1
Bank lending and country risk: an empirical analysis of
European banks.
Master Business Economics, Finance track Master Thesis
Student: Shannon Sewnandan Number: 6148379
Field: Finance
Supervisor: Mark Dijkstra Completion: July 2016
Abstract
This paper studies the impact of country risk on the amount of bank loan supply, when banks face an increase of the Tier 1 ratio set under Basel III. Using a panel data regression with a sample consisting of 19 European countries and 63 European banks, estimates are done over the period 2001-2014. Since the interest on governments bonds increased when governments default in repaying debt, this variable is used to measure country risk. The higher country risk the higher the 10-year interest rate. The first regression estimates the effect of the Tier 1 capital ratio on the amount of bank loan supply. The results are significant and show a reduction of 3.97% in loan supply when the Tier 1 ratio increases with 1 unit. The main finding of this paper shows that banks reduce the amount of loan supply with 5.46% when they face country risk and the Tier 1 capital ratio increases.
Key words
2 Statement of Originality
This document is written by Shannon Sewnandan who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3
TABLE OF CONTENTS
1. Introduction……….3
2. Theoretical background………..5
2.1 Theory………5
2.2 Empirics………...9
3. Methodology and Data………..10
3.1 Methodology………..10
3.2 Data……….13
4. Results………14
4.1 Robustness tests………19
4
Introduction
Banks provide liquidity, monitor services and produce information. In order to strengthen the stability of the international banking system and to set up a fair and consistent international banking system, the Basel Committee in Banking Supervision introduced a capital framework in 1988 which is known as the Basel I accord (Basel Committee, 1988). Over the years, this accord has been adapted because regulators have to adjust capital adequacy standards in the short term in order to achieve an optimal regulation system for banks in the medium to long run. However, when in 2007 several important banks e.g. Lehman brothers, were going bankrupt it was obvious that the capital requirements were still too low (Rossignolo, Fethi & Shaban, 2012). Capital requirements are requirements for banks and other depository
institutions which determines how much liquidity should be held for certain levels of assets. One of the capital requirements is the Tier 1 capital ratio which increases from 4% in Basel II to 6% under Basel III.
The instability of banks has stoked a debate about optimal capital requirements. It is argued that capital requirements are too high for banks, which negatively affects bank loan supply. Overall conclusion from the literature (Aiyar, Calomiris and Wieladek, 2014a, 2014b; Francis and Osborne, 2012; Bridges, Gregory, Nielsen, Pezzini, Radia, and Spaltro, 2014; Noss and Toffano, 2014) is that stricter capital requirement reduces the amount of loan supply by banks, since raising equity is more costly. Besides stricter capital requirements, banks also faced sovereign risk problems, where governments default in repaying their debt. However how country risk (sovereign risk) relates to bank loan supply has not been studied yet. Therefore his study tries to estimate the effect of country risk on bank loan supply when the Tier 1 capital ratio increases. This result in the following research question:
What is the effect of country risk on the amount of bank loan supply, when the Tier 1 capital ratio increases?
Studies of Myers (1977), Amati (2011) and Boot (2013) argue that the cost of capital is high due to e.g. tax benefits, debt overhang problems and implicit government subsidies. It is expected that an increase in bank capital requirements will have a negative impact on the amount of bank loan supply because raising equity is more expensive due to the higher cost of capital. Since bank lending to consumers and corporations is important for economic growth, higher cost of capital could worsen this growth (Ivanshina & Scharfsein, 2009). This is
5 because corporations and consumers are being restricted in doing their investments and
expenditures which partly are being financed with bank loans.
Another consequence of the bankruptcy problems that banks faced was that
governments had to bail out banks that were too important to fail. Governments issued bonds to finance these bail outs in order to prevent recessions. National debt levels rose to high levels and simultaneously governments in Portugal, Italy, Ireland, Greece and Spain (PIIGS) were not able to meet sovereign debt payments (sovereign risk), resulting in the sovereign debt crisis (Barth, Prabha, & Yun, 2012). The compensation banks required for the fact that they had to face sovereign risk is characterized by rising interest rates on governments bonds. Therefore banks can make higher profits by providing more government bonds and at the same time reduce loan supply to corporations and consumers in order to meet the higher Tier 1 ratio of 6% (De Bruyckere et al., 2013). For banks located in one of the PIIGS countries this effect should be stronger because these countries faced lower levels of loan demand since the average GDP growth rates declined between 2008 and 2013 (-23.8% in Greece, -11.2% in Ireland, -10.1% in Spain, -9.2% in Italy and -7.0% in Portugal). When GDP growth declines it means that total demand for final goods and services in an economy declines, including loan demand (Martin & Waller, 2012). This is because consumers and corporations have low levels of expenditures which are partly being financed with bank loans. Since a bank in for example Italy already face lower loan supply (due to lower demand levels) an increase in the Tier 1 ratio should reduce the amount of loan supply by a greater amount than for a bank in e.g. Germany.
In order to test the effect of country risk on the amount of bank loan supply when the Tier 1 ratio increases, a panel data regression is used which contains annual data and controls for time fixed effects. First, the overall effect of an increase in the Tier 1 capital ratio on the amount of loan supply by banks in the European countries is tested. Secondly the effect of country risk is tested. In order to measure country risk, the 10-year interest rate on
government bonds is used. The higher the country risk, the higher is the compensation banks require, resulting in a higher interest rate on government bonds. Results show that the effect of an increase in the Tier 1 ratio reduces the amount of loan supply by 3.97% when a bank faces no country risk. The reduction in loan supply is 5.46% when a bank has to deal with country risk when the Tier 1 ratio increases with 1 unit. Both results are significant. This paper is organized as follows. Chapter 2 contains the relevant literature which will outline this research. Chapter 3 explains methodology and data. Chapter 4 will describe the results of the empirical analysis and the hypotheses are tested in this chapter, while chapter 5 contains the conclusion.
6
2. Literature
2.1 Theory
In response to the recent financial crisis in 2007, the Basel Committee in Banking Supervision (BCBS) introduced the Basel III Accord in 2010. This regulatory framework is an update of Basel II regarding capital, liquidity and leverage requirements (BCBS, 2010a). According to Rossignolo, Fethi &Shaban (2013) the main purpose of Basel III is to strengthen capital and liquidity regulations to make banks more resilient in order to withstand negative shocks such as impact of crisis.
Under the Basel Accord, a bank’s capital consists of Tier 1 and Tier 2 capital. A bank’s Tier 1 capital ratio is calculated by dividing its Tier 1 capital, which is a bank’s
shareholders equity and retained earnings by its total risk weighted assets. Total risk weighted assets are bank’s assets weighted according to risk. Riskier assets such as unsecured loans are given a higher weight than less risky collateralized loans such as mortgages. The Tier 1 capital ratio is used to rate a bank’s capital adequacy, which measures a bank’s risk of insolvency from excessive losses and thus require banks to retain a certain level of capital as protection against unexpected losses. Tier 2 capital, or supplementary capital, consist of revaluation reserves, hybrid capital instruments, subordinated term debt, general loan-loss reserves and undisclosed reserves. Revaluation reserves are reserves that arise when a value of assets becomes greater than the value at which it was previously carried on the balance sheet. Hybrids are instruments that have characteristics of both debt and equity and subordinated debt is any outstanding loan that will be repaid only after all other debt and loans have been settled. The amount of cash and cash equivalents a bank holds in order to cover losses in its loan portfolio are called general loan-loss reserves. The last component of the Tier 2 capital ratio is undisclosed reserves, which are hidden reserves that do not appear on a banks’ balance sheet (BCBS, 2010b). The minimum Tier 1 ratio increases from 4% in Basel II to 6% under Basel III, whereas the minimum total capital ratio (which is Tier 1 plus Tier 2) of 8% increased to 10.5% (BCBS, 2010a). If a bank has a Tier 1 capital ratio lower than 6%, the bank has several ways to increase their Tier 1 ratio. Either by raising new capital or shift its portfolio towards lower risk weighted assets.
According to Jackson et al. (1999), banks are likely to choose the most cost-effective way of fulfilling these capital requirements. As such if the cost of raising new capital is relatively high, banks will reduce their amount of loan supply in order to meet the regulatory requirement. Modigliani & Miller (1958) described a fundamental theory regarding to a firm’s
7 capital structure and the weighted average cost of capital (WACC). They state that the value of a firm, in a perfect capital market (where there are no taxes, no transaction costs, no bankruptcy costs, no asymmetric information between the company and investors, and no effect of debt on a firm’s earnings before interest and taxes), is independent from its capital structure. The value of a firm is determined by its earnings and by the risk captured in its assets. The WACC is the simple weighted average of the cost of equity and the cost of debt. Whether a firm finances its operations with debt or with equity, the weighted average cost of capital should remain constant. For example, no matter which funding the firm uses in order to borrow, there will be no tax benefit from interest payments and thus no changes or benefits to the WACC. Even if the cost of debt is lower than the cost of equity, the weighted average cost of capital does not change with the capital structure. This happens because when a firm increases debt financing which is its cheaper source of financing it also increases the risk for equity-holders, and hence the cost of equity goes up.
However in the real world this theory does not hold, because there are market
imperfections such as taxes. In 1963, Miller and Modigliani proposed a theory to incorporate the tax benefit of debt. This theory states that interest payments are tax deductible and as such create a benefit for the company. For banks this means that it is more attractive to use debt e.g. lend from other banks or the central bank when financing new loans, which makes it more expensive to use equity. This increase in leverage can lead to a so called debt overhang
problem described by Myers (1977), where a bank is in a situation where it has existing debt so great that it cannot easily borrow more money because it faces a high risk of default. Consequently many positive net present value loans will not be undertaken because equity holders have to share the gains from those loans partially with creditors. As such the investments are less attractive to equity holders and they will require a higher return on equity1 which results in a higher cost of capital.
According to Admati (2011) this debt overhang problem is exacerbated by the implicit government guarantee because debt holders know that the government will intervene in order to prevent bankruptcy and therefore banks are encouraged to use more leverage. The
recapitalizations of many banks all over the world in 2008 and 2009 provide examples of implicit government guarantee. Because banks for example, Royal Bank of Scotland and Bank of Ireland were going to collapse, governments had to spend large amounts of public money to avoid a breakdown of the financial system. According to Barth, Prabha & Yun (2012), Bank of Ireland received a bail out amount of 8.04 billion dollars in 2009 whereas
1
The cost of equity is defined as, 𝑅𝑒 = 𝑅𝑓 + 𝛽 ∗ (𝐸(𝑅𝑚) − (𝑅𝑓), where Re is the cost of equity (the required rate of return on equity), Rf is the risk free rate, 𝛽 is the beta coefficient (unsystematic risk) and (𝐸(𝑅𝑚) − (𝑅𝑓) is the market risk premium.
8 Royal Bank of Scotland received 25.43 billion dollars in 2008. These and many other
financial institutions had become too important to fail. Bank equity holdings were diluted through government intervention but bank’s debt holders did not face the risk of loss, because insolvency was prevented by government intervention. To the extent that banks and creditors did not pay for this guarantee, it can be considered as an implicit subsidy (Noss & Sowerbutts, 2012). Alessandri and Haldane (2009) state that since debt holders know that the government will intervene in order to prevent a default, this encourage excessive leverage and excessive risk taking for banks. They will invest in more risky projects. When banks become riskier, shareholders will require a higher return on their investments to compensate for the higher risk they take, resulting in a higher cost of capital. Since banks are facing higher capital requirements, a higher cost of capital encourage banks to reduce the amount of risk weighted assets instead of raising new equity. Boot (2013) provides an explanation for the higher cost of capital. He states that when a bank has a capital surplus, i.e. capital above the minimum requirement, it uses this capital in order to finance other activities to increase return on equity, which is defined as the net income return as a percentage of shareholders equity. According to Boot (2013), using capital increases the cost of capital and may not create value at all. He states that shareholders and other market participants can anticipate that banks will use capital and thus ask for a higher return on equity. This confirms the belief of banks that equity capital is expensive and that the best way to deal with higher capital requirements is to increase risk on the short-term in order to realize the required return in the future.
According to Elliot (2009) (equity) capital is expensive because the required expected rate of return on equity is higher than that on debt. When banks are required to use more of this expensive form of funding, its overall cost of capital would increase. If capital would be cheap, banks would be encouraged to hold high levels of capital. Banks would then be much safer and protected against events such as crises. Unfortunately, the suppliers of capital ask for high returns because their role is to bear the risk from a bank’s loan book, investments and operations. So when capital requirements (Tier 1) increase and banks face high cost of capital, it is more expensive to raise new capital and a bank is more likely to reduce the amount of loan supply.
The increase in capital requirements under Basel III was a response on the financial crisis which revealed that banks capital requirements were still too low. Since the financial crisis banks were facing bankruptcy problems (Ivashina & Scharfstein, 2010). As a
consequence many governments around the globe were forced to rescue big troubled banks that were too important to fail. As governments issued more bonds to finance these bail outs and to prevent recessions, sovereign risk rose along with high levels of debt to GDP ratios
9 (Barth, Prabha & Yun, 2012). In 2010 for example, the debt to GDP ratio in Greece reached 140% and in Italy 110% (Hamdi & Sbia, 2013). In some countries debt became riskier than the risk of banks holding the debt. Acharya et al. (2010) state that the bail out of banks led to a shift of credit risk from the financial sector to sovereign risk, which is a form of country risk, where governments are not able to repay their outstanding loans to banks. For over ten years since the introduction of the Euro, the yields of 10-year bonds issued by European countries have been low and stable. However, as a result of country risk a government face, banks require a higher compensation as counterpart for the greater risk leading to increasing interest rates on the 10-year government bonds (De Bruyckere et al., 2013). Therefore banks can make higher profits by providing more government bonds and at the same time reducing loan supply to corporations and consumers because banks has to meet the Tier 1 capital requirement. In order to study how banks will respond to country risk, when they also have to face a higher Tier 1 capital ratio, this paper will test the following hypothesis: does country risk reduce the amount of bank loan supply when the Tier 1 ratio increases?
I expect that country risk reduces the amount of bank loan supply when the Tier 1 capital ratio increases.
This effect should also differ between banks in different countries. In 2008 Portugal, Italy, Ireland, Greece and Spain (PIIGS) suffered more than other European countries in terms of economic downturn. This is characterized by the macro-economic indicators such as decline in the development rate, increasing deterioration in the budget balance, increase in the debt rates, growth in the current account deficit, high decreases in the value of import and export, decrease in the rates of total investments and increase in the rates of unemployment
Cukurcayir & Tezcan (2013). PIIGS have been experiencing government budget deficits since the financial crisis, because these governments were not able to meet sovereign debt payments after they used too much debt to bail out banks. This resulted in declining levels of their GDP growth rates. The average real GDP growth rate between 2008 and 2013 was -23.8% in Greece, -11.2% in Ireland, -10.1% in Spain, -9.2% in Italy and -7.0% in Portugal (Martin & Waller, 2012). Lower real GDP growth rates implies that the aggregate demand level in a country is low, which is the total demand for final goods and services in an economy at a given time. This implies that the PIIGS countries were facing weaker economies compared to other European countries. When the economy is in a weak form, it is harmful for companies (including banks) since consumers and other corporations are less likely to purchase products and do investments. For this reason also banks in PIIGS countries face weaker economies and consequently lower levels of loan demand since part of these expenditures are being financed through bank loans. Thus is should matter for a bank if it is located in for example Italy or
10 Germany since a bank in Italy has to deal with increasing capital requirements plus lower demand levels of bank loans whereas a German bank does not face lower levels of loan demand. Both banks are expected to reduce loan supply due to an increase in capital
requirements but this effect should be higher in Italy since supply is already reduced in order to meet the lower demand of loans.
2.2 Empirics
Baker & Wurgler (2013) studied the relation between capital requirements and cost of capital. They used data ranging from 1996 until 2011 with a sample consisting of 4000 publicly traded banks. They found that a 1 percentage-point increase in the required Tier 1 capital ratio increased the overall cost of capital by as much as 90 basis points per year.
Aiyar, Calomiris, Hooley, Korniyenko & Wieladeck (2014) estimate the impact of changes to capital requirements on cross-border bank loan supply of banks in the UK where they focuses on loans to non-banks and banks in other countries than the UK. By examining a sample in which each recipient country has relationships with UK-resident banks, they can control or demand effects. Their sample contains of data from 1996 until 2006. They found that a 1% increase in the capital requirement is associated with a reduction in the growth rate of cross-border credit of 5.5%.
Bridges et al. (2014) examine the effect of bank-specific, time-varying capital
requirements in the UK between 1990 and 2011. The main finding of this paper is that banks increased their capital ratios in order to hold buffers above the minimum requirement, by reducing the amount of loan supply. Bridges et al. studied the effect of changing regulatory capital requirements on bank capital and bank lending in different sectors in the economy such as households, commercial real estate corporations, non-real estate and non-financial corporations. For lending to households, they found that a 1% increase in capital
requirements, leads to a 0.8% decrease in loan supply per quartile. Furthermore they found that banks reduce commercial real estate loans by 4% after one quarter when capital requirements increase with 1%.
Francis and Osborne (2012), uses data on the individual capital requirements set by supervisors for all banks in the UK (roughly 150 banks over the sample period) to examine the effects on capital, lending and balance sheet management behavior for the period 1996-2007. They find that banks raise targeted capital ratios when capital requirements increase. In order to achieve these capital ratios, banks reduce the relatively higher risk weighted asset
11 classes (including loans) by a larger amount (2%), than the non-risk weighted asset categories. Calomiris (2013) studied how loan supply of banks responds to an increase in the required equity ratios in the UK. Estimates show that a one percentage-point increase in required equity ratios reduces the supply of lending to domestic financial firms by about 7%. According to Cimiano and Hachure (2011) and Chaim and Cimiano (2003), increased capital requirements reduce lending because lending rates increase. They studied the impact of capital requirements on bank lending rates and loan supply. Higher costs associated with meeting capital requirements through the higher cost of capital, force banks to require a higher risk-adjusted rate of return on bank loans leading to higher lending rates. The higher lending rate reduces demand for these loans, because borrowers have to pay higher interest. They found that a 1% increase in the capital to assets ratio lead large banks to increase their lending rates by 6%, causing loan supply to decline by 1.3% in the long run. Peek & Rosengren (1995) investigates the direct link between a regulatory increase in the capital-to-asset ratio requirements and the shrinkage of bank loans. They used a sample consisting of the 150 largest financial institutions in England, for the period 1989 until 1992. The results show that banks with regulatory increase in the capital to assets ratio shrink their banks loans with 6.9% more than banks without. The shrinkage due to loan supply is caused due to weak loan demand associated with weakness in the economy by the time that the regulatory capital to assets ratio was increased.
3. Methodology
In order to test the effect of country risk on the amount of loan supply by banks due to an increase in the Tier 1 capital ratio, this study will build on the model of Aiyar et al. (2014). I will use the following panel data regression model. First, the overall effect of an increase in the Tier 1 capital ratio on the amount of loan supply by banks in the European countries will be tested. Secondly, to test how the effect on the amount of loan supply by banks by a change in the Tier 1 capital ratio depends on country risk, an interaction term is included. This results in the following regression model:
lnLoansit = β0 + β1 Tier1it + β2 Rgovit + β3 Tier1it * Rgovit + β4 lnAssetsit + β5 Spreadit + β6
12 where lnLoansit denotes the amount of corporate loans and loans to consumers in period t by
bank i. Total loans are measured in millions of dollars. The natural logarithm is used in this model because the distribution of total loans is skewed, so that outliers might influence too much (Stock & Watson, 2012). The Tier1it denotes the Tier 1 capital ratio for bank i in period
y. Rgovit , is the 10-year interest rate on government bonds in period t for banks in country i.
When governments are facing county risk, there is a higher risk that governments fail to repay their loans to banks. The higher this risk, the higher is the 10-year interest rate on government bonds, because suppliers of these bonds e.g. banks require a compensation for the risk they take when lending to governments. Including the interaction term (Tier1it * Rgovit), allows the
effect on the amount of loan supply by banks of a change in the Tier 1 capital ratio to depend on the 10-year interest rate on government bonds.
lnAssetsit, the natural logarithm of total assets, controls for bank size. Total assets are
measured in million dollars. The natural logarithm is used in this model for the same reason as for lnLoansit.
The second control variable is Spreadit, which is the interest rate spread in period t by
bank i. The spread is calculated by subtracting the deposit rate from the lending rate. According to Duprey and Lé (2014), the lending rate is defined as the average rate a bank receives from loans and other interest-accruing activities and the deposit rate is the average rate it pays on deposits. The lower the spread, the more interest a bank has to pay compared to the interest is receives. For this reason a lower spread should reduce a banks’ willingness to lend.
The GDP growth rate is the annual percentage change of real gross domestic product (GDPgrowthit). Real GDP growth determines aggregate demand level, which is the total
demand for final goods and services in an economy at a given time. When aggregate demand is low, corporations and consumers have low levels of expenditures. These expenditures can be financed through loans provided by banks, e.g. mortgage loans or corporate loans for investments. For this reason a lower GDP growth rate should reduce the amount of banks’ loan supply, since demand for loans is lower.
In order to control for an increase in price levels, inflation per country is included. Inflation is measured by the consumer price index (CPI). The reason to include this variable is because when inflation rises, the amount of loan supply in European countries is affected since consumers and corporations have less to spend due to an increase in price levels2
2 This does not hold when wages rises by the same amount. When wages rise at the same level as inflation,
consumers and corporations can spend the same amount as before the rise in inflation. The demand for bank loans will not be affected.
13 (Zemanek, 2010). Since part of their expenditures is financed with bank loans, demand for bank loans should decrease. As a result, fewer loans are being applied, which means that the amount of loan supply by banks should reduce.
Because this sample only contains publicly listed banks, the natural logarithm of market capitalization is included as a control variable (lnMVit). Investors can use market
capitalization as an indicator to value a banks’ net worth in order to make their investment decisions. According to Gambacorta (2004) the decrease in lending is lower for well-capitalized banks that are perceived as less risky by the market.
Sovereignit describes a banks’ exposure to sovereign bonds as percentage of the total
debt outstanding. The sovereign bonds exposure is much higher in some countries which are part of the sample. Acharya (2015) finds that banks’ exposure to sovereign bonds’ riskiness is higher in PIIGS3 countries, meaning that banks in PIIGS countries have higher bonds
exposure to the government of their home country. Since banks in PIIGS faced higher
sovereign risk they also faced a higher 10-year interest rate on government bonds. In order to control for this variation between countries this variable is included.
The matrix FEit controls for year fixed effects, which change over time but are fixed
between banks.
Table 1
Summary statistics
Variables Obs Mean Standard
Deviation Minimum Maximum
Dependent variable
lnLoans (mln $) 969 16.83 2.56 6.74 21.43
Explanatory Variables
Tier 1 capital ratio (%) 969 10.46 3.86 0.60 38.4
10-year interest rate (%) 969 5.39 2.92 1.78 22.35
Control variables
Total Assets (bln $) 969 512 6.32 0,199 3,540
Interest rate spread (%) 969 3.15 1.23 -1.11 8.81
GDP growth (%) 969 2.52 3.19 -5.32 11.27
Inflation (%) 969 2.68 1.57 -4.48 10.32
lnMarketcapitalization (mln $) 969 17.42 2.26 7.66 35.52
Sovereign bond exposure (%) 969 35.05 1.54 0.01 64.34
3
14
4. Data
The main data of bank financials that are used in this study is obtained from Osiris and the European central bank (ECB). Osiris provides data for globally listed companies, including banks. Data about bank related variables, loans, Tier 1 ratio, total assets, and the equity to asset ratio are acquired from Osiris. Interest on government bonds, interest rate spread, GDP growth and inflation, which are country related variables are acquired from the ECB. Market capitalization data at country level is retrieved from the World Bank and data about sovereign bonds exposure is collected from the European Banking Authority.
As this paper tries to study whether country risk affect banks’ loans supply in European countries due to an increase in the Tier 1 capital ratio this sample comprises 19 European countries4. The sample is selected on the basis of banks that are directly under supervision by the European Central Bank (ECB). These European banks are also included in the European Banking Authority (EBA) stress tests. The EBA has been monitoring and assessing the impact of the Basel III rules on a sample of EU banks. Therefore this sample contributes to the fact that for every bank, the Tier 1 capital ratio must be available, since this is one of the main variables this paper has to encounter. Only publicly listed banks and bank holding companies are included in the sample. This is because for non-listed banks there was no data available and for some non-listed banks data was available until 2001. Bank holding companies are included because otherwise important banks as The Royal Bank of Scotland and Dexia will be disregarded in this paper. Removing non-listed banks and banks for which there was no or not enough data available, leaves the sample with 63 European banks to test. The Osiris dataset covers data from 1995 until 2015, but the most comprehensive dataset is only available from 2001-2014.
Table 1 show that the natural logarithm of loans varies from 6.74 to 21.43 in the sample, with an average of 16.83. The Tier 1 capital ratio range between 0.6% to 38.40%. The average Tier 1 ratio according to the sample is 10.46%. This illustrates that, on average, banks meet the Tier 1 capital requirement of 6%. Since banks have two options to meet capital requirements, one the one side they can raise new equity and on the other side they can reduce the amount of risk weighted assets, results will show how banks react to increased capital requirements and if it is indeed consistent with theory about cost of capital. The average 10-year interest on government bonds is 5.39% and varies from 1.78% to 22.35%. This illustrates
4
Austria, Belgium, Cyprus, Denmark, Estonia, France, Germany, Greece, Hungary, Ireland, Italy, Malta, Netherlands, Norway, Poland, Portugal, Spain, Sweden, United Kingdom.
15 that some countries in the sample were facing increasing interest rates on 10-year government bonds since this variable varies between 1.78% and 22.35%. This is consistent with the study of de Bruyckere et al., 2013 where they state that when banks lend to governments who face country risk, banks require a higher compensation as counter party for the risk they face.
5. Empirical results
This chapter discusses the results from the baseline model, which estimates the overall effect of an increase in the Tier 1 capital ratio on the amount of loan supply by banks in the
European countries. Secondly it is estimated how country risk affects the amount of loan supply by banks by a change in the Tier 1 capital ratio.
Table 2 shows that without controlling for year fixed effects and factors that could influence the amount of loan supply by banks there is a significant effect at the 5% level with a coefficient of -3.62. Meaning that a 1 unit increase in the Tier 1 capital ratio reduces the amount of loan supply by banks with 3.62%. When controlling for year fixed effects this effect is again significant at 5%, but loan supply is reduced by 2.96%. In model (9) all control variables are included as well as year fixed effects. This illustrates the overall effect of a change in the Tier 1 capital ratio on the amount of bank loans’ supply is to reduce its lending. This is in line with the theory of Ayer, Calamari’s and Wielded, 2014a, 2014b; Francis and Osborne, 2012; Bridges, Gregory, Nielsen, Pezzone, Radial, and Spaltro, 2014, which claims that holding more equity is expensive and that requiring more equity would increase the costs of capital. Since banks can either raise more equity or reduce its risk weighted assets,
according to the results of this paper banks choose the most cost effective way to meet the capital requirement of 6%. Raising more equity can be expensive because there is no tax benefit. It is more attractive for banks to finance their activities with debt. When banks use more leverage they can face debt overhang problems which can lead to bankruptcy. Since the implicit government guarantees support banks that face bankruptcy, banks are not reluctant to use debt. Banks become riskier and shareholders will require a higher return on equity. For these reasons banks are forced to cut back on providing loans to consumers and corporations in order to meet the Tier 1 capital requirement. The results of this paper find evidence that the Modigliani & Miller (1958) theorem does not hold in practice. The cost of capital is not constant, since banks choose to reduce the amount of loan supply instead of increasing equity which is more expensive. The coefficient -3.97 in model (9) implies that a 1 unit increase in the Tier 1 capital ratio, reduces the amount of loan supply by banks with 3.97%, where this
16 coefficient is significant at a 5% level. This is in line with a study of Bridges et al (2014), where they find a negative relation between capital requirements and banks’ loan supply. Their results show that a 1% increase in capital requirements reduced the amount of loans with 4%.
In models (1)-(9), the variable lnAssets is significant at 1% level. The variable lnAssets shows a positive effect. Meaning that a 1% increase in total assets increases the amount of loan supply by banks with 0.85% regarding to model (9). Since bank loans are part of total assets, the higher total assets the more loans a bank can provide.
In models (1) to (9), the coefficient of the interest spread is negative. In fact, this implies that when the spread increases with 1%, banks reduce their amount of loan supply. This is in contrast with the theory of Duprey and Lé (2014) where they state that the higher the spread, the more interest a bank receives compared to the interest it has to pay. Therefore a banks willingness to lend should increase. Since this coefficient is not significant, no conclusions can be drawn from this estimate.
The GDP growth rate is significant at a level of 10%, in model (5)-(8). The coefficient in model (8) is significant at 10% and implies a decrease of 1.18% in the amount of banks’ loan supply, when there is a 1% increase in GDP growth rate. When controlling for year fixed effects this coefficient becomes significant at 5% and implies a decrease of 1.11% in loan supply. Since this relation is negative, it could be that consumers and corporations use other sources of funding when doing expenditures or investments. For example cash or equity funding instead. A second explanation for this negative relation could be a higher interest rate on corporate loans or loans provided to consumers, e.g. mortgages.
The impact of inflation is positively significant for model (6)-(9) at the 10% level. This implies a 1% increase in inflation, increases the amount of loan supply by banks with 2.45% following the estimates from model (9). This positive relation could possibly be explained by the fact that the model does not account for wage increases. When inflation rises, and wages do not rise at the same level, people have less to spend. Resulting in a lower demand for bank loans.
Finally both variables lnMV and Sovereign are not statistically significant. This means that there is nothing to conclude about the impact of market capitalization and sovereign bond exposure on the amount of banks’ loan supply.
17
Table 2
This table shows the regression results of the fixed effect model, with controlling for year fixed effects. The dependent variable is the natural logarithm of loans. This table reports regression results for the dependent variable loans. Model (1) includes no control variables and no year fixed effects and illustrates only the effect of the Tier1 capital ratio on loans. Model (2) controls for only year fixed effects. Model (3)-(8) contains control variables such as the natural logarithm of assets, interest rate spread, GDP growth, inflation, natural logarithm of market capitalization and sovereign bond exposure. Model (9) includes both year fixed effects and control variables whereas model (8) does control for year fixed effects. The regression coefficients are reported without brackets. The robust standard errors are reported in the brackets.
(1) (2) (3) (4) (5) (6) (7) (8) (9)
Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans)
Tier1 -3.62** -2.96** -2.82** -1.52* -1.76** -2.78** -2.95** -3.88** -3.97** (1.12) (1.25) (1.12) (0.98) (0.93) (1.74) (1.54) (1.21) (1.56) lnAssets 0.92*** 0.97*** 0.91*** 0.91*** 0.94*** 0.99*** 0.85*** (0.20) (0.05) (0.06) (0.06) (0.07) (0.08) (0.11) Spread -1.85 -1.88 -1.69 -1.78 -1.85 -1.93 (0.98) (0.71) (0.29) (0.66) (0.99) (0.95) GDP growth -1.25* -1.22* -1.14* -1.18* -1.11** (0.05) (0.04) (0.06) (0.04) (0.05) Inflation 2.04* 2.32* 2.37* 2.45* (0.88) (1.02) (1.27) (1.32) lnMV 1.33 0.79 0.95 (0.11) (0.18) (0.13) Sovereign -2.03 -2.56 (0.54) (0.22) Constant 1.54*** 1.42*** 1.73*** 1.23*** 1.76*** 1.62*** 1.55*** 1.19*** 1.17*** (0.33) (0.37) (0.22) (0.25) (0.29) (0.14) (0.19) (0.28) (0.24) Year fixed
effects NO YES YES YES YES YES YES NO YES
Observations 969 969 969 969 969 969 969 969 969 R2 0.12 0.14 0.28 0.32 0.39 0.47 0.50 0.55 0.56 Adjusted R2 0.1 0.13 0.25 0.29 0.37 0.46 0.47 0.54 0.55 Standard errors in parentheses
* Two-sided significance at 10% **Two-sided significance at 5% ***Two-sided significance at 1%
Table 3 shows the regression results with respect to country risk when controlling for year fixed effects. Since this paper studies the effect of country risk on the amount of loan supply by banks due to an increase in the Tier 1 capital ratio, an interaction term is added to the
18 regression. By adding the interaction term, the regression tests if the effect of 10-year interest rate on government bonds on the amount of loan supply by banks depends on the Tier 1 capital ratio.
Model (1)-(4) does not include the interaction term, in order to show the effect of the Tier 1 capital ratio and country risk on the amount of bank loan supply separately. Model (1) does not include year fixed effects and control variables. The coefficient of -1.78 is
statistically significant at a 10% level, meaning that a 1 unit increase in the Tier 1 capital ratio reduces the amount of loan supply by banks with 1.78%. The second key variable is the 10-year interest on government bonds which is not significant in model (1). However when controlling for year fixed effects in model (2) this coefficient (-1.98), becomes negatively significant at 10%. Illustrating that the amount of loan supply by banks decreases with 1.98% when the 10-year interest rate increases with 1%. In the model (4), where both year fixed effects and control variables are included, the effect of the Tier 1 ratio (-1.87) and 10-year interest rate (-1.78) are both negatively significant at 5%. Meaning that banks loan supply decreases with 1.87% (1.78%) when there is a 1 unit (%) increase in the Tier 1 ratio (10-year interest rate). When governments face country risk, the interest rate on government bonds increases. Therefore, banks can earn higher profits by providing government bonds. On the other side, when the Tier 1 capital ratio increases, a bank will choose to provide fewer loans to consumers and corporations since it receives lower payoff compared to government bonds.
When including the interaction term in model (5)-(9), the Tier 1 coefficient is significance at the level of 5% as well as the 10-year interest rate when including control variables. The coefficient on Tier1* Rgov is the effect of a 1% change in the Tier 1 ratio and 10-year interest rate on government bonds above and beyond the sum of the individual effects of a 1 unit change in Tier 1 alone and a 1% change in the 10-year interest on government bonds alone. The coefficient on the interaction term becomes significant at the level of 5% when adding control variables in model (8) and (9). Controlling for year fixed effects in model (9) increases the coefficient from 0.52 in model (8) to 0.71. Since the coefficient -0.71 is statistically significant; it implies that the effect on loan supply by banks of a change in the Tier 1 ratio indeed depends on country risk. When a bank provides a loan to a
government who faces country risk and the 10-year interest on this bond is for example 5%, the estimated effect of a 1 unit increase in the Tier 1 ratio results in a reduction in the amount of loan supply by banks of -5.46%, [-1.91 + (-0.71*5)] which is the main finding of this paper. Results of this paper show that when capital requirements increase and banks do not face country risk they reduce loan supply with 3.97% due to the higher cost of capital (table 2). The result in model (9) of table 3 implies that this effect is stronger (5.46%) when a bank
19 faces county risk. A possible explanation for this could be the fact that banks prefer to
increase lending to governments since they receive higher interest rate on these bonds which is consistent with findings of De Bruyckere et al., (2013). They argued that banks require a higher compensation for the fact that they have to face sovereign risk which implies an
increase in the interest rate on government bonds. Therefore banks can make higher profits by providing more government bonds and at the same time reduce loan supply to corporations and consumers in order to meet the higher Tier 1 ratio of 6%.
Table 3
This table shows the regression results of the fixed effect model, with controlling for year fixed effects. The dependent variable is the natural logarithm of loans. This table reports regression results for the dependent variable loans. Model (1) -(2) includes the measure of country risk, no control variables and no interaction term. Model (3) -(4) includes the measure of country risk, all control variables but no interaction term . Model (2) and (4) controls for year fixed effects. Model (1)-(4) test for the effect of the Tier 1 ratio and country risk as two single variables without interacting. Model (5) and (6) test the effect when both Tier 1 and country risk are interacted, but without any controls. Model (7) and (8) includes the interaction term as well as all control variables such as the natural logarithm of assets, interest rate spread, GDP growth, inflation,
natural logarithm of market capitalization and sovereign bond exposure. Where the complete model (8) controls also for year fixed effects. The regression coefficients are reported without brackets. The robust standard errors are reported in the brackets.
Standard errors in parentheses * Two-sided significance at 10% **Two-sided significance at 5% ***Two-sided significance at 1%
(1) (2) (3) (4) (5) (6) (7) (8)
Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans)
Tier1 -1.78* -1.66* -2.42** -1.87** -2.03** -1.86** -1.82** -1.91** (0.36) (0.19) (1.29) (1.41) (1.12) (0.68) (0.48) (0.73) Rgov -1.45 -1.98* -2.25** -1.78** -2.61* -1.88* -1.25** -1.82** (0.65) (0.44) (1.85) (1.65) (1.52) (0.49) (0.33) (0.45) Tier1*Rgov -0.49* -0.68* -0.52** -0.71** (0.21) (0.51) (0.30) (0.36) Constant 1.58*** 1.52*** 1.36*** 2.14*** 2.41*** 2.01*** 1.75*** 2.69*** (0.74) (0.69) (0.88) (1.02) (0.86) (0.77) (0.69) (1.10) Year fixed
effects NO YES NO YES NO YES NO YES Controls NO NO YES YES NO NO YES YES
Observations 969 969 969 969 969 969 969 969 R2 0.28 0.24 0.31 0.36 0.33 0.30 0.39 0.41
20
5.1 Robustness
In order to show that the main findings of this paper are robust, an alternative subsample is used to test the effect of an increase in the Tier 1 ratio and country risk on the amount of banks’ loan supply. Model (1) and (2) compares banks in PIIGS countries with banks in non PIIGS countries and test whether there are differences in the reduction of loan supply by banks. Acharya (2015) finds that banks’ exposure to sovereign bonds’ riskiness is higher in PIIGS countries, meaning that banks in PIIGS countries have higher bonds exposure to the government of their home country. In other words the majority of the PIIGS sovereign debt is held by these countries’ domestic banks. Since Portugal, Italy, Ireland, Greece and Spain face more country risk, banks with sovereign bond exposure in PIIGS countries face more country risk. As a result the interest rate on government bonds increases. In order to compensate investors for the risk of these banks, the cost of capital for banks in PIIGS countries is higher. Furthermore these banks can gain higher payoffs by providing government bonds due to the higher interest on these bonds. Therefore banks in PIIGS countries should reduce their amount of loan supply to consumers and corporations, due to an increase in the Tier 1 capital ratio. This reduction should be larger than the reduction for banks in non PIIGS countries. In order to test this relation, a dummy is used to make a distinction between PIIGS countries and non-PIIGS countries. The dummy takes the value 1 for the 10-year interest rate in PIIGS countries and 0 for non-PIIGS countries.
An important finding of model (1) and (2) is that banks in PIIGS countries react more heavily to an increase in the Tier 1 ratio than non-PIIGS countries. This can be derived from the coefficient -0.45 which is statistically significant at 1%. In other words, when the 10-year interest rate is for example 5%, banks in PIIGS countries reduce the amount of loan supply to a higher extent, namely with 3.44% [-1.19 + (-0.45*5)] than banks in non-PIIGS countries. When controlling for year fixed effects results show that banks in PIIGS reduce their loan supply with 3.71% [-1.26 + (-0.49*5)] compared to non-PIIGS countries.
This result can be an important contribution to policy considerations about capital requirements under Basel III. Currently capital requirements are the same across countries. Since the results of this paper show that banks in PIIGS countries react differently on the Tier 1 ratio than banks in non-PIIGS countries, it should be considered to have different capital requirements for different countries. Banks in PIIGS countries reduce their loan supply by a higher extent than banks in non-PIIGS because they are confronted with low levels of loan demand. Loan demand is low because PIIGS countries have weaker economies compared to non-PIIGS countries due to e.g. lower average GDP growth rates between 2008 and 2013.
21 This will result in a downward spiral where lower economic growth leads to low levels of demand which results in a reduction of loan supply. Simultaneously increasing capital requirements force banks reduce loan supply and there are no possibilities to stimulate the economy. Therefore countries that are facing country risk should face lower capital
requirements in order to stimulate these economies. So that consumers and corporations can increase their demand for bank loans and raise expenditures and investments.
Model (3) and (4) shows the second robustness check of this paper where the Tier 1 capital ratio is replaced by the equity to assets ratio in period t by bank i. This ratio is defined as the market value of equity (E) divided by assets (A), E/Ait, and is calculated by total equity
divided by total assets. In other words this ratio describes the percentage of a bank’s assets that are owned by investors. Since the equity to assets ratio is almost the same as the Tier 1 ratio, where risk weighted assets are used instead of total assets, a robustness test is done in order to test how this ratio affects the amount of bank loan supply. According to Calomiris (2013) forcing banks to raise their equity to asset ratio requirement will reduce banks’
willingness to lend. When banks have to increase their equity to asset ratio, they often choose to cut back on new loans, because it avoids the high costs associated with raising new equity. The equity to assets ratio coefficient of -4.78 implies a significant effect at 1% on loan supply in model (4). This illustrates a 1% increase in the equity to asset ratio, decreases the amount of loan supply by banks with 4.78%, when controlling for year fixed effects. An explanation for this could be the fact that it is more expensive for banks to raise equity, because of the higher costs of equity. Consequently banks choose to reduce the amount of loan supply when the equity to asset ratio increases. This is consistent with a study of
Calomiris (2013), where his results conclude a 1% increase in equity ratios reduces the supply of lending to firms by about 7%. Since the equity to asset ratio is almost the same as the Tier 1 ratio, different capital requirements should be considered as well. Both of these
requirements reduce loan supply to consumers and corporations, which does not progress economic growth. Therefore the minimum leverage ratio should be considered instead of the Tier 1 ratio. This is a non-risk-based leverage ratio and is calculated by dividing Tier 1 capital by the banks’ average total consolidated assets. Since banks have now the possibility to reduce other assets then only risk weighted assets, this ratio should be increased instead of increasing the Tier 1 ratio in order to protect consumers and corporations against lower availability of loan supply.
22
Table 4
Model (1) and (2) of this table shows a robustness test for the effect of country risk in PIIGS versus non PIIGS, where model (1) includes no year fixed effects and model (2) includes this fixed effects. Secondly a robustness test is done for the equity to asset ratio with (4) and without (3) fixed effects. The regression coefficients are reported without brackets. The robust standard errors are reported in the brackets.
(1) (2) (3) (4)
Ln(Loans) Ln(Loans) Ln(Loans) Ln(Loans)
Tier1 -1.19** -1.26*** (1.10) (0.86) dumPIIGS -1.02** -1.18*** (1.33) (1.88) Tier1*dumPIIGS -0.45*** -0.49*** (0.18) (0.23) Equity/Asset ratio -4.67*** -4.78*** (1.56) (1.36) Constant 2.86*** 2.57*** 1.39*** 1.20*** (1.43) (1.29) (0.88) (0.65)
Year fixed effects NO YES NO YES
Controls YES YES YES YES
Observations 605 605 605 605
R2 0.43 0.39 0.48 0.51
Adjusted R2 0.42 0.37 0.45 0.49
Standard errors in parentheses * Two-sided significance at 10% **Two-sided significance at 5% ***Two-sided significance at 1%
23
Conclusion
This paper aims to test the effect of country risk on the amount of bank loan supply when the Tier 1 capital ratio increases. Firstly, the effect of the Tier 1 capital ratio on the amount of bank loan supply is tested for all 19 European countries for the time period 2000 until 2014. By using a panel regression model, with annual data this gives significant results. A 1 unit increase in the Tier 1 capital ratio reduces the amount of bank loan supply with 3.97%. This is consistent with the major studies in this field [(Aiyar, Calomiris and Wieladek (2014); Francis and Osborne (2012): Bridges et al. (2014)]. When testing for the effect of country risk on the amount of loan supply when capital requirements increase, this paper finds again significant results. The main finding of this paper shows that banks reduce the amount of loan supply with 5.46% when they face country risk and when the Tier 1 capital ratio increases. Results conclude that when banks face country risk they reduce the amount of loan supply by a larger amount compared to a situation in which banks do not face country risk. This conclusion can contribute to the overall conclusion whether capital requirements are sufficient or not. The larger decrease in loan supply when banks face country risk can have adverse effects for consumers and corporations since they are limited in order to do expenditures and investments. This can result in lower economic growth since lower expenditure and
investments results in lower productivity, lower employment lower taxes etc. When capital requirements for banks decrease, they can provide more loans, which will stimulate the economy. Governments facing country risk will also benefit from this economic growth, which can contribute to reducing country risk.
The data in the sample represent a large part of banks in PIIGS countries. PIIGS countries have higher exposure to country risk and therefore the results might not be applicable to all banks in Europe. A robustness check for this sample shows indeed that the effect on loan supply is much stronger for PIIGS countries.
In order to have a better understanding in the connection between adjustments of capital requirements and loan supply it is important to relate the type of capital adjustment to loan supply. Bank loan supply should be regressed on different requirements imposed under Basel III to see whether there are requirements that have minimum effect on loan supply.
24 References
Admati, A.R., P.M. DeMarzo, M.F. Hellwig en P. Pfleiderer (2011) Fallacies, irrelevant
facts, and myths in the discussion of capital regulation: why bank equity is not expensive.
Werkdocument van Stanford University.
Aiyar, S., Calomiris, C. W., Hooley, J., Korniyenko, Y., & Wieladek, T. (2014). The international transmission of bank capital requirements: Evidence from the UK. Journal of
Financial Economics, 113(3), 368-382.
Baker, M., & Wurgler, J. (2013). Do strict capital requirements raise the cost of capital?
Banking regulation and the low risk anomaly (No. w19018). National Bureau of Economic
Research.
Barth, J. R., Prabha, A., & Yun, G. (2012). The Eurozone financial crisis: Role of
interdependencies between bank and sovereign risk. Journal of Financial Economic Policy,
4(1), 76-97.
Barrios, S., Langedijk, S., & Pench, L. R. (2010). EU fiscal consolidation after the financial crisis Lessons from past experiences. Bank of Italy Occasional Paper.
Basel Committee on Banking Supervision, 2010a, Basel III: International Framework for Liquidity Risk Measurement, Standards and Monitoring.
Basel Committee on Banking Supervision, 2010b, Basel III: A Global Regulatory Framework for More Resilient Banks and Banking Systems.
Bernoth, K., & Wolff, G. B. (2008). Fool the markets? Creative accounting, fiscal
transparency and sovereign risk premia. Scottish Journal of Political Economy, 55(4), 465-487.
Bridges, J., Gregory, D., Nielsen, M., Pezzini, S., Radia, A. and Spaltro, M. (2014). The impact of capital requirements on bank lending. Working Paper No. 486.
Boot, A. W. (2014). Financial sector in flux. Journal of Money, Credit and Banking, 46(s1), 129-135.
Calomiris, C. W. (2013). Reforming banks without destroying their productivity and value.
Journal of Applied Corporate Finance, 25(4), 14-20.
Chami, R., T. Cosimano (2003), “The nature of Capital Adequacy Constraints under
the Basel Accord”, IMF
Cosimano, T. F., & Hakura, D. (2011). Bank behavior in response to Basel III: A cross-country analysis.
Cukurcayir, S., & Tezcan, K. (2013). Investigations on the euro area public debt crisis: The case of piigs. European Scientific Journal.
De Bruyckere, Valerie, et al. "Bank/sovereign risk spillovers in the European debt crisis."
Journal of Banking & Finance 37.12 (2013): 4793-4809.
25
Working Paper No. 499.
Francis, W. B., & Osborne, M. (2012). Capital requirements and bank behavior in the UK: Are there lessons for international capital standards?. Journal of Banking & Finance, 36(3), 803-816.
Gambacorta, L. (2005). Inside the bank lending channel. European Economic Review, 49(7), 1737-1759.
Guichard, S., Haugh, D., & Turner, D. (2009). Quantifying the effect of financial conditions in the euro area, Japan, United Kingdom and United States.
Hamdi, H., & Sbia, R. (2013). Fiscal policy and economic growth in PIIGS countries: an empirical assessment. Journal of Applied Business Research, 29(5), 1343.
Herring, R. J. (2004). The subordinated debt alternative to Basel II. Journal of financial
stability, 1(2), 137-155.
Jackson, P, C. Furfine, H. Groeneveld, D. Hancock, D. Jones, W. Perraudin, L. Redecki, M. Yoneyama (1999), “Capital Requirements and Bank Behaviour: The Impact of the
Basle Accord”, Basle Committee on Banking Supervision, Working Paper
Martin, F. M., & Waller, C. J. (2012). Sovereign debt: a modern Greek tragedy. Review, 94 Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation finance and the theory of investment. The American economic review, 48(3), 261-297.
Myers, S. C. (1977). Determinants of corporate borrowing. Journal of financial economics,
5(2), 147-175.
Noss, J., & Sowerbutts, R. (2012). The implicit subsidy of banks. Bank of England Financial
Stability Paper, (15).
Peek, J. & Rosengren, E. (1995). Bank regulation and the credit crunch. Journal of Banking &
Finance, 19(3), 679-692.
Sgherri, S., & Zoli, E. (2009). Euro area sovereign risk during the crisis. IMF working papers, 1-23.
Thakor, A. V. (1996). Capital requirements, monetary policy, and aggregate bank lending: theory and empirical evidence. The Journal of Finance, 51(1), 279-324.
Van den Heuvel, S. J. (2002). The bank capital channel of monetary policy. The Wharton
School, University of Pennsylvania, mimeo.
Zemanek, H. (2010). Competitiveness within the euro area: the problem that still needs to be solved. Economic Affairs, 30(3), 42-47.
