1
‘Too Big to Fail’ or ‘Too Small to Survive’?
Commercial bank size, commercial bank risk and the moderating effect of national culture post-crisis in the European Union
Supervisor: Dr. N. Selmane Student number: s2362252 Name: Martijn van Lier
Study Programme: MSc IFM
Other UoG programs for which you have submitted this proposal: x
Field Key Words: Bank risk, bank size, national culture, commercial banks, non-performing
loans, Hofstede, cross-country studies.
Abstract
2
1. Introduction
Bank risk taking has got a lot of attention in the literature, especially after the global financial crisis of 2007-2009. During the crisis it turned out that there were huge cross-country differences in non-performing loans. Non-performing loans are a reflection of bank risk taking behavior. These loans are defined as the loans overdue by more than ninety days. They can be divided into household, commercial bank and mortgage loans (Louzis et al., 2011). The amount of non-performing loans in an economy is a critical indicator of financial stability and for sustainable and rapid economic growth financial stability is considered as an absolute necessity. Non-performing loans are also a reflection of the efficiency of the resource allocation process (Ranjan and Dhal, 2003). Furthermore, they can be an indicator of the start of a banking crisis (Reinhart and Rogoff, 2010). Therefore it is of crucial importance for financial risk managers, regulatory authorities and policy makers to know what the determinants are of non-performing loans and bank risk.
One determinant of bank risk which is investigated a lot is bank size. The relationship between bank size and bank risk has been researched extensively over the past decades with multiple risk measures besides non-performing loans. The results however are inconsistent. Empirical evidence is available for a positive, negative and non-linear relationship between bank size and bank risk. Modern portfolio and intermediation theory suggests that large banks are better able to diversify their risks due to their size (Markowitz, 1952; Boyd and Runkle, 1993). However, when banks become so big that the government will bail them out in case of bankruptcy, they are incentivized to take on more risks. In other words they become ‘too big to fail’ (Financial Stability board, 2010; Dam and Koetter, 2012; Kaufman, 2014). Other scholars have combined both arguments resulting in a non-linear relationship between bank size and bank risk (de Haan and Poghosyan, 2012). This means that there is a negative relation, up to a certain threshold. When banks become ‘too big to fail’ the relationship switches to a positive one.
3
influences the decision making of human participants and therefore affects bank risk taking behavior. This leads to the following research question:
How does national culture influence the relationship between commercial bank size and commercial bank risk in the European Union in the post-crisis period?
This study uses a sample of 188 commercial banks in 26 countries of the European Union in the period 2012-2015. Most of the literature on bank size and bank risk US based (e.g. Boyd and Runkle, 1993; Demsetz and Straham, 1997; Stiroh and Rumble, 2006) therefore this study is comprised of European banks only. Commercial banks are financial intermediaries with high leverage that use customer deposits to make loans to firms and individuals. They offer mainly wholesale and retail services (Hefferman, 2005; Giuliano, 2010). This research puts focus on commercial banks since they adjust their risk profiles much slower than investment banks due to their longer time horizon (Haberman, 1987). Therefore the influence of the economic cycle on the results is reduced. Hofstede’s cultural dimensions are used as proxies for national culture, which assumed are constant over the sample period (Hofstede, 2001). These dimensions are on the following levels: power distance, uncertainty avoidance, individualism, masculinity and long-term orientation. National culture was one of the factors that caused cross-country differences in bank risk-taking in the pre-crisis period (Ashraf et al., 2016). After the crisis the regulatory powers in the banking industry shifted increasingly from the national to the European level (Sum, 2016). Due to this convergence there is less room for national flexibility. The post-crisis period is selected to see if national culture is still of influence. Control variables are used on bank, industry and country level. The study controls for bank growth, industry concentration and GDP per capita (Ashraf et al., 2016).
4
commercial bank size and commercial bank risk will be strengthened in countries that score high on power distance, uncertainty avoidance and long-term orientation.
This study contributes to the literature of the international financial management topic in multiple ways. First of all, the vast majority of research into non-performing loans is based on economic theories which assume that people act as rational agents. Either macroeconomic variables, like unemployment, monetary conditions (Rinaldi and Sanchis-Arellano, 2006) and the business cycle (Boss et al. 2009) or microeconomic variables, like bank-specific characteristics as efficiency or moral hazard problems (Berger and DeYoung, 1997) are used as explanatory variables. However, social scientists have discovered the last decades that people do not tend to act as rational agents but have all kinds of irrational flaws in their behavior (Kahneman and Tversky, 2011). Therefore, social theories can provide an excellent help in order to get a better understanding of the concept of non-performing loans. To my knowledge, this is the first study that combines finance theories with social theories in the research between bank size and bank risk. Secondly, this study contributes to a better understanding of national culture and to what extent international managers need to take it into account. In a highly globalized financial world the need to do business in a variety of cultures is an irreversible necessity. Therefore it is essential to know how national culture influences the business world. If national culture still affects bank risk this means that managers need to adapt the strategy of the bank to the risk tendencies of the different countries they operate in. Thirdly, this study contributes to policy makers and financial regulators. They have to take into account the cultural differences between member states when drafting and implementing new directives and laws if they want to be effective instead of a ‘one size fits all’ approach. In practice this will turn out to be very hard because it clashes with fair competition and a level playing field. Finally, investors need to take national culture into account when managing their portfolios in order to optimize the risk-return tradeoff.
The rest of the paper is organized as follows. Section 2 reviews previous studies and relates my contribution to the literature, Section 3 provides the hypotheses development and the conceptual model. Section 4 explains the data and research methods, Section 5 presents the empirical results. Section 6 concludes and provides limitations and suggestions for further research.
5
The literature review covers the following topics. Section 2.1 provides a general description of banks and their risk taking behavior. Section 2.2 goes into the relationship between bank size and bank risk. Section 2.3 elaborates on national culture and its main influences. Section 2.4 describes the changes in banking regulation post-crisis in the European Union and the implications it has for this study.
2.1 Bank risk-taking behavior
Banks have an important role in national economies. They function as the middleman
between people with a surplus of capital and shortages. In this role they offer a wide array of products namely savings, lending, investment, mediation and advice, payment ownership, guarantee and trust of real estate. Because of their vital role in the economy it is important to know what the determinants are of bank performance. Not only for the sustainability of the banking or financial sector itself, but the economy as a whole (Aduralere and Olufemi, 2016).
In order to perform and survive risk taking is an absolute necessity for corporations (Li et al., 2013). This paper builds on the extant literature of corporate risk-taking in an
international setting. Factors that influence corporate risk-taking can be divided into two categories: country-level influences and firm-level influences. On the country-level are legal origin (common law vs code/civil law), the nature of the financial system (market vs bank-based) and the strength of investor protection. These factors are part of the institutional environment and all have significant influence on corporate risk taking. Companies who operate in common law countries, in a market-based financial system with strong investor protection tend to take more risk. (Claessens et al. 2000; John et al. 2008; Acharya et al. 2011). Factors on the firm-level that influences corporate risk-taking are equity based compensation schemes based on the agency theory (Jensen and Meckling, 1976).
While building on the existing literature of corporate risk-taking it must be said that the banking sector differs fundamentally from other sectors in the economy. The risk taking incentives and opportunities in the banking sector are greater relative to other industries. This is due to the fact that the bank equity has the nature of a call-option (Merton, 1977).
6
explicit deposit insurance scheme bank risk-taking increases due to moral hazard problems (Demirguc-Kunt and Detragiache, 2002; Hoque et al. 2015). Besides the regulatory
determinants of bank risk-taking are macroeconomic indicators such as unemployment, inflation and GDP growth (Chaibi and Ftiti, 2015). Paligorova and Santos (2013) have argued, based on the historically low interest rates, that monetary policy influences bank risk-taking behavior as well. On industry-level, banking industry structure has a significant influence on the risk-taking behavior of banks. In more concentrated industries there are less incentives for banks to take risk (Luo et al., 2016; Martinez-Miera and Repullo, 2010). On firm-level, bank risk-taking is influenced by the ownership structure. Banks tend to take more risks when they have more powerful owners. This is due to the fact that shareholders have more incentives to take risk than creditors (Laeven and Levine, 2009).
2.2 Bank size and bank risk
One of the major determinants of bank risk-taking that has not been mentioned in the previous paragraph is bank size. The literature on this relationship is quite extensive. In order to give an inclusive and clear overview, the existing research is divided in three strands. There are proponents of a negative, positive and a non-linear relationship between bank size and bank risk.
The first stream argues that there is a negative relationship between bank size and bank risk. This view is based on the modern intermediation theory and modern portfolio theory. Modern intermediation theory implies that larger banks can achieve cost efficiencies through economies of scale and have therefore bigger buffers, so they are less likely to fail. In other words, they are less risky (Boyd and Runkle, 1993). Modern portfolio theory implies that larger banks are better able to reduce risk through diversification of their operations (Markowitz, 1952). Proponents of the first stream argue that splitting up banks could lead to greater volatility, so less stability. The evidence for this point of view is considerable. Boyd and Runkle (1993), using data from 122 US Holding banks in 1971-1990, find a negative relation between bank size and bank risk, using earnings volatility as bank risk measure. The results are consistent with Kasman and Kasman (2016) who used Turkish commercial banks in the 2002-2012 period. The same applies for the Pakistani commercial banking sector (Afzal and Mirza, 2012). Moutsianas and Kosmidou (2016) found a negative relation between bank size and bank risk as well, but only after a certain threshold. Their study contained
7
diversified but use this advantage to perform riskier strategies. So in the end it does not lead to less risks. They used data from US bank holding companies from 1980-1993. Furthermore Stiroh and Rumble (2006) did not find evidence for a relation between bank size and earnings volatility at all. They used data from 1816 US financial holding companies in 1997-2001. Laeven and Levine (2005) found that financial conglomerates have a higher market value if they were broken up in individual entities. In other words, they did not find evidence for risk reduction through diversification.
The second stream argues that there is positive relation between bank size and bank risk. This view is based on moral hazard problems, the market risk hypothesis and
complexity. Moral hazard implies that people tend to take more risks when they are insured (Demirguc-Kunt and Detragiache, 2002). When applied to the banking industry this means that banks tend to take more risks when they expect that the government will bail them out when they go bankrupt. This is the case when they become ‘too big to fail’ (TBTF) (Dam and Koetter, 2012; Kaufman, 2014). According to the Financial Stability board (2010), an
international supervisor of the global financial system, TBTF banking institutions are those whose disorderly failure could cause significant disruption to the functioning of financial markets and the economy as a whole because of their size, importance, complexity and systemic interconnectedness. As a result, funding costs for banks who are TBTF are lower compared to banks who are not TBTF (Baker and McArthur, 2009). The TBTF status also results in higher credit and bond ratings (Rime, 2005; Morgan and Stiroh, 2005). Brewer and Jagtiani (2013) found that in order to become a TBTF financial institution banks are willing to pay a premium. Ashraf (2017) has linked political institutions with bank risk-taking behavior. He argued that sound political institutions increases the chances of a government bailout. Proponents of the second stream argue that banks should be split up in order to increase stability (Kaufman, 2014). De Nicolo (2000) finds, in the 1988-1998 period, that bank insolvency risk increases with size. Benefits from size-related advantages were not found or did not weigh-up against an increase in risk-taking. The only exception were small-US bank holding companies. The second argument for a positive relationship between bank size and bank risk is based on the market risk hypothesis (Berger et al. 2016). When banks expand their business cross-country they will face new local environments with which they need to cope. These market specific factors increase the complexity and therefore increase the risk. Examples of these market specific factors are foreign exchange risk, local competition, local culture and economic and political stability. Foreign assets are more risky if they are
8
establish new lending relationships, capture market share and learn the locals informal institutions like language and preferences (Chari and Gupta, 2008; Li and Guisinger, 1992; Brewer and Rivoli, 1990). Berger, El Ghoul, Guedhami and Roman (2016) found evidence for the market risk hypothesis using 15.988 US Commercial Banks in the period from 1989-2010. Furthermore Beck, Demirgüç-Kunt and Levine (2006) found out that there is a positive
correlation between bank size and organizational complexity. Increased complexity could lead to a decrease in transparency. As a consequence it will be more difficult for management to keep control and oversight which would result in higher operational risks (Fu et al., 2014). All things considered, the evidence for the second stream is considerable as well. Using the largest 270 banks in 48 countries, from 1996-2001, Laeven and Levine (2009) find a positive relationship between bank size and risk. These results are consistent in the US banking sector (Jokipii and Milne, 2011; Berger and Bouwman, 2013). The same applies for commercial banks in Bangladesh and China (Rahman et al., 2015; Zhang et al., 2008). Gonzalez (2004) also found that large banks take more risk in a study with 251 banks from 36 countries. Bhagat, Bolton and Lu (2015) find evidence for the positive relation between the size of financial institutions and risk, in the pre-crisis period (2002-2006) and the crisis period (2007-2009), but not in the post-crisis period (2010-2012). Their study contained commercial banks, investment banks and insurance companies.
The third stream combines elements of the two previous streams by arguing that there is a U-shaped relationship between bank size and bank risk. This group argues that there is a negative relationship, due to the advantages of diversification, up to a certain threshold when banks become ‘too big to fail’. This is the point where the risk increases again, so the
relationship is non-linear and U-shaped. De Haan and Poghosyan (2012) find evidence for this claim using US bank holding companies from 1995-2010. Risk is measured in this study as earnings volatility.
In order to explain some of controversy regarding the relationship between bank size and bank risk, this research adds national culture as a moderating variable into the equation.
2.3 National culture
9
(Hofstede, 2001; Ashraf, 2017). National culture can be measured in multiple ways. In this study Hofstede’s five cultural dimensions are used.
Williamson (2000) has argued that national culture has a direct and indirect influence on managerial decision making and thus the behavior of corporations. He has made a
framework in which he suggests that national culture is a social institution which conditions all lower level institutions like governance structures. Based on this argument scholars have divided the influences of national culture on macro and micro level (Mihet, 2013).
On the macro level culture can explain the institutional, legal and economic environment of a country (Mihet, 2013). Cross-cultural differences explain much of the variance in corporate governance, investor protection, judicial efficiency, financial systems and transparency of accounting systems, (Doidge et al., 2007; Hope, 2003; Radenbaugh et al., 2006; Kwok and Tadesse, 2006). Furthermore national culture influences savings rates and income redistributions which impact economic development (Guiso et al., 2006).
On the micro level culture explains differences between entrepreneurial orientation in countries (Lee and Peterson, 2000). National culture influences decision making even in the top of organizations where managers are assumed to be aware of the cultural differences. Culture had significant effects on decision making of executives (Tse et al., 1988). This effect was still visible two decades later (Graham et al., 2010).
There is another strand of literature that investigates national culture and banking. Recent studies have examined the link between national culture and diverse aspects of banking like corruption in bank lending, dividend policy, accounting conservatism, earnings quality and bank performance in the crisis (Kanagaretnam et al., 2011; Kanagaretnam et al., 2014; Zheng et al., 2013, Zheng et al., 2014; Boubakri et al., 2017).
Multinational corporations have to cope with cross-country differences in national cultures. For these organizations both the national culture of home and host countries can be of importance. Ashraf and Arshad (2017) compared the influence of national culture in home and host countries of foreign bank subsidiaries. They showed that national culture of parent banks home country has a bigger influence on the risk taking behavior than the national culture of the host country. They argue that this is due to the fact that head office decisions are made in home countries. This study does not take into account the distinction between
10
regulation post-crisis provoke the question if there is still room for the influence of national culture (Sum, 2016).
2.4 Post-crisis banking regulation in the European Union
The Basel Committee on Banking Supervision (BCBS) has the purpose to encourage convergence toward common banking approaches and standards. The committee was established by the G-10 in 1974. Since 2009 all of the G-20 countries are represented in the committee. In order to reach their goal the committee issues the Basel accords. Technically, these accords are not binding regulation. The committee formulates recommendations and broad supervisory standards and guidelines. Subsequently member authorities will implement these through their own national systems.
When the global financial crisis hit in 2007 the finance ministers of the G-20
concluded that “(…) important causes of the global financial crises were excessive risk taking and faulty risk management practices in financial markets, (...), as well as deficiencies in financial regulation and supervision”. As a response to the crisis came Basel III with the main goal to strengthen banks capital positions and their resistance to shocks in order to prevent bankruptcies and contagion; to improve risk management; and to enforce better transparency, governance and disclosure in the banking industry (Sum, 2016).
In order to implement Basel III the European Union issued the CRD IV/CRR Package. This package consists of Capital Requirements Directives IV (Directive 2013/36/EU) and Capital Requirements Regulation (Regulation No. 575/2013). While the directive leaves room for national flexibility, the regulation prevents national divergence in the banking regulatory framework. The regulation is part of the ‘Single Rulebook’ that aims to provide a single set of harmonized prudential rules which institutions throughout the EU must respect (EBA, 2016).
11
can also participate in the SSM and SRM, provided they join the banking union (Sum, 2016). Member states are currently debating about the third pillar, a European Deposit Insurance Scheme (EDIS).
Based on this convergence in banking regulation and supervision there is less room for national flexibility and regulatory arbitrage. Therefore it is expected that the influence of national culture will be weaker post-crisis compared to the pre-crisis period. Because of constraints regarding data availability, this research only investigates the post-crisis period.
3. Hypotheses development
This study builds further on previous research regarding the relationship between bank size and bank risk. While the literature is inconsistent, altogether the most evidence is available for a negative relationship between bank size and bank risk (e.g. Boyd and Runkle, 1993; Afzal and Mirza, 2012; Kasman and Kasman, 2016; Moutsianas and Kosmidou, 2016). Therefore this is the expected relationship. This leads to the following hypothesis:
H1. Commercial bank size is negatively associated with commercial bank risk.
In order to explain some of the controversy regarding the inconsistency in the literature national culture is added as a moderating variable in order to test if it affects the relation between commercial bank size and commercial bank risk. Previous research has linked national culture with bank risk taking (Ashraf et al., 2016; Kanagaretnam et al., 2014). However, these studies both used data from the pre-crisis period only. The earlier described changes in post-crisis banking regulation leave less room for the influence of national culture. Therefore this study uses the post-crisis period to see if national culture affects the relation between bank size and risk. Hofstede’s five cultural dimensions are used in order to formulate the rest of the hypotheses. These are individualism, masculinity, power distance, uncertainty avoidance and long-term orientation (Hofstede, 2001).
12
overconfidence (Chui et al., 2010). Furthermore the concern for other stakeholder welfare is likely to be lower in individualistic cultures (Kiridaran et al., 2013). Finally, individualistic decisions have higher variance than decisions in groups (Shupp and Williams, 2008). These arguments suggest that bank risk-taking will be higher in individualistic countries. Building further on the first hypothesis, it is expected that the negative relationship between
commercial bank size and commercial bank risk will be partially offset in individualistic countries. This leads to the following hypothesis:
H2. The negative relationship between commercial bank size and commercial bank risk will be weaker in countries that score high on individualism.
Masculinity (MASC) in national culture refers to the extent of toughness in a country. Femininity refers to tenderness. In masculine countries there is a clear separation of gender roles. Men are though, assertive, competitive and focus on material success. Women are concerned with the quality of life and are modest and tender (Hofstede, 2001). Masculine countries focus on social recognition and ego. In low masculine or feminine countries the roles over gender overlap. These countries focus more on relationships. In masculine
countries people tend to be more self-confident and ambitious. Because achievers are admired and there is room for machismo, masculinity is linked with more financial risk-taking (Meier-Pesti and Penz, 2008). Therefore it is expected that the negative relationship between bank size and risk will be partially offset in masculine countries. This leads to the following hypothesis:
H3. The negative relationship between commercial bank size and commercial bank risk will be weaker in countries that score high on masculinity.
Power distance (PDI) in national culture measures the power inequality between superiors and subordinates. It means the accepted unequal distribution of power in a country (Hofstede, 2001). In countries who score high on power distance individuals have less
freedom and autonomy in the decision making process. This leads to conservatism. Countries who score low on power distance have more social mobility. This means that individuals are trying to improve their positions (Thompson et al., 2009). When applying this to the
entrepreneurial perspective power distance has a significant negative influence on
13
also a negative influence on risk taking (Mihet, 2013). Therefore it is expected that the negative relationship between bank size and bank risk will be strengthened in hierarchical countries. This leads to the following hypothesis:
H4. The negative relationship between commercial bank size and commercial bank risk will be strengthened in countries that score high on power distance.
Uncertainty avoidance (UAI) in national culture refers to the extent to which people of a country feel uncomfortable with uncertain situations. These people feel a need for
predictability in their lives and try to avoid uncertain situations (Hofstede, 2001). In financing and investing decisions however the future is always uncertain. It is therefore unsurprisingly that the amount of uncertainty in financial contracts has a significant influence on financing and investment decisions (Aggarwal and Goodell, 2014). It has to be mentioned that
uncertainty avoidance is not the same as risk-aversion (Hofstede, 2001). However, there is a difference between known and unknown risks. This is the part where high-UAI and low-UAI countries differ. Both countries will take known risks but low-UAI tend to take more
unknown risk. (Ashraf et al., 2016). These arguments suggest that bank risk-taking will tend to be higher in low-UAI countries. Therefore it is expected that the negative relationship between size and risk will be strengthened in high-UAI countries. This leads to the following hypothesis:
H5. The negative relationship between commercial bank size and commercial bank risk will be strengthened in countries that score high on uncertainty avoidance.
The fifth dimension is Confucian dynamism which is referred in the literature as long versus short term orientation (LTO). Countries who score high on long-term orientation have a tendency to accept delayed gratification of social needs and materials (Beugelsdijk et al., 2015). In countries who score high on short-term orientation the focus is on the profit of next quarter or next year, with freedom and achievement as major working values (Hofstede and Minkov, 2010). Therefore it is expected that countries with a long term orientation tend to take less risk. This leads to the following hypothesis:
14
Figure 1 presents the conceptual model of the hypothesized relationships.
Figure 1. Conceptual model
4. Data and research methods
Data on bank-level is downloaded from the Bankscope database. The sample comprises of panel data from 581 commercial banks in the European Union, including the United Kingdom, in the 2012-2015 period. In order to prevent for differences due to accounting standards, only banks who comply with IFRS are used. The European Union is selected because most banking research is US based (e.g. Boyd and Runkle, 1993; Demsetz and Straham, 1997; Stiroh and Rumble, 2006). Furthermore the post-crisis time period is selected because of the convergence in banking regulation described in Section 2.4. After filtering for missing observations 188 banks remained in the final sample. Appendix 1 contains a list with the names of all used banks. The sample is supplemented with data on industry- and country level, which is downloaded from the Global Financial Development Database of the World Bank. The cultural value dimensions of the EU Member States are retrieved from Hofstede’s website. There is no available cultural data from Cyprus and no available bank data from Lithuania so the final dataset contains 26 countries.
4.1 Dependent variable
In the extant literature bank risk has been measured in multiple ways. Examples are the volatility of equity returns, the volatility of earnings and the volatility of net interest margin
Commercial Bank risk Uncertainty avoidance
Long term orientation Individualism
Masculinity Power distance
Managerial decision making
15
(Laeven and Levine, 2009; Moutsianas and Kosmidou, 2016). According to Klomp and de Haan (2012) there is no general accepted definition of bank risk. Therefore they applied factor analysis on 25 indicators of banking risk. The indicators were divided in six groups: Capital adequacy, asset quality, managerial qualities, earnings and profitability, liquidity and market risk management. They concluded that the percentage of non-performing loans divided by total loans was one of the factors that captured most of the variance of the 25 indicators of bank risk. Because of constraints regarding the data availability, only this indicator is used in this paper to measure bank risk. Non-performing loans are defined as the loans overdue by more than ninety days (Louzis et al., 2011).
Other scholars have often used a Z-score as proxy for bank risk (Laeven and Levine, 2009; Kanagaretnam et al., 2014; Ashraf et al., 2016) . This is the return on assets plus the capital asset ratio divided by the standard deviation of asset returns. The Z-score indicates the distance from insolvency, a higher Z-score indicates higher bank stability. The Z-score is used in this paper as well to provide robustness checks on country-level.
4.2 Independent variable
Following the literature bank size is measured as the natural logarithm of total assets (de Haan and Poghosyan, 2012; Moutsianas and Kosmidou, 2016).
4.3 Moderating variables
There are multiple models developed to measure national culture (Schwartz, 1994). This paper employs Hofstede’s framework (1984) which is extensively used and validated (Li et al., 2013; Mihet, 2013; Ashraf et al., 2016). Based on employees survey data from IBM subsidiaries in the 1970s Hofstede identified four cultural dimensions on a scale from 0-100. In the following years Hofstede validated and extended his dataset to more countries and introduced a fifth dimension. Table 1 contains a list with the scores from all the countries in the sample.
Table 1
Hofstede’s cultural dimensions for countries in the sample
Country EU member since PDI IDV MAS UAI LTO
Austria 1995 11 55 79 70 60
Belgium 1958 65 75 54 94 82
Bulgaria 2007 70 30 40 85 69
16 Czech Republic 2004 57 58 57 74 70 Denmark 1973 18 74 16 23 35 Estonia 2004 40 60 30 60 82 Finland 1995 33 63 26 59 38 France 1958 68 71 43 86 63 Germany 1958 35 67 66 65 83 Greece 1981 60 35 57 100 45 Hungary 2004 46 80 88 82 58 Ireland 1973 28 70 68 35 24 Italy 1958 50 76 70 75 61 Latvia 2004 44 70 9 63 69 Lithuania 2004 42 60 19 65 82 Luxembourg 1958 40 60 50 70 64 Malta 2004 56 59 47 96 47 Netherlands 1958 38 80 14 53 67 Poland 2004 68 60 64 93 38 Portugal 1986 63 27 31 100 28 Romania 2007 90 30 42 90 52 Slovakia 2004 100 52 100 51 77 Slovenia 2004 71 27 19 88 49 Spain 1986 57 51 42 86 48 Sweden 1995 31 71 5 29 53 United Kingdom 1973 35 89 66 35 51
Note: PDI = Power distance IDV = Individualism MAS = Masculinity UAI = Uncertainty avoidance LTO = Long-term orientation.
Based on the argument of cultural theorists that culture tends to remain stable over time, it is assumed that the cultural scores do not change over the time-period from 2012-2015
(Hofstede, 2001). There are scholars who argue, based on the societal value change perspective, that cultural values could change over time due to economic development or globalization (Inglehart and Baker, 2000). However, recent research rejects this claim
(Beugelsdijk et al., 2015). These scholars conclude, by replicating Hofstede’s dimension, that there has not changed much regarding the relative positions of countries.
4.4 Control variables
Control variables are used on bank, industry and country-level. Growth influences risk-taking behavior on bank-level (Rahman et al., 2015; Ashraf et al., 2016). Growth is measured as the percentage of asset growth on a yearly basis. On industry-level is controlled for bank
17
GDP per capita because the level of economic development influences the level of risk at banks. High-income countries have better developed risk management techniques which leads to a lower amount of risk (Ashraf et al., 2016). All used variables, their definitions and data sources are summarized in table 2.
Table 2
Variable definitions and data sources
Variables Definition Data source
Dependent variable
Bank_risk Percentage of non-performing loans divided by total loans (yearly)
Bankscope
Independent variable
Bank_size Logarithm of total assets in USD (yearly) Bankscope Moderator
IDV Individualism index Hofstede's site
MAS Masculinity index Hofstede's site
PDI Power distance index Hofstede's site
UAI Uncertainty avoidance index Hofstede's site
LTO Long-term orientation index Hofstede's site
Control variables
Bank_growth Percentage of total asset growth (yearly) Bankscope Bank_concentration Percentage of assets of largest three
banks divided by all assets in one country (yearly).
World Bank
GDP_per_capita Logarithm of GDP per capita in USD (yearly)
World Bank Robustness check
Country Z-score (ROA+(equity/assets))/S.D. ROA) World Bank
4.5 Empirical model
In order to analyze the relationship between commercial bank size and commercial bank risk and the moderating effect of national culture, the following model is constructed:
𝑴𝒐𝒅𝒆𝒍(𝟏): 𝐵𝑎𝑛𝑘 𝑟𝑖𝑠𝑘𝑖,𝑗
= 𝛽0+ 𝛽1∗ 𝐵𝑎𝑛𝑘 𝑠𝑖𝑧𝑒𝑖,𝑗+ 𝛽2∗ 𝐵𝑎𝑛𝑘 𝑔𝑟𝑜𝑤𝑡ℎ𝑖,𝑗+ 𝛽3 ∗ 𝐵𝑎𝑛𝑘 𝑐𝑜𝑛𝑐𝑒𝑛𝑡𝑟𝑎𝑡𝑖𝑜𝑛𝑗+ 𝛽4∗ 𝐺𝐷𝑃 𝑝𝑒𝑟 𝑐𝑎𝑝𝑖𝑡𝑎𝑗 + 𝛽5∗ 𝑃𝐷𝐼𝑗
18
In the regression model i and j are the subscripts that refer to bank and country respectively. 𝛽0
is a constant, 𝛽1 to 𝛽9 are the coefficients and ε is the error term. Bank_risk is the dependent variable, measured by the percentage of non-performing loans over total loans. Bank_size is the independent variable measured by the log of total assets in US dollars. Bank_growth, Bank_concentration and GDP_per_capita are the control variables. The interaction effect of national culture is provided by the coefficients 𝛽5 to 𝛽9. Because the data is collected over a period of time and cross-sectional the pooled ordinary least squared (OLS) regression method is used.
5. Results
The descriptive statistics of all the variables, before logging, are provided in table 3.
Table 3
Summary statistics.
Variables Countries Observations Mean S.D. Min Max
Bank_risk 26 752 10.36 10.15 0.00 53.55 Bank_size (mln) 26 752 167,000 402,000 61 2,670,000 Bank_growth 26 752 5.15 17.90 -48.66 121.89 Bank_concentration 26 104 64.95 13.73 34.32 94.67 GDP_per_capita 26 104 35,195 17,161 7,062 107,036 IDV 26 26 65.37 17.45 27 89 MAS 26 26 53.19 22.29 5 100 PDI 26 26 50.96 19.26 11 100 UAI 26 26 68.14 22.30 23 100 LTO 26 26 55.90 12.42 24 83
Note: IDV: Individualism, MAS: Masculinity, PDI: Power distance, UAI: Uncertainty avoidance, LTO: Long-term orientation.
19
here as well. Luxembourg has the least concentrated commercial banking sector (34,32%) while the sector of Sweden is the most concentrated (94,67%). The average GDP per capita (in USD) is 35,195. Luxembourg has the highest income per capita with 107,036 while Bulgaria is on the other side of the spectrum with 7,062. Looking at the descriptive statistics of the cultural dimensions, the United Kingdom is the most individualistic country in the sample (89). Slovenia is the most collectivistic country with the lowest score (27). Slovakia scores the highest on masculinity (100) and power distance (100) where Sweden is the most feminine (5) and Austria the least hierarchical (11). Greece scores the highest on uncertainty avoidance (100) whereas Denmark is on the other side of the spectrum (23). Finally, Germany is the most long-term oriented (83) opposed to Ireland (24). To what extend the variables are correlating with each other is provided in table 4.
Table 4 Correlation matrix Variables 1 2 3 4 5 6 7 8 9 10 1 Bank_risk 1.00 2 Bank_size -0.19* 1.00 3 Bank_growth -0.16* -0.26* 1.00 4 Bank_concentration -0.04 0.03 -0.03 1.00 5 GDP_per_capita -0.32* 0.22* -0.08* 0.23* 1.00 6 IDV -0.42* 0.16* 0.02 0.09* 0.51* 1.00 7 MAS 0.03 -0.07* 0.04 -0.37* -0.24* 0.20* 1.00 8 PDI 0.23* -0.16* 0.04 -0.37* -0.69* -0.58* 0.26* 1.00 9 UAI 0.26* -0.06 0.07 -0.31* -0.53* -0.54* 0.20* 0.66* 1.00 10 LTO -0.13* -0.07 0.02 0.00 -0.11* 0.05 0.29* 0.32* 0.25* 1.00 Note: This table reports Pearson correlation coefficients between each pair of variables.
*Significance at 5% level. IDV: Individualism, MAS: Masculinity, PDI: Power distance, UAI: Uncertainty avoidance, LTO: Long-term orientation.
There is a significant negative correlation between bank_risk and bank_size, which is consistent with the modern portfolio and intermediation theory. The control variables
20
on the long-term have less risk. The coefficients of individualism, power distance and
uncertainty avoidance are all the opposite of what the theory suggests in Section 3 (Hofstede, 1984, 2001). There is a negative correlation between bank_size and bank_growth. This suggests that larger banks have a lower growth rate, in line with the law of diminishing returns (Shephard and Fare, 1974). Furthermore bank_size correlates positively with GDP_per_capita suggesting that large banks are in higher-income countries.
Bank_concentration correlates positive and significant with GDP_per_capita, suggesting that the level of competition between banks is lower in high-income countries. A possible
explanation for this could be that financial markets are more mature in high-income countries causing a consolidation phase in the industry. The cultural dimensions of Hofstede correlate quit a lot with each other. Besides the positive correlation of masculinity with power distance, uncertainty avoidance and long-term orientation, all the correlations are in line with the theory. The low number of observations (26) could be a possible explanation for the deviating results.
The results of the pooled OLS regressions are provided in table 5. The first model only includes the dependent, independent and control variables. According to the results of the first model there is a negative and significant relationship between commercial bank size and commercial bank risk. This is in line with modern portfolio and intermediation theory and provides support for the first hypothesis. It suggests that when banks become bigger they are better able to diversify their risks. This result is confirmed again in the second and third model. The second model includes also Hofstede’s cultural dimensions as additional explanatory variables which are all statistically significant at the minimum level of 5%. According to the results there is a negative relationship between individualism and bank risk, suggesting that banks in individualistic countries tend to take less risk. This is not in line with previous research. The same applies for the positive relationship between uncertainty
avoidance and bank risk. The other three cultural dimensions are in line with the existing literature (Ashraf et al., 2016). The third model includes also the interaction effects for the moderating effect of national culture. Individualism, masculinity and uncertainty avoidance are the only three moderators that are significant. Of these three, only the coefficient of uncertainty avoidance is in line with the expectations and provides support for the fifth hypothesis. Summarizing, the results of table 5 provide only support for the first and fifth hypothesis. There is a negative relation between commercial bank size and commercial bank risk and this relation is strengthened in countries who score high on uncertainty avoidance.
21
Table 5
Panel regression results with OLS: Bank risk, bank size and national culture.
Variables Bank_risk Bank_risk Bank_risk
(1) (2) (3) Bank_size -1.640*** -1.517*** -13.666*** (0.000) (0.000) (0.000) Bank_growth -0.130*** -0.118*** -0.109*** (0.000) (0.000) (0.000) Bank_concentration 0.020 0.041 0.094*** (0.422) (0.134) (0.002) GDP_per_capita 0.000*** 0.000*** 0.000*** (0.000) (0.000) (0.009) IDV -0.214*** -0.967*** (0.000) (0.000) MAS 0.065*** -0.209 (0.000) (0.201) PDI -0.080*** 0.084 (0.007) (0.764) UAI 0.046** -0.543*** (0.025) (0.006) LTO -0.129*** 0.197 (0.000) (0.456) Size*IDV 0.116*** (0.000) Size*MAS 0.041* (0.060) Size*PDI -0.024 (0.525) Size*UAI 0.089*** (0.002) Size*LTO -0.049 (0.164) Constant 27.824*** 41.278*** 116.165*** (0.000) (0.000) (0.000) Banks 188 188 188 Observations 752 752 752 R-squared 0.169 0.288 0.327 Adjusted R-squared 0.165 0.279 0.315
Time fixed effects No No No
Note: Dependent variable is Bank_risk in all models. All models are estimated using
22
5.1 Cross-section and time effects
Before testing for cross-section and time effects it is necessary to determine the appropriate model. In general, a random effects model is preferred over a fixed effects model because it is more efficient. However, a random effects model can only be used on the condition that the error term ε is unrelated with the explanatory variables. Otherwise the fourth Gauss-Markov assumption about the OLS method is violated. To check this a Haussmann test is performed (Hausman, 1978). The null hypothesis is that the error ε is uncorrelated with the explanatory variables. This hypothesis is rejected (P=0.000), so the random effects is not appropriate and the fixed effects should be preferred. When testing for time fixed effects, no significance is found. This implies that the relationship between commercial bank size and commercial bank risk does not change over the time period (2012-2015). Dummies are added for the 26 countries to test for cross-section fixed effects. Together all the dummies are jointly-significant using an F-test (P=0.000). Individually no significance is found. One major drawback of the fixed effects model is that any variable that does not vary over time will cancel out. So it is possible that there are differences cross-country but do not appear in the results because they are time-invariant.
5.2 Robustness tests
One of the limitations of the results provided in table 5 is that the number of banks across countries is not evenly distributed. This could have consequences for the results regarding the country-level variables because the cultural dimensions of Italy weigh much heavier compared to, for example, Portugal. This can be seen in table 6 below.
Table 6
Spreading of banks in sample
Country Number of banks Number as percentage
23 Ireland 4 2% Luxembourg 4 2% Romania 4 2% Bulgaria 3 2% Czech Republic 3 2% Germany 3 2% Slovenia 3 2% Sweden 3 2% Belgium 2 1% Estonia 2 1% Hungary 1 1% Latvia 1 1% Malta 1 1% Portugal 1 1% Total 188 100%
Note: Percentages ar rounded.
The country with the most banks in the sample is Italy, followed by the United Kingdom, Croatia, Denmark, France and Poland. These six countries have 119 (63%) banks in the sample. The other twenty countries have combined 69 (37%) banks in the sample. It stands out immediately that the number of banks across countries are far from evenly distributed. This could have consequences for the results and therefore should be addressed. To address this issue the regression is rerun with the dependent variable (bank_risk) averaged on country-level. This means a drop in observations from 752 to 104 but fixes the issue of the skewed distribution of countries in the sample. Because the issue does not affect the bank-level variables, only the second and third model of the regression of table 5 are rerun. In order to strengthen these results, due to the relatively low number of observations, additional testing is done with the country Z-score as an alternative commercial bank risk measure. The country’s Z-score captures the probability of default of a country’s commercial banking system. The Z-score is calculated in the 1998-2014 period and is retrieved from the World Bank Financial Development database. As mentioned earlier a higher Z-score implies higher bank stability, so this should give the opposite results. To address this issue the Z-score is multiplied with -1 and subsequently logged because it is highly skewed. The results are provided in table 7.
24
Table 7
Panel regression results with OLS: Country average scores
Variables Bank_risk Bank_risk Bank Z-score Bank Z-score
1 2 3 4 Bank_size -2.043* -8.034 -0.147* -0.572 (0.100) (0.350) (0.067) (0.301) Bank_growth -0.159* -0.078 -0.021*** -0.017*** (0.096) (0.415) (0.001) (0.007) Bank_concentration -0.069 0.012 0.001 0.006* (0.118) (0.815) (0.758) (0.070) GDP_per_capita -4.260*** -1.254 -0.768*** -0.672*** (0.007) (0.543) (0.000) (0.000) IDV -0.145** -0.540 0.010*** 0.022 (0.014) (0.365) (0.007) (0.566) MAS 0.054* -1.292** -0.005*** -0.084** (0.065) (0.029) (0.007) (0.027) PDI -0.073 0.715 -0.001 -0.093 (0.114) (0.479) (0.737) (0.155) UAI 0.038 -0.330 -0.001 0.079* (0.360) (0.630) (0.643) (0.075) LTO -0.130*** 0.541 -0.011*** -0.031 (0.003) (0.401) (0.000) (0.454) Size*IDV 0.067 -0.002 (0.405) (0.634) Size*MAS 0.183** 0.011** (0.023) (0.027) Size*PDI -0.093 0.014 (0.501) (0.120) Size*UAI 0.054 -0.012* (0.580) (0.063) Size*LTO -0.097 0.003 (0.267) (0.610) Constant 88.228*** 83.468 7.102*** 8.918* (0.000) (0.294) (0.000) (0.084) Countries 26 26 26 26 Observations 104 104 104 104 R-squared 0.485 0.542 0.544 0.598 Adjusted R-squared 0.436 0.470 0.500 0.535
Note: Dependent variable is Bank_risk and Bank Z-score. All models are estimated using
25
The results of the first and second model in table 7 reflect the country averages of the bank-level variables: Bank_risk, bank_size and bank_growth. The other control variables
bank_concentration and GDP_per_capita were already on country-level, just as Hofstede’s cultural dimensions. Comparing the first two models of table 7 with the second and third model of table 5, the coefficients are largely in line, but some significance-levels differ. When looking at the first two models of table 7, only the first model provides support for the first hypothesis, showing a negative relationship between commercial bank size and commercial bank risk that is significant. The only significant moderator is masculinity, but the coefficient is not in line with the expectations, providing no support for any hypothesis. The coefficient of uncertainty avoidance is just as in table 5 positive, but this time not significant. The first two models in table 7 predict the level of commercial bank risk, with 44% and 47%
respectively when looking at the adjusted R-square. This is a bit higher than the 28% and 32% of the second and third model of table 5.
The results of the third and fourth model in table 7 reflect the country’s Z-score, the alternative bank risk measure. Only the third model shows a negative relationship between commercial bank size and commercial bank risk that is significant, providing support for the first hypothesis. Regarding the moderating effect of national culture only masculinity and uncertainty avoidance are significant, but their coefficients are not in line with the
expectations so this does not provide evidence for the hypotheses. When comparing all the models of table 7, the fourth model is the best predictor of commercial bank risk, with an adjusted R-squared of 54%.
26
Table 8
Comparing most important results of table 7 and table 5
Variables Bank_risk (individual banks) Bank_risk (country average) Bank Z-score (country average) (1) (2) (3) Bank_size -13.666*** -8.034 -0.572 (0.000) (0.350) (0.301) Bank_growth -0.109*** -0.078 -0.017*** (0.000) (0.415) (0.007) Bank_concentration 0.094*** 0.012 0.006* (0.002) (0.815) (0.070) GDP_per_capita 0.000*** -1.254 -0.672*** (0.009) (0.543) (0.000) IDV -0.967*** -0.540 0.022 (0.000) (0.365) (0.566) MAS -0.209 -1.292** -0.084** (0.201) (0.029) (0.027) PDI 0.084 0.715 -0.093 (0.764) (0.479) (0.155) UAI -0.543*** -0.330 0.079* (0.006) (0.630) (0.075) LTO 0.197 0.541 -0.031 (0.456) (0.401) (0.454) Size*IDV 0.116*** 0.067 -0.002 (0.000) (0.405) (0.634) Size*MAS 0.041* 0.183** 0.011** (0.060) (0.023) (0.027) Size*PDI -0.024 -0.093 0.014 (0.525) (0.501) (0.120) Size*UAI 0.089*** 0.054 -0.012* (0.002) (0.580) (0.063) Size*LTO -0.049 -0.097 0.003 (0.164) (0.267) (0.610) Constant 116.165*** 83.468 8.918* (0.000) (0.294) (0.084) Countries 26 26 26 Observations 752 104 104 R-squared 0.327 0.542 0.598 Adjusted R-squared 0.315 0.470 0.535
Note: All models are estimated using pooled OLS regressions. P-values are computed
27
masculinity. Because all three models provide evidence for this claim, it is worthwhile to investigate in future research why this is the case. The fourth hypothesis stated a positive interaction effect of power distance. In none of the models a significant interaction effect is found for power distance. The fifth hypothesis stated a positive interaction effect of
uncertainty avoidance. This is the case in the first model, but in the third model the opposite is found. So in the end the results are mixed regarding this claim. The sixth and last hypothesis stated a positive interaction effect of long-term orientation. Just as was the case with power distance, no significant interaction effect was found at any of the three models.
Summarizing, only partial evidence is found that supports the first hypothesis, implying a negative relationship between commercial bank size and commercial bank risk. Regarding the moderating effect of national culture, the results are contradicting the literature or
non-existent at all.
6. Conclusions, limitations and suggestions for further research
28
This study contributes to the literature by taking social theories into account in the relationship between commercial bank size and commercial bank risk. As regards to the practical aspects, this study has implications for bank managers, investors, policy makers and regulators in the financial world. Commercial banks need to increase the size of their business if they want to optimize their risk management. Furthermore managers of international banks need to take national culture into account when designing and executing their strategy because the strategy is usually set at the headquarters. Investors need to take the results into account in order to optimize the risk-return tradeoff of their portfolios. In practice this means assigning a higher risk premium to smaller banks when constructing their portfolios. After the global financial crisis, policy makers and regulators have talked a lot about banks that are ‘too big to fail’. The results of this paper suggest that banks need size if they want to manage their risks properly. Hence, this paper provides more support for the counter argument that small banks do not have enough scale and diversification opportunities, or in other words, are ‘too small to survive’.
29
Appendix 1. List of used banks in the sample.
Bank Name Banca Simetica S.P.A.
Abanca Corporacion Bancaria SA Banco Bilbao Vizcaya Argentaria SA-BBVA
Abanka d.d Banco Comercial Português, SA-Millennium bcp
Abbey National Treasury Services Plc Banco de Sabadell SA
ABLV Bank AS Banco di Credito P. Azzoaglio SpA
ABN AMRO Group N.V. Banco di Desio e della Brianza SpA-Banco Desio
AEGON Bank NV Banco di Lucca E Del Tirreno SpA
Aktia Bank Plc Banco di Napoli SpA
Alandsbanken Abp-Bank of Aland Plc Banco di Sardegna SpA
Alior Bank Spólka Akcyjna Banco Popular Espanol SA
Allianz Bank Financial Advisors S.p.A. Banco Santander SA
Allied Irish Banks plc Bank BPH SA
Alpha Bank AE Bank für Tirol und Vorarlberg AG-BTV (3 Banken
Gruppe)
Arbejdernes Landsbank A/S Bank Handlowy w Warszawie S.A.
AS SEB Pank Bank Leumi (UK) Plc
Attica Bank SA-Bank of Attica SA Bank Millennium
Axa Banque Bank Ochrony Srodowiska SA - BOS SA-Bank
Ochrony Srodowiska Capital Group
Banca 5 SpA Bank of China (UK) Ltd
Banca Aletti & C. Spa Bank of Ireland-Governor and Company of the Bank
of Ireland
Banca Cambiano Bank of Scotland Plc
Banca Capasso Antonio SpA Bank of Valletta Plc
Banca Carige SpA Bank Polska Kasa Opieki SA-Bank Pekao SA
Banca del Fucino SpA Bank Zachodni WBK S.A.
Banca del Monte di Lucca SpA Banka Kovanica dd Varazdin
Banca del Piemonte Bankia, SA
Banca del Sud SPA Bankinter SA
Banca di Imola SpA BankNordik P/F
Banca di Sassari SpA Banque Degroof Petercam Luxembourg SA
Banca di Sconto e Conti Correnti di Santa Maria Capua Vetere SpA
Banque Fédérative du Crédit Mutuel
Banca Federico del Vecchio SpA Banque Internationale à Luxembourg SA
Banca Finnat Euramerica SpA Barclays Bank Plc
Banca Galileo SpA Belfius Banque SA/NV-Belfius Bank SA/NV
Banca Generali SpA-Generbanca BGL BNP Paribas
Banca Ifis SpA BinckBank NV
Banca Interprovinciale Societa per Azioni BNP Paribas
Banca Mediolanum SpA BNP Paribas Fortis SA/ NV
Banca Monte dei Paschi di Siena SpA-Gruppo Monte dei Paschi di Siena
Boursorama SA
Banca Nazionale del Lavoro SpA-BNL Bradford & Bingley Plc
Banca Nuova SpA BRD-Groupe Societe Generale SA
Banca Passadore & C. SpA British Arab Commercial Bank Plc
Banca Popolare di Milano Spa Bulgarian-American Credit Bank
Banca Profilo SpA Banca Prossima Spa
Banca Promos SpA Banca Santa Giulia SpA
Banca Sella SpA Länsförsäkringar Bank AB (Publ)
Caixabank, S.A. Le Crédit Lyonnais (LCL) SA
Cassa Centrale Banca Credito Cooperativo del Nord Est SpA
Lloyds Bank Plc
Cassa di Risparmio di Cesena SpA Lollands Bank A/S
Ceska Sporitelna a.s. Lyonnaise de Banque SA
Citibank Europe Plc mBank SA
30
Credit Agricole Bank Romania SA National Bank of Greece SA
Credit Agricole Corporate and Investment Bank SA-Credit Agricole CIB
National Westminster Bank Plc - NatWest
Crédit Agricole Friuladria spa Natixis SA
Credit Europe Bank N.V. NIBC Bank NV
Crédit Industriel et Commercial SA - CIC Nordfyns Bank A/S
Credito Emiliano SpA-CREDEM Nordjyske Bank A/S
Danske Andelskassers Bank A/S Nova Kreditna Banka Maribor d.d.
Danske Bank A/S Nykredit Bank A/S
Danske Bank Plc Oberbank AG
De Volksbank N.V. OP Corporate Bank plc
Demir-Halk Bank (Nederland) N.V-DHB Bank OTP Bank Plc
Deutsche Bank AG OTP banka Hrvatska dd
Deutsche Bank Mutui SpA OTP Banka Slovensko, as
Deutsche Bank SpA Partner Banka dd
Erste & Steiermärkische Bank dd Patria Bank S.A.
Eurobank Ergasias SA Permanent TSB Plc
Europe Arab Bank Plc Piraeus Bank SA
Extrabanca SPA Podravska Banka
Farbanca SpA Prima banka Slovensko a.s.
Findomestic Banca SpA Privredna Banka Zagreb d.d-Privredna Banka Zagreb
Group
FinecoBank Banca FinEco SpA-Banca FinEco SpA Raiffeisen Bank International AG
First Investment Bank AD RCI Banque SA
Fynske Bank A/S Royal Bank of Scotland Plc (The)
Getin Noble Bank SA Santander UK Plc
Hrvatska Postanska Bank DD Sberbank Slovensko, as
HSBC Bank plc Secure Trust Bank Plc
HSBC France SA Skandinaviska Enskilda Banken AB
Hypotecni banka a.s. SKB Banka DD
IBL Istituto Bancario del Lavoro SpA Societe Generale - Splitska Banka dd
Imprebanca SpA Société Générale SA
ING Bank NV Spar Nord Bank
ING Bank Slaski S.A. - Capital Group Standard Chartered Bank
Intesa Sanpaolo Laan & Spar Bank A/S
Istarska Kreditna Bank Umag d.d. Suedtirol Bank AG - Alto Adige Banca SP A.
J&T banka d.d. Sumitomo Mitsui Banking Corporation Europe
Limited-SMBCE
Jyske Bank A/S (Group) Svenska Handelsbanken AB
Karlovacka banka d.d. Swedbank As
KBL European Private Bankers SA Sydbank A/S
Komercni Banka Kreditna Banka Zagreb
Tatra Banka a.s.
Teximbank-PEB Texim AD
Transilvania Bank-Banca Transilvania SA UniCredit Bank AG
UniCredit Bank Austria AG-Bank Austria UniCredit SpA
Unipol Banca Spa Veneto Banka d.d.
Vseobecna Uverova Banka a.s. VTB Capital Plc
31
References
Acharya, V. V., Amihud, Y., Litov, L., 2011. Creditor rights and corporate risk-taking. Journal of Financial Economics 102, 150–166.
Aduralere, I. I., Olufemi, A. M., 2016. Bank reforms and banking sector performance: Evidence from Nigeria. International Journal of Economics and Research, 7(1), 36-55.
Afzal, A., Mirza, N., 2012. Size, Diversification and Risk: Preliminary Evidence from Commercial Banks in Pakistan. Pakistan Journal of Commerce and Social Science 6(2), 282- 296.
Aggarwal, R., Goodell, J. W., 2014. National cultural dimensions in finance and accounting scholarship: an important gap in the literatures? Journal of Behavioral and Experimental Finance 1, 1–12.
Angkinand, A. P., Sawangngoenyuang, W., Wihlborg, C., 2010. Financial liberalization and banking crises: a cross-country analysis. International Review of Finance 10, 263–292.
Ashraf, B. N., 2017. Political institutions and bank risk-taking behavior. Journal of Financial Stability 29, 13-35.
Ashraf, B. N., Arshad, N., 2017. Foreign bank subsidiaries’ risk-taking behavior: Impact of home and host country national culture. Research in international business and finance 41, 318-335.
Ashraf, B. N., Zheng, C., Arshad, S., 2016. Effects of national culture on bank risk-taking behavior. Research in International Business and Finance, 37, 309-326.
Baker, D., McArthur, T., 2009. The value of the ‘‘too big to fail’’ big bank subsidy. Center for Economic and Policy Research Paper, Washington, D. C.
32
Berger, A. N., Bouwman, C. H. S., 2013. How does capital affect bank performance during financial crisis? Journal of Financial Economics 109, 146–176.
Berger, A. N., DeYoung, R., 1997. Problem loans and cost efficiency in commercial banks. Journal of Banking and Finance, 21, 849–870.
Berger, A. N., El Ghoul, S., Guedhami, O., Roman, R. A., 2016. Internationalization and bank risk. Research working paper, The Federal Reserve Bank of Kansas City.
Beugelsdijk, S., Maseland, R., van Hoorn, A., 2015. Are scores on Hofstede's dimensions of national culture stable over time? A cohort analysis. Global Strategy Journal 5, 223–240.
Bhagat, S., Bolton, B., Lu, J., 2015. Size, leverage, and risk-taking of financial institutions. Journal of Banking and Finance 59, 520-537.
Boss, M., Fenz, G., Pann, J., Puhr, C., Schneider, M., Ubl, E. 2009. Modeling credit risk through the Austrian business cycle: an update of the OeNB Model. OeNB Financial Stability Report, 17, 85–101.
Boubakri, N., Mirzaei, A., Samet, A., 2017. National culture and bank performance: evidence from the recent financial crisis. Journal of Financial Stability 29, 36–56.
Boyd, J. H., Runkle, D. E., 1993. Size and performance of banking firms: testing the predictions of theory. Journal of Monetary Economics 31, 47–67.
Beltratti, A., Stulz, R. M., 2012. The credit crisis around the globe: Why did some banks perform better? Journal of Financial Economics 105, 1-17.
Brewer, E. I., Jagtiani, J., 2013. How much did banks pay to become too-big-to-fail and to become systemically important? Journal of Financial Services Research 43 (1), 1–35.
33
Brimmer, A. F., Dahl, F. R., 1975. Growth of American international banking: Implications for public policy. The Journal of Finance 30, 341–363.
Chaibi, H., Ftiti, Z., 2015. Credit risk determinants: evidence from a cross-country study. Research in International Business and Finance 33, 1–16.
Chang, T., Chen, C. C., 2017. The Impact of Bank Size on Profit Stability in China. Journal of Applied Finance and Banking 7(2), 59-70.
Chari, A., Gupta, N., 2008. Incumbents and protectionism: The political economy of foreign entry liberalization. Journal of Financial Economics 88, 633–656.
Claessens, S., Djankov, S., Nenova, T., 2000. Corporate risk around the world. World Bank Policy Research Working Paper, Washington D. C.
Dam, L., Koetter, M., 2012. Bank Bailouts and Moral Hazard: Evidence from Germany. The Review of Financial Studies 25(8), 2343 – 2380.
De Haan, J., Poghosyan, T. 2012. Size and earnings volatility of US bank holding companies. Journal of Banking and Finance 36, 3008–3016.
De Haan, J., Poghosyan, T., 2012. Bank size, market concentration, and bank earnings volatility in the US. Journal of International Financial Markets, Institutions and Money 22, 35–54.
De Nicolo, G., 2000. Size, charter value and risk in banking: an international perspective. International Finance Discussion Paper 689, Board of Governors of the Federal Reserve System.
Demirgüç-Kunt, A., Detragiache, E., 2002. Does deposit insurance increase banking system stability? An empirical investigation. Journal of Monetary Economics 49(7), 1373– 1406.
34
Doidge, C., Karolyi, G. A., Stulz, R. M., 2007. Why do countries matter so much for corporate governance? Journal of Financial Economics 86, 1-39.
Financial Stability Board, 2010. Reducing the Moral Hazard Posed by Systemically Important Financial Institutions: Interim Report to G20 Leaders. Financial Stability Board, Basel.
Fu, X. M., Lin, Y.R., Molyneux, P. 2014. Bank competition and financial stability in Asia Pacific. Journal of Banking and Finance 38, 64–77.
Giuliano, I., 2010. Investment Banking: A Guide to Underwriting and Advisory Services. Springer, New York.
Gonzalez, F., 2004. Bank regulation and risk-taking incentives: An international comparison of bank risk. Journal of Banking and Finance 29, 1153-1184.
Graham, J. R., Harvey, C. R., Puri, M., 2010. Managerial Attitudes and Corporate Actions. Journal of Financial Economics 109, 103-121.
Guiso, L., Sapienza, P., Zingales, L., 2006. Does Culture Affect Economic Outcomes? Journal of Economic Perspectives 20(2), 23-48.
Haberman, G., 1987. Capital Requirements of Commercial and Investment Banks: Contrasts in Regulation. Quarterly Review, Federal Reserve Bank of New York.
Hausman, J. A., 1978. Specification Tests in Econometrics. Econometrica 46(6), 1251-1271.
Hefferman, S., 2005. Modern Banking. John Wiley & Sons, England.
Hofstede, G., 1984. Culture's consequences: international differences in work-related values. CA: Sage Publications, Newbury Park,
Hofstede, G., 2001. Culture's consequences. 2nd ed. Sage Publications, Thousand Oaks.
35
Hope, O. K., 2003. Firm-Level Disclosures and the Relative Roles of Culture and Legal Origin. Journal of International Financial Management and Accounting 14, 218-48.
Hoque, H., Andriosopoulos, D., Andriosopoulos, K., Douady, R., 2015. Bank regulation, risk and return: evidence from the credit and sovereign debt crises. Journal of Banking and Finance 50, 455–474.
Inglehart, R., Baker, W. E., 2000. Modernization, cultural change, and the persistence of traditional values. American Sociological Review, 19–51.
Jensen, M., Meckling, W., 1976. Theory of the firm: managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3, 305-360.
John, K., Litov, L., Yeung, B., 2008. Corporate governance and risk-taking. Journal of Finance 63, 1679–1728.
Jokipii, T., Milne, A., 2011. Bank capital buffer and risk adjustment decisions. Journal of Financial Stability 7, 165-178.
Kahneman, D., 2011. Thinking, fast and slow. Macmillan, London.
Kanagaretnam, K., Chee Yeow, L., Lobo, G. J., 2014. Influence of national culture on accounting conservatism and risk-taking in the banking industry. Accounting Review 89, 1115– 1149.
Kanagaretnam, K., Lim, C. Y., Lobo, G., 2011. Effects of national culture on earnings quality of banks. Journal of International Business Studies, 42(6), 853-874.
36
Kasman, A., Kasman, S., 2016. Bank size, competition and risk in the Turkish banking industry. Empirica 43(3), 607-631.
Kaufman, G. G., 2014. Too Big to Fail in Banking: What Does it Mean? Journal of Financial Stability 13, 214-223.
Klomp, J., De Haan, J., 2012. Banking risk and regulation: Does one size fit all? Journal of Banking and Finance 36, 3197-3212.
Kreiser, P., Marino, L., Dickson, P., Weaver, K., 2010. Cultural Influences on Entrepreneurial Orientation: The Impact of National Culture on Risk Taking and Proactiveness in SMEs. Entrepreneurship Theory and Practice, 959-984.
Kwok, C. C. Y., Tadesse, S., 2006. National Culture and Financial Systems. Journal of International Business Studies 37, 227-47.
Laeven, L., Levine, R., 2007. Is there a diversification discount in financial conglomerates? Journal of Financial Economics 85, 331–367.
Laeven, L., Levine, R., 2009. Bank governance, regulation and risk taking. Journal of Financial Economics 93, 259–275.
Lee, S. M., Peterson, S. J., 2000. Culture, Entrepreneurial Orientation, and Global Competitiveness. Journal of World Business 35(4), 401-16.
Li, J., Guisinger, S., 1992. The globalization of service multinationals in the “triad” regions: Japan, Western Europe and North America. Journal of International Business Studies 23, 675– 696.
Li, K., Griffin, D., Yue, H., Zhao, L., 2013. How does culture influence corporate risk-taking? Journal of Corporate Finance, 23, 1-22.
37
Luo, Y., Tanna, S., De Vita, G., 2016. Financial openness, risk and bank efficiency: Cross-country evidence. Journal of Financial Stability 24, 132-148.
Markowitz, H., 1952. Portfolio Selection. Journal of Finance 7(1), 77–91.
Martinez-Miera, D., Repullo, R., 2010. Does competition reduce the risk of bank failure? Review of Financial Studies 23, 3638–3664.
Meier-Pesti, K., Penz, E., 2008. Sex or gender? Expanding the sex-based view by introducing masculinity and femininity as predictors of financial risk taking. Journal of Economic Psychology, 29(2), 180-196.
Merton, R. C., 1977. An analytical derivation of the cost of deposit insurance and loan guarantees. Journal of Banking and Finance 1(1), 3-11.
Mihet, R., 2013. Effects of culture on firm risk-taking: a cross-country and cross-industry analysis. Journal of Cultural Economics, 37(1), 109-151.
Morgan, D. P., Stiroh, K. J., 2005. Too Big to Fail after All These Years. Working paper, Federal Reserve Bank of New York.
Moutsianas, K. A., Kosmidou, K. 2016. Bank earnings volatility in the UK: Does size matter? A comparison between commercial and investment banks. Research in International Business and Finance 38, 137-150.
Paligorova, T., Santos, J. A. C., 2013. Monetary policy and bank risk-taking: Evidence from the corporate loan market. Journal of Financial Intermediation, Forthcoming.
Radenbaugh, L. H., Gray, S. J., Black, E. L., 2006. International Accounting and Multinational Enterprises 6th edition. John Wiley and Sons, Hoboken.
