of this research will contribute to multiple fields. The main goal of this research is to find whether or not banks with high capital ratios have better stock performances when compared to low capitalized banks. The relation that is researched focuses on how investors perceive capital ratios of banks after the financial crisis. At the moment of writing there was no research available that focused specifically on the period after the crisis, which is characterized by a decreasing trust in the RWA and an increasing focus on weighted and unweighted capital ratios, both by investors and regulators. Besides, this research will contribute to the literature on RWA density. Currently, only a few articles mention RWA density and none of the previous papers has incorporated the RWA density ratio into quantitative analysis, to see whether there is an influence, and if yes which effect this ratio has on the stock performances of banks. Furthermore, the outcomes can be used by banks to see whether their RWA calculations need to be more transparent and whether there is a lower and upper bound of RWA density. This last part is also useful for regulators, since it will provide them with usable boundaries for future capital requirements.

2. Theoretical background

Banks’ capital levels and the capital ratios are a well-researched subject. Shrieves and Dahl (1992) find a positive association between changes in risk and capital. This implies that capital requirements help to limit the total risk exposure. Diamond and Rajan (2000) identify the effects of minimum capital requirements, as they state that a binding capital requirement makes the bank saver and increases the ability for banks to pay outside investors. Besides, they state that an increase in capital requirements will cause a credit crunch for the cash poor and can create an increase in debt for the cash rich. With this risk of a “credit crunch” for the

banks that have lower capital ratios, it could imply that the stocks of those banks will perform suboptimal because of the expected credit crunch risk of those parties. Following this research, it could be argued that investors prefer stock of better capitalized banks, due to this increase demand for such stocks, it could be argued that those banks will perform better on the stock market.

Since the introduction of Basel II in 2004 there is an increased focus on the influence of capital ratios and the risk exposure of banks. Miles, Yang and Marcheggiano (2012) try to estimate the long-run costs and benefits from the increased capital requirements. They find that the optimal amount of equity capital is higher than the ratios banks had in the past, and that even the new requirements of Basel III are not sufficient. Furthermore, they test whether the Modigliani-Miller theorem holds for banks. This theorem states that without distortions, it makes no difference whether a firm finances itself with debt or equity. They conclude that this theorem does not hold for banks. This implies that a higher bank capital can increase the cost for the bank, which could influence the stock return of banks that are better capitalized.

Tier 1 capital perform better during the crisis. Yang and Tsatsaronis (2012) use the Fame & French three factor model on a sample of 50 banks and find that equity market rewards leverage with higher returns, this leverage also creates higher stock price volatility. However, the stricter capital rules reduce the leverage and lower the required return in the stock market. The most important article for this research is the article “Bank Capital: Lessons from the Financial Crisis” (Demirguc-Kunt et al., 2012). In their research, they study whether or not better-capitalized banks experienced higher stock returns during the financial crisis. They differentiate among multiple types of capital ratios; the Basel risk-adjusted ratio, the leverage ratio, the Tier 1 and Tier 2 ratios, and the tangible equity ratio. They construct two ratios, namely one based on RWA and one on unweighted assets. Their findings show that before the financial crisis, differences in capital did not have much impact on stock return. However, during the crisis period there is no significant influence of the risk weighted ratios on the stock performances. Furthermore, they find that there is a significant positive effect of the leverage ratio and the risk weighted ratio on the stock performances of larger banks that have at least one billion dollar in operating income.

Chan-Lau, Liu & Schmittmann (2015) investigate the effect of capital ratios on a worldwide sample of banks between July 2006 and July 2011. They do not find evidence that the Tier 1 ratio influences the stock performances. However, they find, pre-crisis and post-crisis, that banks with lower leverage have significant higher stock performances. This finding supports the finding of Demirguc-Kunt et al., (2012) that unweighted capital ratios have a positive influence on stock performances of banks.

Demirguc-Kunt et al., (2013), Das & Sy (2012), Chau-Lau et al. (2015), Hogan (2015) and Beltratti & Stulz (2012) a model based on bank and country characteristics is used. In this research the latter method is used. In the method section the multiple models will be explained. This method is preferred because it is usable, with minor changes, for answering all the sub-questions. Based on the literature one would expect that the capital ratios have a positive influence on the stock performances of banks in the period after the crisis. Furthermore one could assume that this effect differs between the capital ratios.

However, theory is less clear on the difference between the influence of unweighted and weighted capital ratios. In order to calculate the RWA for the weighted capital ratios, banks can use two methods, namely the standardized approach (SA) and the internal rate based (IRB) approach. The difference between those two methods is that in the SA the assets of the bank are weighted using a given framework and ratings from rating agencies. In the internal rate based (IRB) approach the bank has more influence to provide input. In this approach there are two methods, the advanced IRB (A-IRB) and the foundation IRB (F-IRB). Figure 1 shows the two approaches and to which extent the bank can calculate the values.

The SA used predefined risk weights for assets based on their credit ratings. The IRB method uses the model K, a standard model provided by the ECB, to calculate the capital required.

The three blocks between the brackets in figure 1 represent the input for this model. The F-IRB provides the bank with the opportunity to calculate the probability of default (PD). This is the probability that the counterparty will default within one year. In the F-IRB this probability is calculated by the bank, using a model approved by their central bank. The exposure at default (EAD), lost given default (LGD) and maturity (M) are provided by their central bank. The EAD reflects the amount expected to be outstanding when counterparty defaults, this is after any credit risk mitigation. It is important to note that the EAD reflects used balances and the commitments for undrawn amounts. The LGD is the estimated ratio of the loss on an exposure upon default of the counterparty. The M is based on the profile of a loan’s contractual principal payments over time. The more advanced approach to calculate the

RWA is the A-IRB. Under this method, the bank must calculate the effective maturity and provide their own estimates on EAD, LGD and PD (BIS, 2001a). Almost all of the larger, equity above 5 billion euros, European banks have shifted from the standardized approach to calculate their RWA towards an A-IRB approach.

the different ratios will be compared. The Tier 1 ratio and Tier 1 and 2 ratio will be compared using the method expressed by Hogan (2015). In the method part of this research, a more detailed explanation will be given on how the model of Hogan(2015) works. Based on the literature it is expected that the ratios using unweighted assets have a stronger positive influence on the stock performances.

The last step in this research is to research the influence of the RWA density. The RWA density is the ratio between RWA and total assets (RWA/Total Assets). As mentioned earlier on, from a research point of view, there is an increased focus on the weighted capital ratios. Furthermore, there is also an increase in the attention for unweighted capital ratios (Demirguc-Kunt et al., 2012). However, there is almost no research available on the difference between the unweighted ratios of a bank and the weighted ratios. Hogan (2015) use both ratios to find which one has a larger influence on the risk of the company, in which the risk is measured as the beta. This research does highlight part of the relation between a banks weighted ratio and unweighted ratio. At the moment of writing, little is known about the relation between RWA density and stock performances.

In 2012 Le Leslé and Avramova published a paper in which they highlight the usefulness of the RWA density. They find that the density differs per region. This is explained by the differences in regulatory framework, for example, in the EU it is more common to use A-IRB for the calculation of RWA. In America the standardized approach is often used. Furthermore, supervisory frameworks differ per region and so are the accounting and legal frameworks. Recently there is a shift in the interpretation of the RWA density of banks. The RWA density can be seen as a good indicator of the risk profile of a bank. Typically, banks with low RWA density, closer to zero, hold riskier assets. Appendix B provides an overview of the different stakeholders and their specific concerns regarding RWA.

IRB approach for their credit portfolio, become less RWA dense once IRB is approved. The effect cannot be explained by modelling choices, or improved risk-measurement alone. They find, in line with the theory of risk-weight manipulation, a decline in risk-weights mostly among weakly capitalized bank. Based on the fact that the equity of a bank is used as backstop in cases of shortages it is important to note that regulators and market participants should prefer banks with a low density. However, due to the discrepancy of the RWA and the distrust of investors and regulators, the perception has changed dramatically (Bruno, Nocera & Resti, 2014). Le Leslé and Avramova (2012) state that many investors view higher RWA’s as more reliable. This implies that banks with high RWA density are an indication of more prudent risk measurement, where banks are less likely to optimize the computation of their risk-based capital ratios. They find that the A-IRB average RWA density is around 1.6 times lower than in the RWA density for banks that use SA. However, Das & Sy (2012) find that, during the 2007-2008 crisis period, for every one percentage point higher RWA density the stock return is 0.06 percentage points lower. With this ratio, the risk of a bank’s assets can be put in perspective. The riskier the assets of the bank are, based on the SA or IRB, the closer the ratio will be to 1. Using this ratio will shed light on whether or not this ratio has a positive or negative influence on the relation between the risk weighted capital ratio and leverage ratio on the stock performances.

3. Methodology

Demirguc-Kunt et al., (2012). Equation 2 will be used for testing the first hypotheses for the weighted and unweighted capital ratios.

𝑦_𝑖𝑡 = 𝛼 + 𝛽₁𝑅𝑎𝑡𝑖𝑜_𝑖𝑡−1+γ₁𝐿𝑖𝑞_𝑖𝑡−1+γ₂𝐴𝑠𝑠𝑒𝑡𝑠_𝑖𝑡−1+γ₃𝐷𝑒𝑝𝑜𝑠𝑖𝑡_𝑡−1 (2)

+γ₄𝑃𝑟𝑜𝑣𝑖𝑠𝑖𝑜𝑛_𝑖𝑡−1+γ₅𝐵𝑢𝑠𝑖𝑛𝑒𝑠_𝑖𝑡−1+γ₆𝐵𝑒𝑡𝑎_𝑖𝑡−1+γ₇𝑃/𝐸_𝑖𝑡−1+γ₈𝑀/𝐵_𝑖𝑡−1+ 𝜇_𝑖𝑡

The variables, formulas and the data sources are explained in Appendix C. The equation will be used four times, every time the “Ratio” will be changed. In the equation, yit is the

bank’s stock return between the end of the quarter t-1 and the end of quarter t and is calculated by using the natural log, equation 3, of the Datastream return index (RI). The RI is based on the stock return of the period plus the dividend payout, which is completely invested in the same stock. This RI measurement is often used in research and provides a return measurement that is comparable between companies (Demirguc-Kunt et al., 2012).

𝑦_𝑖𝑡 = ln( 𝑅𝐼𝑡

𝑅𝐼𝑡−1) (3)

will react only on information that is available. Because the values are only published once per quarter, it is expected that the effect will be measurable after the announcement until new values are published. The models are estimated using ordinary least squares (OLS) with fixed period and country effects. Robust standard errors are created using White diagonal standard errors. In the result part a more in-depth explanation for choosing those test is provided.

The first sub-question focuses solely on the effects between the four individual capital ratios and stock return. In this stage, it is not yet researched whether higher quality equity, so Tier 1, has a stronger influence on the stock return. Based on the literature it is expected that higher capital ratios have a positive influence on the stock performances. With this first sub-question there is not yet tested for the relation between the ratios and whether some are significantly better. Following this, the same hypothesis will be tested four times, every time using a different ratio. The following hypotheses will be tested:

H1.1: The unweighted Tier 1 capital ratio has a positive influence on the stock performances

H1.2: The unweighted Tier 1 & 2 capital ratio has a positive influence on the stock

performances

H1.3: The weighted Tier 1 capital ratio has a positive influence on the stock performances

H1.4: The weighted Tier 1 & 2 capital ratio has a positive influence on the stock

performances

In order to find out if the ratios based on unweighted assets are a better estimator, one cannot just simply use the individual t-statistics for two variables to test the hypothesis. In order to test the hypotheses a new model is estimated. For this model a new parameter is added,

𝑇𝑂𝑇𝐶𝐴𝑃𝑖𝑡−1 = 𝑤𝑒𝑖𝑔ℎ𝑡𝑒𝑑 𝑟𝑎𝑡𝑖𝑜𝑖𝑡−1+ 𝑢𝑛𝑤𝑒𝑖𝑔ℎ𝑡𝑒𝑑 𝑟𝑎𝑡𝑖𝑜𝑖𝑡−1 (4)

This method is in line with the research of Hogan (2015). Due to the fact that we investigate both the Tier 1 and Tier 1 plus Tier 2 equity in equation (5), the equation will be used twice.

𝑦_𝑖𝑡 = 𝛼 + 𝛽₁𝑅𝑎𝑡𝑖𝑜_𝑖𝑡−1+γ₁𝑇𝑂𝑃𝐶𝐴𝑃_𝑖𝑡−1+γ₂𝐴𝑠𝑠𝑒𝑡𝑠_𝑖𝑡−1+γ₃𝑆𝑖𝑧𝑒_𝑖𝑡−1+γ₄𝐷𝑒𝑝𝑜𝑠𝑖𝑡_𝑡−1 (5)

+γ₅𝑃𝑟𝑜𝑣𝑖𝑠𝑖𝑜𝑛_𝑖𝑡−1+γ₆𝐵𝑢𝑠𝑖𝑛𝑒𝑠_𝑖𝑡−1+γ₇𝐵𝑒𝑡𝑎_𝑖𝑡−1+γ₈𝑃/𝐸_𝑖𝑡−1+γ₉𝑀/𝐵_𝑖𝑡−1+ 𝜇_𝑖𝑡

The p-value of the ratio coefficient tells whether there is a significant difference between the weighted and unweighted ratios. When this is significant, the value of the ratio coefficient will show which of the ratios provides a better measurement. If the coefficient is negative, then the weighted ratio is a better measure of risk than the unweighted. However, when the coefficient is positive, it will support the expectation that the unweighted ratio provides a better measurement (Hogan, 2015). With this equation (5) the following two hypotheses will be tested:

H2.1: Unweighted Tier 1 capital ratios have a higher influence on the stock performances

than weighted Tier 1 capital ratios.

H2.2: Unweighted Tier 1 & 2 capital ratios have a higher influence on the stock

performances than weighted Tier 1 & 2 capital ratios.

density of the period. The low RWA density group has a density smaller than the mean, minus half a standard deviation. The high RWA density group contains all the banks that have a density at least as high as the mean plus half a standard deviation. By including a dummy variable into equation 2 a new equation is formed:

𝑦_𝑖𝑡 = 𝛼 + 𝛽₁𝑅𝑎𝑡𝑖𝑜_𝑖𝑡−1+γ₁𝐿𝑖𝑞_𝑖𝑡−1+γ₂𝐴𝑠𝑠𝑒𝑡𝑠_𝑖𝑡−1+γ₃𝐷𝑒𝑝𝑜𝑠𝑖𝑡_𝑡−1+γ₄𝑃𝑟𝑜𝑣𝑖𝑠𝑖𝑜𝑛_𝑖𝑡−1 (6) +γ₅𝐵𝑢𝑠𝑖𝑛𝑒𝑠𝑖𝑡−1+γ6𝐵𝑒𝑡𝑎𝑖𝑡−1+γ7𝑃/𝐸𝑖𝑡−1+γ8𝑀/𝐵𝑖𝑡−1+ 𝐷ℎ𝑖𝑔ℎ𝑖𝑡−1+ 𝐷𝑙𝑜𝑤𝑖𝑡−1+ 𝜇𝑖𝑡

In order to find out if the dummy influences the stock return, the dummy should be significant. Due to the fact we look at outliers, the high and low group are tested. This implies that when one of those two is significant, the value of the coefficient shows if it has a positive or negative influence. The average group is not included into the equation to prevent perfect multicollinearity. The RWA density is calculated by RWA/total assets per quarter. The two hypotheses that will be tested are:

H3.1: The unweighted Tier 1 capital ratio has an influence on the stock performances and is

influenced by the RWA density.

H3.2: The weighted Tier 1 capital ratio has an influence on the stock performances and is

influenced by the RWA density.

4. Data

For the sake of consistency, the data is only gathered from DataStream and Bankscope. Missing data is not complemented with data from other sources, and due to this, the dataset is unbalanced. In order to check for validity, randomly selected samples are compared with the data from the annual reports and Basel pillar 3 disclosures. All independent variables, except total assets, are ratios, and these ratios are calculated using the currency of the home country of the bank. For the total assets, the currency is standardized to the Euro because most of the banks are based in an Euro country. The exchange rate at the closing date of the statement is used. For all the other independent variables, the ratio is calculated so there is no need for using the exchange rate. The dependent variable RI is extracted from DataStream. This is then converted into a log return between two periods using formula 3.

For the dataset all Banks within the EU are used; those who are public listed during the whole period, and have at least 5 billion of equity in quarter 2, 2015. This cutoff point is used to exclude small banks that work in niche markets. In total, 39 banks match those criteria2_. Table 1 presents the geographical distribution of the banks per country. The dataset is well distributed among countries, Switzerland is missing due to the fact that it does not have implemented CRD IV and CRR.

2 _{The dataset is available upon request.}

Table 1: Geographical distribution of dataset

Country Nr. of Banks Italy 7 Spain 6 Germany 5 France 5 United Kingdom 5 Sweden 4 Poland 2 Denmark 1 Ireland 1 Netherlands 1 Portugal 1 Belgium 1

During the time period, quarter two 2010 until the third quarter of 2015, the stock market has seen its ups and downs. Graph 1 shows the average Ln RI of the dataset per quarter. The second dip in 2011 is clearly visible. The average stock return of the sample was negative over the whole period.

Graph 1: Average Ln RI of European banks from 2010 until 2015. For the calculation the ln(RIit/RIit-1) is

used. In which the RI is the Return Index from DataStream. Based on a total sample of 858 observations. When focusing on the capital ratios a surprising trend is witnessed. Graph 2 shows the four different capital ratios and the averages over time. When analyzing the capital ratios over time, it is remarkable that on average the weighted capital ratios have increased by 50% over a time horizon of 5 years and that the unweighted capital ratios are almost constant over time.

-0,40 -0,30 -0,20 -0,10 0,00 0,10 0,20 0,30 L n r etu

rn Average Ln RI European banks

Average Ln RI 0% 5% 10% 15% 20%

Average capital ratios between 2010 and 2015 of European banks

Tier 1 weighted Tier 1 & 2 weighted Tier 1 unweighted Tier 1 & 2 unweighted

This discrepancy between the weighted and unweighted ratios could underline the hesitations that more and more investors have with the weighted capital ratios. How is it possible that on average the weighted ratios per bank has sharply declined when the denominator does not differ between weighted and unweighted capital ratios? The only explanation is that the assets of the banks have become less risky. This could be explained by two effects; first it could be caused by a shift in the banks’ portfolios, secondly it could be caused by an optimization of the risk weight of their assets. The RWA density supports the expectation that European banks have decreased their RWA compared to their total assets. In graph 3, the average RWA density (μ) is plotted and upper and lower boundaries of one standard deviation (σ) are provided. This provides a clear overview on how the RWA density

differs between the banks and shows that on average there is decline in the density. This decline implies that the risk weight of the assets has declined between 2010 and 2015.

Graph 3: The average (μ), one standard deviation (σ) higher and lower RWA density of European banks per quarter between 2010 and 2015. Based on a total sample of 688 observations.

0% 10% 20% 30% 40% 50% 60% 70% 80%

Average RWA density

The descriptive statistics of the dataset3 are provided in Table 2. Due to the fact that the dataset is unbalanced, the number of observations per variable differ.

Table 2: Descriptive Statistics dataset

Nr. of obs. Mean Std. Dev. Median Maximum Minimum

Ln RI 858 -0,002 0,026 0,020 0,606 -1,212

Tier 1 weighted ratio 670 0,122 0,029 0,117 0,257 0,050 Tier 1 & 2 weighted ratio 666 0,147 0,032 0,144 0,289 0,057 Tier 1 unweighted ratio 691 0,052 0,019 0,049 0,135 0,016 Tier 1 & 2 unweighted

ratio 688 0,063 0,022 0,061 0,137 0,017 TOTCAP1 670 0,174 0,037 0,168 0,330 0,068 TOTCAP12 666 0,209 0,040 0,206 0,340 0,085 Business ratio 767 0,534 0,154 0,571 0,817 0,130 Total assets (ln) 767 19,952 1,141 19,786 21,849 17,487 Provision ratio 762 0,003 0,005 0,002 0,064 0,001 Deposit ratio 767 0,573 0,131 0,571 0,888 0,096 Liquidity ratio 767 0,201 0,118 0,184 0,591 0,018 Market to Book 873 0,996 0,717 0,840 4,820 -2,130 Price to Equity 711 20,629 28,955 13,800 440,700 1,500 BETA 870 1,513 0,573 1,440 4,371 -0,317

Table 2: The descriptive statistics of the whole dataset, containing the data between 2010 and 2015 on a quarterly level. RI(ln) is the dependent variable, for the calculation the ln(RIit/RIit-1) is used. The RI is the

Return Index from DataStream. RI is the return index and measures the stock return plus dividend. The tier 1 and tier 2 ratios use the RWA for the weighted ratio and total assets for the unweighted ones. TOTCAP 1 is the sum of Tier 1 ratio weighted and unweighted. TOTCAP12 is the sum of Tier 1 and 2 ratio weighted and unweighted. Based on the total sample of 39 banks and quarterly observations.

The average of the Ln RI shows that there is a negative overall return in the period between 2010 and 2015. Furthermore, it shows that the Tier 2 capital is only a small part of the equity provision of banks. Besides that, there is a large difference between the bank’s Tier 1 weighted capital ratio. One bank has a maximum of almost 26% but there is also a bank that has a Tier 1 weighted ratio of 5%, which is extremely low and even below the Basel III minimums. The same holds for the unweighted capital ratios. The business ratio and deposit

Table 3: Correlation between independent variables Provi si o n r at io B ET A D epos it r at io L iqui di ty r at io Ln A ss et s TO TC A P12 TO TC A P1 We ig ht ed Ti er 1 rat io U nw ei ght ed T ier 1 r at io We ig ht ed Ti er 1 & 2 r at io U nw ei ght ed T ier 1 & 2 ra ti o PE Busi ne ss r at io M TB V Provision ratio _1,000 BETA _{0,130 1,000} Deposit ratio _{0,263 0,233 1,000} Liquidity ratio _{-0,397 -0,208 -0,430 1,000} ln Assets -0,261 -0,195 -0,543 0,606 1,000 TOTCAP12 -0,119 0,036 0,207 -0,031 -0,167 1,000 TOTCAP1 -0,164 0,124 0,179 -0,033 -0,179 0,839 1,000

Weighted Tier 1 ratio _{-0,305 -0,031 -0,130 0,248 0,146 0,718 0,877 1,000}

Unweighted Tier 1 ratio _{0,189 0,311 0,594 -0,498 -0,624 0,494 0,553 0,083 1,000}

Weighted Tier 1 & 2 ratio _{-0,280 -0,107 -0,111 0,272 0,183 0,856 0,760 0,898 0,019 1,000}

Unweighted Tier 1 & 2 ratio _{0,223 0,243 0,578 -0,499 -0,618 0,546 0,391 -0,066 0,923 0,034 1,000}

PE -0,024 0,010 0,008 0,007 0,031 -0,030 0,008 0,026 -0,030 -0,001 -0,055 1,000

Business ratio 0,311 0,259 0,415 -0,778 -0,763 0,077 0,067 -0,230 0,536 -0,249 0,552 -0,054 1,000 MTBV _{-0,012 -0,019 0,027 -0,093 0,012 0,096 0,038 0,014 0,054 0,060 0,089 0,023 0,016 1,000}

5. Results

This section reports the outcome of the statistical analyses in order to answer the research question and its sub-questions. For this research the statistical program EViews 8.0 is used. As mentioned before, the dataset is based on DataStream and Bankscope. For the analysis ordinary least squares (OLS) method is used. This statistical method focuses on minimizing the squared residual of the prediction. The easiest way of testing this panel data set is using a simple pooled OLS. However, this would assume that there is no heterogeneity, which means that the same relationship holds for all the data. Since the panel contains banks in multiple countries over a larger time period, it is expected that there is at least some heterogeneity.

First, all the tests will be performed using OLS with random effects. The problem with random effects is that, in order to be reliable, the error term should be uncorrelated with the explanatory variables. In order to find whether using random effects is allowed, the Hausman test is performed. The null hypothesis of this test is that the error term is uncorrelated with the independent variables. For all the estimations the Hausman test is significant, and therefore random effects are not allowed. In order to perform the regressions, fixed effect are preferred. Even in the European Union there are differences between countries that may influence the dependent variable. Based on this cross-sectional fixed effects are added.

the standard errors is not constant, White diagonal standard errors are used. The first part of the research focuses on the influence of the multiple capital ratios. Table 4 contains the outcomes of those regressions, between those four regressions only the ratio changed.

Table IV: Results on hypotheses focused on the influence of weighted and unweighted capital ratios. For the analyses the RI is used as dependent variable. The RI is calculated using Ln(RIit/RIit-1) is used. The model is

estimated four times, once for every hypothesis. The ratios are calculated by dividing Tier 1 by RWA for the weighted ratio and total assets for the unweighted ratio. The same is done for Tier 1 & 2 but then the sum of Tier 1 & 2 is used. In order to create robust standard errors White diagonal standard errors are used.

Table 4: Results on influence of four different capital ratios on stock performances

H1.1 H1.2 H1.3 H1.4

Tier 1 weighted ratio 0,815** (0,400)

Tier 1 & 2 weighted ratio 0,248 (0,269)

Tier 1 unweighted ratio 2,742***

(1,075)

Tier 1 & 2 unweighted ratio 0,975

(0,741) Deposits ratio 0,086 (0,112) 0,083 (0,111) 0,059 (0,109) 0,073 (0,108) Provision ratio -14,529*** (3.379) -15,082*** (3,512) -14,352*** (3,223) -14,878*** (3,523) Business ratio 0,337** (0,155) 0,308** (0,151) 0,221* (0,191) 0,233* (0,137) Liquidity ratio 0,166 (0,156) 0,162 (0,158) 0,076 (0,142) 0,091 (0,146) Assets 0,018 (0,017) 0,018 (0,017) 0.031* (0,019) 0,020 (0,018) Price-earnings ratio -0,001 (0,001) -0,001 (0,001) -0,001 (0,001) -0,001 (0,001) M to B 0,003 (0,009) 0,001 (0,009) -0,001 (0,008) -0,001 (0,008) Beta -0,003 (0,012) -0,003 (0,020) -0,015 (0,013) -0,007 (0,013) Constant -0,674 (0,4132) -0,566 (0,422) -0.846* (0.467) -0,579 (0,443)

Country fixed effects Yes Yes Yes Yes

Period fixed effects Yes Yes Yes Yes

Adj. R-square 0.402 0,396 0.410 0.399

Nr. Observations 523 521 542 541

From the results of the analyses, multiple conclusions are drawn. Tier 1 weighted capital ratios have a significant influence on the stock performances of banks (p<0,05). This effect does not hold when Tier 2 capital is included. This shows that investors value higher equity. For the unweighted capital ratios, the same relation is found. Tier 1 unweighted capital ratios is significant on a 1% level. However, when Tier 2 is added there is no significant relation between unweighted capital ratio and the RI of banks. Those results imply that after the financial crisis, equity with a higher quality is valued by investors and that it has a positive influence on the RI. For all the four tests the adjusted R-square is around 0,4 and all regressions have at least 520 observations.

Table 5: Results on unweighted versus weighted capital ratios

H2.1 H2.2

Tier 1 unweighted ratio 1.526 (1.147)

TOTCAP1 0.577*

(0,336)

Tier 1 & 2 unweighted ratio 0,347

(0,834) TOTCAP12 0.184 (0,255) Deposits 0,067 (0,114) 0,077 (0,122) Provision -14,163*** (3,208) -15,035*** (3,499) Business ratio 0,272** (0,138) 0,278 (0,142) Liquidity 0,114 (0,151) 0,134 (0,156) Assets 0,029 (0,019) 0,021 (0,019) Price-earnings ratio -0,001 (0,001) -0,001 (0,001) M to B 0,002 (0,009) 0,001 (0,009) Beta -0,012 (0,014) -0,006 (0,011) Constant -0,903* (0,497) -0,628 (0,448)

Country fixed effects Yes Yes

Period fixed effects Yes Yes

Adj. R-square 0.408 0,395

Nr. Observations 523 521

*** 0.01 significant **0.05 significant *0.10 significant

The last step is to test whether the RWA density is influencing the relation between the capital ratios and stock performances. In order to measure this effect, three groups are created. The banks are grouped every quarter based on the mean RWA density of the period and the standard deviation. The first group, low RWA dense, contains all the banks that have a RWA density lower than the mean minus half a standard deviation. The second group, average RWA dense, is for the banks that have a RWA density between the upper boundary of the mean plus half a standard deviation and the lower boundary of the mean minus half a standard deviation. The last group, so the high dense banks have a density above the mean plus half a standard

deviation. For the regression, the low dense and high dense group are added into the equation. Only the capital ratios based on Tier 1 have a significant influence in the first test, thus

Table 6: Results on influence of RWA density

H3.1 H3.2

Tier 1 weighted ratio 0,821*

(0,446)

Tier 1 unweighted ratio 3,712***

(1,201)

Low RWA density 0,0014

(0,020)

0,054** (0,023)

High RWA density -0,007

(0.024) -0,027 (0,020) Deposits 0,087 (0,112) 0,048 (0,108) Provision -14,523*** (3,377) -13,859*** (3,185) Business ratio 0,333** (0,152) 0,296** (0,143) Liquidity 0,155 (0,160) 0,146 (0,149) Assets 0,020 (0,019) 0,028 (0,020) Price-earnings ratio -0,001* (0,000) -0,001 (0,001) M to B 0,003 (0,010) -0,002 (0,009) Beta -0,003 (0,013) -0,012** (0,012) Constant -0,700 (0,473) -0,885* (0,489)

Country fixed effects Yes Yes

Period fixed effects Yes Yes

Adj. R-square 0,400 0,416

Nr. Observations 473 485

*** 0.01 significant **0.05 significant *0.10 significant

Table 6: RWA density and its influence on the stock performances of banks. The group of low RWA density has a RWA density below the mean minus half a standard deviation. The group of high RWA density has a RWA density above the mean plus half a standard deviation. The group with an average RWA density is left out. The groups are formed every quarter. White diagonal standard errors are used. For the analyses the RI is used as dependent variable. The RI is calculated using ln(RIit/RIit-1) and is collected using

6. Conclusion

Since the financial crisis in 2008, the focus of investors and regulations on the capital position of banks has increased. Using data on publicly listed European banks with more than 5 billion in equity and internationally active, the effect of capital ratios on stock performances between 2010 and 2015 is researched. The goal is to find to which extent the stock performances of banks are impacted by their capital ratios. The first analyses find that capital ratios based on Tier 1 capital have a significant influence and that capital ratios based on Tier 1 and Tier 2 do not have a significant influence. It is notable that for both the unweighted and weighted capital ratio the significant influence is only present for the Tier 1 capital. Based on the findings of this research, evidence is provided that investors do value higher capital ratios and that there is a higher demand for banks that have high ratios of Tier 1 capital, both weighted and unweighted. This means that the sole influence of Tier 1 capital is significant but when the Tier 1 and 2 of a bank is combined the relation is not significant anymore. The findings on the influence of Tier 1 capital are in line with the literature, however the not significant influence of Tier 1 and 2 combined differs. The outcomes of this part of the research can help banks by deciding how to shape future equity emissions. For banks it could be more valuable to issue products that count under the Tier 1 requirements and by doing so, increasing the potential of higher stock returns.

suggested that the trust in RWA is declining. The results from this research show that investors still trust the RWA of European banks. This contradiction could be explained by the time period of this research. For this research the time period was 5 years in order to collect enough data for a valid analysis. It could be possible that the distrust has grown in the last few years and that this effect is not measurable yet. As the data show, the weighted capital ratios are steeply rising but the unweighted capital ratios stay constant during the entire period. This trend cannot continue for another 5 years.

With the RWA density a new parameter to measure the differentiation between a bank's RWA and total assets is presented. Before this research, the effect of the density on the stock performances was never been researched. The interesting thing about RWA density is that it provides a ratio that can be used to measure the riskiness of the assets of a bank. When a bank has a high RWA density, it means that the risk weights of their total assets are high. In order to see whether investors take into account a bank’s ratio between RWA and Total assets, three

7. Limitations

As the conclusion shows, there is a positive causality between the capital ratios of banks and their stock performances. However, there are some limitations in this research. First of all, the focus of this research was on banks that are listed and had an equity above 5 billion Euro. It could be argued that, due to those cut-off points, characteristics important for investors were left out. In order to check for the robustness a lower cut-off point could be used. By adding such banks, it could happen that the dataset will become too diverse. Secondly, due to the fact that the focus is on stock returns, all banks that are not listed were excluded. As this research shows, higher capital is valued by investors and has a significant influence on the stock performances of European banks. The underlying motivation to only focus on European banks, all exposed to the same regulations CRD IV and CRR, offers the opportunity to research other countries. However, the challenge with this is that most of the countries adopt their own legal framework.

8. Literature list

Acharya, V., Schnabl, P., Suarez, G., 2013. Securitization without risk transfer. Journal of Financial Economics 117, 515-536.

Avdjiev, S., Kartasheva, A., and Bogdanova, B., 2013, CoCos: a primer. Bank of International Settlement Quarterly Review 3, 43-56.

Bank of England, 2012, Financial stability report.

Bank of International Settlement, 2014, Basel III monitoring report.

Basel Committee on Banking Supervision,1999, Principles for the management of credit risk.

Basel Committee on Banking Supervision, 2001a,Consultative Document: The internal Ratings-Based Approach.

Basel Committee on Banking Supervision, 2001b, Consultative Document: Pillar 2.

Basel Committee on Banking Supervision, 2010, Basel III: A global regulatory framework for more resilient banks and banking systems.

Beltratti, A., Stulz, R., 2012, The Credit Crisis around the Globe: Why Did Some Banks Perform Better? Journal Of Financial Economics 105, 1-17.

Chan-Lau, J., Liu, E.., Schmittmann, J., 2015, Equity returns in the banking sector in the wake of the great recession and the European sovereign debt crisis. Journal Of Financial Stability 16, 164-172.

Crotty, J., 2009, Structural causes of the global financial crisis: a critical assessment of the 'new financial architecture'. Cambridge Journal Of Economics 33, 563-580.

DNB, 2009, DNB Timeline ING.

Das, S., Sy, A., 2012, How Risky Are Banks’ Risk Weighted Assets? Evidence from the Financial Crisis. working paper, IMF.

Demirguc-Kunt, A., Detragiache, E., Merrouche, O., 2012, Bank Capital: Lessons from the Financial Crisis. Journal of Money, Credit and Banking 45, 1147-63.

Diamond, D., Rajan, R., 2000, A theory of Bank Capital. Journal of Finance 55, 2431-65.

Dijsselbloem, J., 2013, Onteigeningsbesluit SNS REAAL en SNS Bank.

Fama, E., French. K., 1992, The Cross-section of expected stock returns. The Journal of Finance 47, 427-465.

Gow, I., Ormazabal, G., Taylor, G., 2009, Correcting for Cross-Sectional and Time-Series Dependence in Accounting Research. The Accounting Review 85, 483-512.

Hogan, T., 2015, Capital and risk in commercial banking: A comparison of capital and risk-based capital ratios. The quarterly review of economics and Finance 57, 32-45.

Le Leslé, V., Avramova, S., 2012, ‘Revisiting Risk-Weighted Assets. Working paper, IMF.

Mariathasan, M., Merrouche, O., 2014, The manipulation of Basel risk-weights. Journal of Financial Intermediation 23, 300-321.

Miles, D., Yang, J., Marcheggiano, G., 2013, Optimal bank capital. The Economic Journal 123, 1-37.

Shrieves, R., Dahl D., 1992, The relationship between risk and capital in commercial banks. Journal of Banking and Finance 16, 439-457.

Tabachnick, B, Fidel, L, 2012, Using Multivariate statistics. Pearson Education; Boston

Yang, J., Tsatsaronis, K., 2012, Bank stock return, leverage and the business cycle. Bank of International Settlement Quarterly Review 1, 45-59.

Appendix A: Basel III outline

The Basel Committee on Banking supervision was originated in 1974, following the problematic breakdown of the way exchange rates were managed. After this event many banks incurred large foreign currency losses. This triggered the collapse of the Bankhaus Herstatts in 1974 which eventually spread the turmoil to other international banks. The Basel Committee was designed to improve the cooperation between its member countries. Its aim is to increase financial stability by improving the knowledge and quality of the banking supervision. Currently, the BIS consist of sixty central banks and monetary authorities The Latin American debt crisis of the eighties caused the capital ratios of the main international banks to decline and at the same time the risks grew. This problem was partly caused by the differences in national capital requirements. Within the committee there was a strong need for a multinational accord that strengthen the stability of the international banking industry. In 1987 the first Basel capital accord was approved by the members. In this accord the first capital ratio was included. At the end of 1992 banks had to hold a minimum of 8% of capital to RWA. Until the publication of Basel II, it was intended that the accord would evolve over time.

Over time the dissatisfaction on the effectiveness of the regulation grew. There were two main problems with the accord, first the insufficient risk sensitivity and secondly the concern of regulatory capital arbitrage. Due to the simplified risk weighting structure, banks were able to sell or securitize assets for those which are believed to have a higher capital charge than the one the market would impose and hold the assets with a lower quality for which the capital charge was relatively low (Zicchino, 2006). Besides this, it encouraged banks to hold the assets that had the highest risk in a certain risk weight bucket (Acharya, Schnabl & Suarez 2013).

pillar focus on the minimum capital requirements and, so on credit risk, market risk and operational risk. The second pillar focus on the supervisory review process and contains a framework for banks which includes capital allocation and risk management. Furthermore, it provides a supervisory framework which is designed for evaluating the internal system, assessing the risk profiles and reviewing the compliance with all regulation measures. The third pillar tries to increase the transparency and is focused on disclosure requirements. Figure 2 provides an overview of the three pillar structure:

Figure 2: Basel II Revised Capital framework three pillar structure

The financial crisis of 2008 uncovered the shortcomings of the Basel II agreements. Banks relied too much on the external credit rating agencies, the lack of focus on leverage, the intertwine of the global banking sector, and the inability of the banks to absorb shocks. In order to improve the regulations, the Basel committee published Basel III; A global regulatory framework for more resilient banks and banking systems in 2010 (BIS, 2011). This formed the bases for Basel III. The aim of the members is to phase-in the arrangements, stepwise before 1 January 2019. The Basel III framework for more resilient banks and banking systems was published in 2010 and an improved version in 2011. The goal is to increase the capital buffer of the banks and reduce the risk exposure. Furthermore, it tries to decrease the possibility of regulatory arbitrage. New in the accords is the leverage ratio, which could be seen as a backstop to prevent dubious practices, for example underrate the loan to reduce capital requirements, but also more legitimate forms of regulatory capital arbitrage. For banks that use the Standardized approach for their RWA, Basel III improves the risk exposure to market risk but does not change the standardized approach on credit risk (Das & Sy, 2012). This market risk is covered in the first pillar.

Pillar 1

Table 7: Types of Tier 1 and 2 equity

Common Equity Tier 1 Additional Tier 1 Tier 2

Common shares Instruments issued by the bank that meet criteria

Instruments issued by the bank that meet criteria for Tier 2 Share premium Share premium from additional

Tier 1 capital

Share premium from Tier 2 capital

Retained earnings Preferred shares

Table 7 VII: frequent types of Tier 1 and Tier 2 capital (BIS, 2011)

As table 7 shows, the requirements for common equity Tier 1 are clearest. In order to meet the requirements of Basel, many banks created new products that meet the requirements for additional Tier 1 and Tier 2. An example of such a product is the contingent convertible, commonly known as CoCo’s. Those CoCo’s are hybrid capital securities that absorb losses

Table 8: Timeline of Basel III implementation

Phases 2013 2014 2015 2016 2017 2018 2019

Capital Leverage ratio parallel run 1 Jan 2013 – 1 Jan 2017. Disclosure starts 1 Jan 2015 Migration to Pillar 1 Minimum Common Equity

Capital Ratio

3,5% 4,0% 4,5% 4,5%

Capital Conservation Buffer 0,625% 1,25% 1,875% 2,5%

Minimum common equity plus capital conservation buffer

3,5% 4,0% 4,5% 5,125% 5,75% 6,375% 7,0%

Phase-in of deductions from CET1*

20% 40% 60% 80% 100% 100%

Minimum Tier 1 Capital 4,5% 5,5% 6,0% 6,0%

Minimum Total Capital 8,0% 8,0%

Minimum Total Capital plus conservation buffer

8,0% 8,625% 9,25% 9,875% 10,5%

Capital instruments that no longer qualify as non-core Tier 1 capital or Tier 2 capital

Phased out over 10 years horizon beginning in 2013

Liquidity Liquidity coverage ratio – minimum requirement

60% 70% 80% 90% 100%

Net stable funding ratio

To prevent liquidity shortages that occurred during the financial crisis, Basel III introduced the liquidity measures. Two ratios are introduced; Liquidity Coverage Ratio (LCR) and the Net Stable Funding Ratio (NSFR). The LCR focuses on the assets that are easily and without much value loss converted into cash. The ratio is calculated by dividing the high quality liquid assets by the total net cash outflows over the next 30 calendar days. The NSFR focuses on the funding of long term assets. This ratio is calculated by dividing the available amount of stable funding by the required amount of stable funding. For the capital requirements, three risk are taken into account. The first one, credit risk, is the most important one in the Basel agreements and can be defined as

“the potential that a bank borrower or counterparty will fail to meet its obligations in accordance with agreed terms” (BIS, 1999).

Besides the aforementioned credit risk, there are two more types of risk that are tried to counter with the Basel framework. Following EBA (2015) market and operational risk are defined as:

Market risk: “The risk of losses in on and off-balance sheet positions arising from adverse

movements in market prices.”

Operational risk: “the risk of losses stemming from inadequate or failed internal processes,

people and systems or from external events. Operational risk includes legal risks but excludes

reputational risk and is embedded in all banking products and activities.”

Pillar 2

With the introduction of Basel II, the committee tried to cover the gaps that were identified in Basel I. The aim of processes is to improve the connection between the bank’s risk profile, the management of this risk, capital planning and the mitigation of the risks. Furthermore, it provides a framework to cope with the other risk. In the first pillar, market, credit and operational risk are handled. However, liquidity, reputational, systemic, strategic and legal risks are not incorporated. This Pillar has two major goals. The first one is that institutions should establish strong, effective and complete strategies, and processes. This has to be done in order to maintain and asses the amounts, types, and distribution of capital inside the organization so that it meets the corresponding risk profile. The second goal focuses on the review and evaluation of the process by the supervisory body. It is important that banks have adequate arrangements, processes, mechanisms and strategies in place to ensure a coverage of the risk to which they are exposed (BIS, 2001b).

Pillar 3

Appendix B: Key concerns of multiple stakeholders on RWA. Cited

from Le Le Leslé and Avramova (2012, p. 7-8)

Key concern Possible impact

Regulators’ main concerns Reliability and Accuracy of capital ratios

 Inaccurate measurement of risk, both on and off-balance sheet

 Understatement of risk

 Tail risk not captured properly, thus low probability/high impact events mispriced

 Reported capital adequacy ratios can substantially overstate banks’ capital adequacy and send wrong signals about the true solvency and resilience of banks

 Capital ratio is not a reliable indicator of stress, possibly delaying necessary restructuring/ recovery/resolution

Adequacy of capital

 RWAs as a percentage of total assets have trended down in recent reporting despite a heightened risk environment, leading to concerns that low RWA calculations reflect insufficient capital

 RWAs decrease due to “optimization”, “model changes”, “data cleaning”, “parameter updates”, etc…

 Banks with similar business models may have very different capital levels

 Imperfect match between risk and capital

 Some banks are under-estimating risk

 The lower the RWA density, the higher the scope for error in the calculation of capital requirements

Pro-cyclicality

 RWAs, which rely on ratings and are mostly based on historical parameters, may be too low in good times & rise too late in bad times

 Probability of default: “point in time” versus “through the cycle” affects cyclicality of capital

 Calculation of RWA may amplify pro-cyclicality of capital requirements, as banks de-leverage in a downturn to reduce their RWAs, or increase them in good times, thereby amplifying the crisis or building up an asset bubble

Risk-taking incentives and risk management

 Banks may “game” the system by underestimating risks to optimize their capital beyond what prudence requires

 RoE type targets may incentivize banks to aggressively manage their RWAs down

 Lack of prudence and excessive management discretion in pushing capital down may result in aggressive risk-taking and could potentially lead to bank failure, with significant related social and economic costs

Banks’ main concerns Competitive advantage

 Some banks worry that the banks with the lowest RWAs could benefit from an undue competitive advantage (due to lower capital requirements), and capture market shares thanks to more aggressive pricing power

 G-SIFI capital surcharge will be calculated as a percentage of risk-weighted assets, not

 Least conservative banks could gain ground, which could threaten global financial stability

 Based on RWAs, the G-SIFI surcharge does not necessarily penalize the largest banks (in terms of total assets)

of total assets, which could favor some banks over others in terms of additional capital

 Uneven regulations and supervision of banks’ RWA practices across jurisdictions

 Model approvals are neither uniformly robust nor uniformly reviewed

buffers is variable

 Supervisory practices vary excessively, and some banks get a more lenient treatment

Investors’ & Markets’ main concerns

Comparability of capital ratios

 RWAs are subjective and vary from one bank to the next, and it is challenging to compare capital ratios across banks, both cross border & within countries

 Markets may prefer a simpler, more objective and easier to compare measure such as the leverage ratio

Credibility of capital ratios

 Different methodologies may lead banks, regulators, and markets to distrust each other’s on reported RWAs

 Could lead to a confidence crisis, where markets become reluctant to lend to banks, ultimately resulting in a liquidity crisis

Opacity and complexity of internal models

 The formula for calculating RWAs is very complex in itself and leaves large potential for different interpretations

 Difficult for markets to gauge the quality of internal models and the robustness of methodologies used by IRB banks (a difficulty also faced to a certain extent by supervisors)

 Large cross-border banks often rely on a myriad of models, each measuring a small portion of the assets under specific rules of various jurisdictions, and it is not unusual for G-SIFIs to employ several dozens of models simultaneously

 Markets may doubt RW based capital measure & adopt leverage ratio instead

 Regulators may be tempted to over-ride internal models and impose minimum risk weights floors

Appendix C: Variables and data sources

Dependent variable Formula Source

Return index (RI)

𝑅𝐼_𝑡 = 𝑅𝐼_𝑡−1∗ 𝑃𝐼𝑡 𝑃𝐼𝑡−1(1 + 𝐷𝑌𝑡 100∗ 1 𝑁) DataStream

Independent variables Formula Source

Tier 1 weighted ratio 𝑇𝑖𝑒𝑟 1

𝑅𝑊𝐴 Bankscope

Tier 1 & 2 weighted ratio 𝑇𝑖𝑒𝑟 1 + 𝑇𝑖𝑒𝑟 2

𝑅𝑊𝐴 Bankscope

Tier 1 unweighted ratio 𝑇𝑖𝑒𝑟 1

𝑇𝑜𝑡𝑎𝑙 𝐴𝑠𝑠𝑒𝑡𝑠 Bankscope

Tier 1 & 2 unweighted ratio 𝑇𝑖𝑒𝑟 1 + 𝑇𝑖𝑒𝑟 2

𝑇𝑜𝑡𝑎𝑙 𝐴𝑠𝑠𝑒𝑡𝑠 Bankscope Liquidity ratio 𝐿𝑖𝑞𝑢𝑖𝑑 𝑎𝑠𝑠𝑒𝑡𝑠 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠 Bankscope Assets _{𝐿𝑛(𝑡𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠)} Bankscope Deposits ratio 𝐷𝑒𝑝𝑜𝑠𝑖𝑡𝑠 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠 Bankscope Quality ratio 𝐿𝑜𝑎𝑛𝑠 𝑙𝑜𝑠𝑠 𝑝𝑟𝑜𝑣𝑖𝑠𝑖𝑜𝑛𝑠 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠 Bankscope Business ratio 𝑁𝑒𝑡 𝑙𝑜𝑎𝑛𝑠 𝑇𝑜𝑡𝑎𝑙 𝑎𝑠𝑠𝑒𝑡𝑠 Bankscope Stocks Beta 5 years covariance between the bank’s quarterly

stock return and the country stock market return

DataStream

Market to book value of equity 𝑆𝑡𝑜𝑐𝑘 𝑃𝑟𝑖𝑐𝑒 𝑝𝑒𝑟 𝑆ℎ𝑎𝑟𝑒 𝑆ℎ𝑎𝑟𝑒ℎ𝑜𝑙𝑑𝑒𝑟𝑠′_{𝐸𝑞𝑢𝑖𝑡𝑦 𝑝𝑒𝑟 𝑆ℎ𝑎𝑟𝑒}

DataStream

Price-earnings ratio 𝑀𝑎𝑟𝑘𝑒𝑡 𝑣𝑎𝑙𝑢𝑒 𝑝𝑒𝑟 𝑠ℎ𝑎𝑟𝑒 𝐸𝑎𝑟𝑛𝑖𝑛𝑔𝑠 𝑝𝑒𝑟 𝑠ℎ𝑎𝑟𝑒