Master Thesis
MSc International Economics & Business
University of Groningen
Faculty of Economics & Business
An empirical study of foreign bank presence on
financial stability using bank- and country-level data
January 2017
Author:
Supervisor:
Vivian Marit van Breemen
(s2670127)Prof. dr. J. de Haan
vivianmarit@gmail.com
jakob.de.haan@rug.nl
Co-supervisor:
Acknowledgements
Contents
1. Introduction ... 2
2. Literature Review ... 4
2.1 Foreign bank entry ... 4
2.2 Financial stability ... 5
2.3 Heterogeneity of (foreign) banks ... 7
2.4 Heterogeneity of host- and home- market ... 9
2.5 Conceptual Model ... 10
3. Data & Method ... 11
3.1 Data ... 11 3.2 Method ... 15 3.2.1 Bank-level Analysis ... 15 3.2.2 Country-level Analysis ... 16 4. Empirical Result ... 17 4.1 Bank-level Analysis ... 17 4.1.1 Robustness Check ... 21 4.2 Country-level Analysis ... 24
5. Discussion & Limitation ... 27
6. Conclusion ... 29
References ... 30
Databases ... 33
Appendices ... 34
I. Data ... 34
III. Correlation matrix ... 39
IV. Heteroscedasticity ... 40
V. Hausman test ... 41
VI. Income group ... 42
List of Tables
TABLE 1. Foreign bank share ... 12TABLE 2. Ownership, income group and region ... 12
TABLE 3. The effect of foreign bank presence on bank-level financial stability by controlling for bank and country level characteristics ... 18
TABLE 4. The effect of foreign banks with above/below median size and profitability on bank-level financial stability ... 20
TABLE 5. The effect of foreign bank presence on bank-level financial stability, introducing a lag term, ROE and interaction variable... 23
TABLE 6. The effect of foreign banks presence on country-level financial stability ... 25
TABLE 7. The effect of foreign bank presence on country-level financial stability, introducing host country characteristics ... 26
TABLE 8. The effect of foreign bank presence on bank-level financial stability, introducing host country characteristics ... 42
1
An empirical study of foreign bank presence on
financial stability using bank- and country-level data
Vivian Marit van Breemen
University of GroningenABSTRACT
This paper examines the impact of foreign bank presence on bank- and country-level financial stability via a panel data study, using data for more than 400 banks from 40 countries for the period 1999-2013. The empirical results show (i) a significant negative effect of foreign bank presence on bank-level as well as country-level financial stability (ii) foreign banks with high profitability have a positive effect on bank-level financial stability and foreign banks with low profitability negatively impact level financial stability (iii) small foreign banks negatively impact bank-level financial stability while large foreign banks have a positive influence.
JEL classification: F23, G21, G32
2
1. Introduction
In the last decade foreign bank presence increased substantially (Lehner & Schnitzer, 2008) especially toward developing countries (Beck, 2008). Besides, a fast consolidation of banks within and across countries took place. As a consequence, financial conglomerates provide commercial and investment banking services as well as insurance and pensions. This, however, increased concerns among policymakers about the financial stability and location of banks. Some institutions might be too-big-to-fail and as a result take excessive risk (Beck, 2008). In fact, the internationalization of banking liberalization is closely related to the global financial crisis in 2007-2009 (Beck, De Jonghe & Schepens, 2012). Consequently, factors affecting bank stability are of major interest to bank supervisors and regulators in order to ensure stability in the financial system (Adusei, 2015).
3
Moreover, to my knowledge this is the first paper that analyses the impact of foreign bank presence on host country-level financial stability. This might be of great importance to policymakers to ensure financial system stability (Demirgüç-Kunt et al., 1998). Besides, it contributes to existing literature in the fields of foreign bank entry, financial stability and bank heterogeneity. Furthermore, this paper studies foreign bank presence on bank-level financial stability by using a rich data set over fourteen years with most recent available data on foreign bank ownership of Claessens & Van Horen (2015). Financial stability is measured using the Z-score, a proxy for bank’s probability of failure (Lee & Hsieh, 2014; Beck et al., 2012; Leaven & Levine, 2009), nonperforming loans and the capital ratio (equity to total assets).
The findings of the paper are as follows. First, foreign bank presence has a significantly negative effect on bank-level as well as country-level financial stability. Second, foreign banks with below median profitability decreases bank-level financial stability, while foreign banks with above median profitability increase bank-level financial stability. Third, small foreign banks negatively influence financial stability while large foreign banks positively influence bank-level financial stability.
4
2. Literature Review
The first section provides previous literature on the impact of foreign bank entry, while section 2.2 describes previous studies on financial stability. Section 2.3 states the literature on the effect of foreign bank heterogeneity and section 2.4 explains the heterogeneity of the host- and home-country characteristics. Finally, section 2.5 provides a conceptual framework.
2.1 Foreign bank entry
Foreign bank entry creates higher competition in the banking sector of the host country, which generates positive welfare effects (Lehner & Schnitzer, 2008) and financial stability since it stimulates innovation and efficiency improvements (Akins, Li, NG, Rosticus, 2016). Foreign banks are attracted by host country’s government via liberalization of the banking markets since they stimulate foreign bank entry to improve the country’s banking system, caused via spill over effects from foreign banks or increased competition (Lehner & Schnitzer, 2008). Lehner & Schnitzer (2008) study the impact of foreign bank entry on the host country by taking into account the differences in screening abilities of banks. Domestic banks in a closed economy have low screening capabilities; foreign bank entry might increase their screening ability due to positive spill over effects from foreign banks towards domestic banks (Claessens et al., 2001), since foreign banks: encourage technological improvements and modern banking techniques; enlarge international capital availability; strengthen the quality and availability of financial services; and improves the legal and supervisory framework(Levine, 1997). Barajas, Steiner & Salazar (2000) also found a positive impact of foreign bank entry on domestic bank operations due to increased competition, lower intermediation costs and improved loan quality. On the other hand, the entrance of foreign banks causes a reduction in profitability and margins for domestic banks, which may reduce their stability (Claessens et al., 2001). Likewise, Stiglitz (1990) found that foreign bank entry results in possible difficulties for domestic banks, local entrepreneurs and the government. Foreign bank entry creates higher competition in the host country and as a result domestic banks face higher costs to compete with the (often large) foreign banks. Besides, small local companies face problems regarding financial access as foreign banks mainly focus on larger firms.
5
liberalization of East Asian countries. Besides, it shows that the financial globalization increased rapidly before the global financial crisis (Claessens & Van Horen, 2014) caused by communication and technology development, deregulations and economic integrations (Claessens & Van Horen, 2012). As a consequence of the global financial crisis, foreign bank entry decreased since banks featured capital losses and yet countries that were influenced by the financial crisis are not as interesting for investments (Claessens & Van Horen, 2014).
2.2 Financial stability
Systemic risk has gained considerable amount of attention from academia and policy makers, particularly after the US subprime mortgage crises (Gang & Qian, 2015). Systemic risk is the connection in bank portfolio values and credit interlinkages, causing one bank’s bankruptcy to affect other banks in the system, resulting in enormous economic and social costs (Elsinger, Lehar & Summer, 2006). Hence, systemic risk is the risk of a financial crisis and its spill over to the whole economy (Acharya, Pedersen, Philippon & Richardson, 2010) and is identified by an initial shock, propagation and magnification mechanism as well as disruption of the financial system (Bijlsma, Klomp & Duineveld, n.d.). Causes of systemic risk include (i) correlated exposures that can result in several bank defaults at the same time in case of shocks and (ii) banks facing distress might use interbank liabilities which causes insolvency for other banks as well. Hence, insolvency of one bank that use interbank liabilities of other banks might stimulate contagious defaults, the so called domino effect. The default of banks may be either due to fundamental default, caused by market and credit risk loss, or due to contagious default, caused by contagion effects (Elsinger et al., 2006). Contagion is defined as the transmission of individual shocks that affects one or several banks which will propagate to more banks or to the economic sector. Contagion or similar shocks affecting banks at the same time are causes of systemic risk (Hasman, 2013). A joint distress in the financial system causes large externalities and is very expensive since it is harder to acquire capital for financial firms. As a result, funding liquidity problems might arise due to tension in the interbank market. Furthermore, externalities increase with bank size since large banks frequently have important roles in interbank lending (Laeven, Ratnovski & Tong, 2014).
6
advantage hypothesis which predicts that domestic banks have advantages over foreign banks. The advantages of domestic banks are caused by increased costs for foreign banks for offering similar financial services or by lower revenues of foreign banks. In contrast, the global advantage hypothesis predicts competitive advantages of foreign bank compared to domestic banks due to the usage of better technologies caused by higher competition in the foreign bank home country. Angkinand & Wihlborg (2010) found results in line with the home field advantage hypothesis, hence foreign bank entry decreases financial stability; they also used the Z-index as a proxy for stability. Demirgüç-Kunt et al. (1998) state that foreign banks might improve efficiency, innovation, supervision and regulation which in turn stimulate economic growth and reduce systemic risk. On the contrary, foreign banks might decrease host country’ stability via excessive borrowing and easing of international capital flows. Moreover, they show, using a Korean case study, that foreign bank entry improves the host country’s banking system efficiency and spur growth due to lower overhead costs. Furthermore, they found that foreign banks lower the probability of a financial crisis in the host country. Besides, foreign bank entry increases competition which creates better bank efficiency but might as well destabilize the banking sector (Demirgüç-Kunt & Detragiache, 2005).
To summarize, mixed empirical evidence exists when it comes to the effect of foreign bank entry. On the one hand, it might increase financial stability due to the stimulation of innovation and efficiency but, on the other hand, increased market power might cause increased risk-taking of banks that might be too-big-to-fail (Akins et al., 2016). Henceforth, based on the results found in previous studies (Lee & Hsieh, 2014; Angkinand and Wihlborg, 2010) the following hypothesis is tested:
H1) Foreign bank presence is negatively related to bank-level financial stability in the host
country.
7
H2) Foreign bank presence is negatively related to country-level financial stability in the host
country.
2.3 Heterogeneity of (foreign) banks
Claessens & Van Horen (2015) stressed the importance of further research including whether it matters how much and what variation there is among foreign banks active in a particular country. They state that it is important to differentiate bank types by country of origin, size, degree of international operations and profitability on financial sector development and risks. In line with Van Oordt & Zhou (2014), this paper studies the effect of size and profitability of foreign banks while controlling for bank funding structures.
The size of banks is an important aspect for bank supervision and regulation and has gained attention during the global financial crisis in 2007-2009. Some authors pose that large banks accounted for the financial crisis (Adusei, 2015). Size and scope limits for banks is one way of reducing the risk of banks that are too-big-to-fail (Viñals, Surti, Narain, Erbenova & Chow, 2013). Two theories explain the relationship between bank size and bank stability where (1) the agency theory (Jensen & Meckling, 1976) states that personal benefits are key in managers’ decisions and actions where larger banks provide larger compensation and thus larger bank size is a result of managers own interest in acquiring higher personal returns, representing poor governance, while (2) the stewardship theory states a positive association between bank size and stability since it argues that managers are reliable and act accordingly (Donaldson & Davis, 1991). The former theory predicts a negative relationship between bank size and bank-level financial stability while the latter theory suggests a positive relationship between bank size and bank-level financial stability.
8
assets in excess of US$10-50 billion), but emphasize that insufficient amount of capital of large banks is also a cause of individual bank risk. Furthermore, externalities tend to be larger for large banks, since large banks have more important roles in international lending as they contribute higher amounts of liquidity and are involved in economies of scale and scope activities. Large banks are more diversified which decreases risks and permits banks to have less stable funding and capital. Furthermore, larger bank might be active in market-based activities that necessitate high fixed costs and economies of scale. Whereas, small banks may be more involved in traditional lending type. However, when large banks have less capital and fewer deposits their impact on systemic risk increases. Overall, large banks create more systemic risk than smaller banks (Leaven et al., 2014) therefore the following hypothesis is tested:
H3) Presence of large foreign banks has a more negative effect on financial stability than the
presence of small foreign banks.
Increased profitability results in higher bank stability since the bank has more funds to meet contingencies (Adusei, 2015). More profitable banks might improve financial stability if their profits become part of equity capital and in turn strengthen the capital base (Flamini, McDonald & Schumacher, 2009). Profitable banks are able to create higher capital buffers from their earnings and are therefore less sensitive to shocks in the financial system. Correspondingly, Van Oordt & Zhou (2014) found a negative relation between capital ratios of banks and systemic risk. Besides capital buffers, also profitability is negatively related to systemic risk, where a 1 percentage point increase in return on equity decreases a bank’ individual risk and sensitive to shocks in the financial system by 0.4% (Van Oordt & Zhou, 2014). Moreover, higher capital buffers decrease the risk of default and increases financial stability. Strong banks tend to have more equity and less debt as well as assorted funding structures. As a consequence, regulations (intended by Basel III) require higher capital buffers to reduce financial fragility (IMF, 2013).
9
(2010) show that foreign banks are not significantly more profitable than domestic bank, where the mode of entry is of importance since greenfield banks feature a higher return of asset than domestic banks. The following hypothesis is tested based on previous literature:
H4) Presence of foreign banks with higher profitability has a more positive effect on financial
stability than that of foreign banks with lower profitability.
2.4 Heterogeneity of host- and home- market
The differences between foreign banks and domestic banks are caused by the home and host market conditions. In fact, the different economic development of host countries shows a remarkable dissimilarity in foreign bank entry. Foreign bank entry increased substantially in emerging and developing countries with 74% and 113%, while in OECD countries foreign bank entry increased remarkably less with 38%. Henceforth, foreign banks in emerging and developing countries account for dominant roles in financial intermediation. Furthermore, OECD countries tend to have a larger foreign bank export share than emerging and developing countries (Claessens & Van Horen, 2014).
10
scale advantages of foreign banks. Hence, these advantages do not counterpart the information disadvantage of foreign banks (Claessens & Van Horen, 2012).
This study takes into account the host country characteristics and divides between foreign banks entering emerging, developing or OECD countries, as in line with Claessens & Van Horen (2012). Home country characteristics are not taken into account due to lack of data availability for the selected time period but is definitely recommended for future studies as will be explained in section 6.
2.5 Conceptual Model
The conceptual model visualized in figure 1 provides an overview of the model analysed in this paper. The impact of foreign bank presence on financial stability is tested, where both bank-level and country-level data is used to test financial stability. To clarify the conceptual model below, two separate regression models are used to test the effect on the different financial stability levels. Furthermore, several bank- and country-level characteristics influence the financial stability. Additionally, the relationship between foreign bank presence and financial stability is extensively explained in section 2.2 and likewise provided in the conceptual model under the header ‘how/ why related’ using key words.
FIGURE 1. Conceptual Model
The first hypothesis is related to the effect of foreign bank presence on bank-level financial stability, while the second hypothesis addresses the country-level financial stability. Separate data is used to test these two hypotheses, as explained in section 3.0.
Foreign bank presence
Financial stability How/why related: spill over effect,
increased competition, risk-taking behavior
Influenced by bank- and country- specific characteristics: profitability, size, funding
structure, GDP (growth), inflation
11
3. Data & Method
Section 3.1 extensively describes the data used in this paper and provides an overview of the countries included as well as the bank ownership, region and income group of the countries in the sample. Furthermore, section 3.2 visualizes the empirical model.
3.1 Data
The extended ownership database of Claessens & Van Horen (2015) is exploited to measure foreign bank presence. This database provides information of 5,498 different type of banks in 138 countries for a time period between 1995 and 2013. The ownership database covers at least 90% of each banking system in terms of assets by including host countries with more than 5 active banks and only reports the 100 largest banks for each country in terms of assets. Foreign bank presence is measured as the assets held by foreign banks as a share of total assets in the country. A bank is defined to be foreign if at least 50% of the shares are held by foreigners and the home country of the largest foreign shareholder is stated as the country of origin (Claessens & Van Horen, 2014). The ownership of bank is included as a dummy variable that is 0 if the bank is domestic owned and 1 if is foreign owned (FOR). The database makes it possible to calculate the foreign banks share (FBS) via the number of foreign banks to the total number of banks operating in the country (Demirgüç-Kunt et al., 1998). Besides, the database provides the index number of every bank which makes it possible to combine the data with the Bankscope database.
12 TABLE 1. Foreign bank share
TABLE 2. Ownership, income group and region
Country Income Group No. of foreign banks ‘99 No. of foreign banks ‘13 Foreign bank share ‘13 Country Income Group No. of foreign banks ‘99 No. of foreign banks ‘13 Foreign bank share ‘13
Argentina EM 34 20 0.32 Ghana DEV 6 12 0.63
Armenia DEV 4 12 0.80 Honduras DEV 4 9 0.53
Austria OECD 5 10 0.12 Hungary EM 25 20 0.80
Belgium OECD 10 13 0.46 India EM 6 8 0.12
Bolivia DEV 5 3 0.30 Indonesia EM 26 31 0.48
Bosnia and Herzegovina
DEV 5 14 0.64 Latvia EM 7 11 0.55
Brazil EM 54 50 0.40 Lebanon EM 17 16 0.36
Bulgaria EM 14 17 0.65 Lithuania EM 2 6 0.75
Canada OECD 22 19 0.37 Macedonia DEV 3 8 0.67
China EM 7 29 0.20 Mexico EM 19 16 0.37
Colombia EM 9 8 0.42 Morocco EM 5 4 0.36
Costa Rica DEV 13 9 0.21 Mozambique DEV 10 11 0.85
Croatia EM 14 16 0.52 Paraguay DEV 13 7 0.64
Czech Republic EM 14 13 0.62 Peru EM 10 11 0.69
Denmark OECD 3 5 0.08 Poland EM 30 32 0.76
Dominica DEV 2 3 0.08 Romania EM 13 23 0.82
Egypt, Arab Rep.
EM 5 13 0.54 Russian
Federation
EM 22 31 0.17
Estonia EM 2 6 0.75 Switzerland OECD 22 17 0.20
France OECD 5 4 0.04 United Kingdom
OECD 38 48 0.58
Germany OECD 11 14 0.14 United States OECD 13 18 0.31
Mean Standard deviation Ownership Foreign 0.70 0.46 Domestic 0.30 0.46 Income Group OECD 0.08 0.27 Developing 0.45 0.50 Emerging 0.47 0.50 Regions
Eastern Europe and Central Asia (ECA) 0.14 0.35
Middle East and North Africa (MENA) 0.07 0.25
Organisation for Economic Co-operation and Development (OECD)
0.48 0.50
Sub-Saharan Africa (SSA) 0.01 0.07
Latin America and Caribbean (LAC) 0.23 0.42
South Asia (SA) 0.07 0.26
13
The data for all the bank-level financial stability indicators is obtained via Bankscope, furthermore, these measures are in line with several studies (Berger, Klapper & Turk-Ariss, 2009; Lee & Hsieh, 2014). Moreover, country-level financial stability is measured using nonperforming loans to total gross loans, data obtained via the World Bank (2016b), and the Z-index provided by Datamarket (2016). The Z-score is a proxy for bank’s probability of failure (Lee & Hsieh, 2014; Beck et al., 2012; Leaven & Levine, 2009) and used as a measure of financial stability since it addresses the distance from insolvency (Beck et al., 2012). The Z-score compares the country’s banking system buffer, via capitalization and returns, with the volatility of the returns (Datamarket, 2016). Higher financial stability and thus a lower bank risk is implied by a higher Z-score value (Lee & Hsieh, 2014). Furthermore, the natural logarithm of the Z-score is used since the indicator is highly skewed (Beck et al., 2012). The Z-score is calculated as follows:
𝑍𝑖,𝑡 =
𝑅𝑂𝐴𝑖,𝑡+ (𝐸𝐴)
𝑖,𝑡
𝜎 (𝑅𝑂𝐴)𝑖,𝑡
Where ROA is the return on average assets, E/A stands for the equity to asset ratio, σ (ROA) is the standard deviation of the return on average assets, for bank i at time t. The standard deviation is calculated using previous 4 year moving average of return on average assets, as suggested by Delis, Tran & Tsionas (2009). Besides ROA, also ROE is widely used as a measure for bank profitability. Whereas this study, as in line with Lee & Hsieh (2014), modifies the original Z-index by replacing ROA with ROE to construct an additional bank-level financial stability measure, also using a 4 year moving average of return on average equity to calculate the standard deviation. Note the bank-level Z-score using ROA (ZA) and using ROE (ZE) are calculated via Eq. 1, with data from Bankscope. While the country-level Z-score (CZA) using ROA, is provided at Datamarket. Bank nonperforming loans to total gross loans, hereafter nonperforming loans, is the value of nonperforming loans divided by the total value of the loan portfolio. Where the amount of loans defined as nonperforming include the gross value of loans as stated on the balance sheet (World Bank, 2016b). Nonperforming loan ratio is used as a fragility indicator as in line with multiple studies (Levy Yeyati & Micco, 2007; Demirgüç-Kunt & Detragiache, 2000). Demirgüç-Kunt & Detragiache (2000) state that systemic risk occurs when emergency measures are enabled to help the banking systems, for e.g. deposit freezes and bank holidays, or if huge nationalizations arise. Besides, systemic risk exists when the nonperforming assets reached at least 10% of the total
14
assets. Bank’s nonperforming loans (NPL) is obtained via Bankscope (2016) and data on countries nonperforming loans (CNPL) is acquired via World Bank (2016b). A lower nonperforming loans ratio indicates higher bank/ country stability (World Bank, 2016b). Lastly, the capital ratio, equity to total assets (ETA), is used to indicate overall bank risk (Berger et al., 2009). Moreover, lower bank risk is indicated with higher capital ratio values (Lee & Hsieh, 2014),
Multiple variables are included in the bank-level analysis to control for bank specific characteristics. The banks’ profitability is indicated using return on average assets (ROA) which compares efficiency and operational performance of banks. Return on assets is calculated using the returns created from assets financed by the bank. Another profitability measure used is the return on average equity (ROE), indicating the return on shareholder funds. The ratio of liquid assets to deposit and short-term funding (FUND) is used to measure the percentage of customer and short-term funds that can be obtained if they would instantaneously be withdrawn. Higher FUND percentages indicate higher bank liquidity and less vulnerability to a bank run (Bankscope, 2016). Lastly, bank size (SIZE) is included as a control variable as the natural logarithm of total assets in US dollar, in line with Klomp & De Haan (2011). Data of the above described bank specific characteristics are obtained from Bankscope.
15
To summarise, the effect of foreign bank presence on bank- and country-level financial stability is analysed in this paper. Whereby, nonperforming loans and the Z-score are the indicators used to measure financial stability on both bank- and country-level. An extensive list of the variables, their definition and source is provided in Appendix I.
3.2 Method
3.2.1 Bank-level Analysis
This study applies a panel data approach by using a fixed effects model including country- and time-specific fixed effect for the period 1999-2013. Henceforth, a Hausman test for endogeneity is performed and the model is tested on heteroscedasticity and correlation. The basic model to answer hypothesis 1 is similar to the model used by Leaven & Levine (2008):
𝐼𝑖𝑗,𝑡 = 𝛼0+ 𝛼1𝐹𝑂𝑅𝑖𝑗,𝑡+ 𝛼2𝐵𝑖𝑗,𝑡+ 𝛼3𝑋𝑖,𝑡+ 𝜂𝑖 + ∅𝑡+ 𝜀𝑖𝑗,𝑡
Where 𝐼𝑖𝑗,𝑡 is the dependent variable including a set of risk exposure indicators to measure
bank-level financial stability. The bank-level financial stability indicators used are the Z-score using ROA (ZA), Z-score using ROE (ZE), nonperforming loans (NPL) and the capital ratio (ETA), as in line with previous studies (Lee and Hsieh, 2014; Beck et al., 2012; Berger et al., 2009; Leaven & Levine, 2008). Furthermore, i refers to the country number; j refers to the bank number; t is the time period; and lastly, 𝛼 is an estimated parameter. The independent variable, 𝐹𝑂𝑅𝑖𝑗,𝑡 indicates
foreign bank ownership in country i for bank j at time t and is an indicator variable indicating 1 if the bank is foreign owned and 0 otherwise. Next, 𝐵𝑖𝑗,𝑡 is a set of bank specific variables for bank
j in country i at time t, and 𝑋𝑖,𝑡 is a set of country specific variables for country i at time t. Lastly,
unobserved country specific effect 𝜂𝑖, time specific effect ∅𝑡 and the error term are included. The bank specific variables include the natural logarithm of bank size (SIZE), return on average assets (ROA), return on average equity (ROE) and liquid assets to deposit and short-term funding (FUND). The country specific variables are GDP per capita (GDP), GDP growth (GDPG), inflation (INF) and regulatory quality (REG). Heterogeneity of foreign banks is analysed by calculating the above and below median bank size and profitability of foreign banks. The median values are used in order to avoid biased values due to skewness of values in the sample. A slope-indicator variable stating 1 if the foreign bank features above median bank size or profitability and 0 otherwise. The following regression models is created to analyse hypothesis 3:
16
𝐼𝑖𝑗,𝑡 = 𝛼1𝐹𝑂𝑅 (𝑆𝐼𝑍𝐸 > 𝑚𝑒𝑑𝑖𝑎𝑛)𝑖𝑗,𝑡+ 𝛼1𝐹𝑂𝑅 (𝑆𝐼𝑍𝐸 < 𝑚𝑒𝑑𝑖𝑎𝑛)𝑖𝑗,𝑡+
𝐵𝑖𝑗,𝑡−1+ 𝛼3𝑋𝑖,𝑡+ 𝜂𝑖 + ∅𝑡+ 𝜀𝑖𝑗,𝑡
A similar model is created to test the differences in banks’ profitability using above and below median return on average assets. The model below is created to analyse hypothesis 4:
𝐼𝑖𝑗,𝑡 = 𝛼1𝐹𝑂𝑅 (𝑅𝑂𝐴 > 𝑚𝑒𝑑𝑖𝑎𝑛)𝑖𝑗,𝑡+ 𝛼2𝐹𝑂𝑅 (𝑅𝑂𝐴 < 𝑚𝑒𝑑𝑖𝑎𝑛)𝑖𝑗,𝑡+
𝐵𝑖𝑗,𝑡−1+ 𝛼3𝑋𝑖,𝑡+ 𝜂𝑖 + ∅𝑡+ 𝜀𝑖𝑗,𝑡
Where the variables in both Eq. (3) and (4) are similar to Eq. 2, apart from the interaction variable of foreign banks with above and below median size (Eq. 3) and foreign banks with above and below median ROA (Eq. 4).
3.2.2 Country-level Analysis
In addition to the bank-level analysis, this study also analyses the effect of foreign bank presence on country-level financial stability. The selected time frame (1999-2013), countries and country specific control variables are similar to the bank-level analysis. Besides, the two dependent variables are the nonperforming loans on country-level and the Z-Score on country level. The regression equation used to answer hypothesis 2 is as followed:
𝐶𝑖,𝑡 = 𝛼0+ 𝛼1𝐹𝐵𝑆𝑖,𝑡 + 𝛼2𝑋𝑖,𝑡+ 𝜂𝑖+ ∅𝑡+ 𝜀𝑖,𝑡
Where Ci,t is the dependent variable including the country-level risk exposure indicator
nonperforming loans (CNPL) and the country-level Z-score using ROA (CZA). Besides, i refers to the country number; t is the time period; and 𝛼 is an estimated parameter. The independent variable of the country-level analysis is the number of foreign banks to the total number of banks in the country (FBS). 𝑋𝑖,𝑡 is a set of country specific variables for country i at time t. Lastly, unobserved country specific effect 𝜂𝑖, time specific effect ∅𝑡 and the error terms are included. The country specific control variables are similar to the bank-level analysis and include GDP per capita (GDP), GDP growth (GDPG), inflation (INF) and regulatory quality (REG). Bank specific characteristics are not included in the model due to lack of data availability for the selected time period. For clarification on which variable is included in the bank- or country-level analysis, together with the definition and source of the variables used, see Appendix I.
(3)
(4)
17
4. Empirical Result
The first section provides the results of the effect of foreign bank presence on bank-level nonperforming loans, including several robustness checks in section 4.1.1. Section 4.2 provides the result of foreign bank presence on country-level financial stability.
4.1 Bank-level Analysis
A fixed-effect model with robust standard errors and country and year fixed effects is used to study the effect of foreign bank presence on bank-level financial stability. An overview of the summary statistics is given in Appendix II and the correlation matrix in Appendix III. The heteroscedasticity and Hausman test are provided in Appendix IV and V. The heteroscedasticity test shows that there is evidence of heteroscedastic data which is solved by using robust standard errors and the Hausman test shows that the fixed effect model is superior to the random effects model. The different income groups (developing, emerging and OECD countries) as stated by Claessens & Van Horen (2014) are pooled together in this regression, resulting in approximately 6000 observations of 40 countries.
18
TABLE 3. The effect of foreign bank presence on bank-level financial stability by controlling for bank and country
level characteristics
Note: The regression includes country and year fixed effects and clustered robust standard errors at the country-year
level given in parentheses. (*), (**), (***) denote significance levels of 10%, 5% and 1%, respectively. The panel data regression includes the dependent variables ‘ZA’ representing the Z-index for ROA and ‘ZE’ stating the Z-index for ROE, ‘ETA’ which stand for equity/total assets and ‘NPL’ nonperforming loans to total gross loans. The independent variable ‘FOR’ is a dummy variable stating that the bank is foreign owned relative to domestic owned. The country level control variables include ‘GDP’ GDP per capita in US$, ‘GDPG’ GDP growth in %, ‘INF’ inflation in %. ‘REG’ is the regulatory quality ranging from -2.5 to 2.5. The bank level control variables are ‘SIZE’ natural logarithm of total assets in US$, ‘ROA’ return on average assets in %, ‘ETA’ equity to total assets in % and ‘FUND‘ liquid assets to deposit and short term funding in %. The selected time frame for all variables is the period 1999 to 2013.
The control variables that are significant for both regression results in column 1 and 2 have similar coefficients signs. GDP has a negative effect on bank-level financial stability, the results are statistically significant at a 1% significance level. Inflation is positively related to bank-level financial stability and statistically significant at a 1% and 10% significance level for the ZA and ZE indicators. A predicted inflation rate where as a result interest rates can be changed might
(1) (2) (3) (4)
VARIABLES ZA ZE ETA NPL
FOR -0.0735* -0.153*** -0.322 39.89
(0.0425) (0.0555) (1.033) (66.02)
GDP -3.75e-14*** -2.59e-14*** 1.49e-12*** 4.00e-11***
(8.10e-15) (1.00e-14) (1.82e-13) (9.43e-12)
GDPG -0.000865 0.00283 -0.0332 -0.858 (0.00172) (0.00220) (0.0345) (2.331) INF 0.00465*** 0.00337* 0.00788 -3.837* (0.00138) (0.00178) (0.0262) (2.209) REG 0.151*** 0.0517 -1.122 -108.7* (0.0399) (0.0516) (1.342) (64.64)
FUND 9.88e-07 -9.23e-06** -3.66e-05 -0.00250
(3.68e-06) (4.68e-06) (5.80e-05) (0.00499)
ROA 0.0468*** 0.0465*** 0.134 -20.99*** (0.00246) (0.00466) (0.141) (5.471) SIZE 0.139*** 0.112*** -6.569*** -75.09** (0.0214) (0.0274) (0.815) (34.81) Constant 0.526*** 0.0576 50.58*** 761.2*** (0.127) (0.163) (4.812) (228.4) Observations 6,051 5,795 6,359 1,483 R-squared 0.076 0.025 0.139 0.094 Number of ID 408 408 424 99
Country FE YES YES YES YES
19
explain the positive effect of inflation on bank-level financial stability (Perry, 1992). Next, regulatory quality positively influenced bank-level financial stability using the ZA indicator and is statistically significant at a 1% level. Liquid asset to deposit and short term funding negatively influences bank-level financial stability using the ZE indicator, statistically significant at a 10% significance level. Return on average assets and bank size positively influences bank-level financial stability and are statistically significant at a 1% significance level for both dependent variables.
Table 4 shows the results of foreign bank presence on bank-level financial stability, taking into account bank heterogeneity. Heterogeneity is measured by difference in bank size and bank profitability since this tends to influence financial stability. Columns 1 and 5 represent the regression results for foreign bank presence with below median profitability (FOR*LOWROA) and columns 2 and 6 show the results for the foreign bank presence with above median profitability (FOR*HIGHROA) in terms of ROA. The independent variable is significant at a 1% significance level for both the dependent variables. Hence, a 1% increase in foreign bank presence with below median profitability decreases the bank-level financial stability with 2.78% (column1). Whereas a 1% increase in foreign bank presence with above median profitability increases bank-level financial stability with 2.43 % (column2). The coefficient is also negative for foreign bank with below median profitability (column 5) and positive for foreign bank presence with above median profitability (column 6), using the dependent variable ZE. Henceforth, hypothesis 3 stating that presence of foreign banks with high profitability has a more positive effect on bank-level financial stability than that of foreign bank with lower profitability cannot be rejected. This is in line with Adusei (2015) and Van Oordt & Zhou (2014) stating that increased bank profitability results in higher financial stability, since increased amount of funds results in higher ability of meeting contingencies. Consequently, banks are less sensitive to shocks in the financial system.
20
TABLE 4. The effect of foreign banks with above/below median size and profitability on bank-level financial stability
(1) (2) (3) (4) (5) (6) (7) (8) VARIABLES ZA ZA ZA ZA ZE ZE ZE ZE FOR*LOWROA -0.278*** -0.414*** (0.0389) (0.0413) FOR*HIGHROA 0.243*** 0.328*** (0.0351) (0.0357) FOR*SMALLSIZE -0.135** -0.116* (0.0631) (0.0662) FOR*LARGESIZE 0.118* 0.0492 (0.0635) (0.0652)
GDP -1.84e-14 -2.19e-14* -2.26e-14* -2.39e-14* -1.15e-14 -1.62e-14 -1.71e-14 -1.69e-14
(1.23e-14) (1.28e-14) (1.25e-14) (1.26e-14) (1.34e-14) (1.45e-14) (1.42e-14) (1.42e-14)
GDPG -0.00289 -0.00288 -0.00213 -0.00219 0.00146 0.00163 0.00261 0.00243 (0.00261) (0.00260) (0.00258) (0.00259) (0.00235) (0.00234) (0.00235) (0.00235) INF 0.00514** 0.00530** 0.00587*** 0.00592*** 0.00335 0.00368* 0.00427* 0.00432* (0.00207) (0.00208) (0.00215) (0.00215) (0.00207) (0.00221) (0.00220) (0.00223) REG 0.219*** 0.198*** 0.236*** 0.225*** 0.0879 0.0617 0.103 0.100 (0.0646) (0.0646) (0.0670) (0.0671) (0.0664) (0.0678) (0.0692) (0.0696)
FUND 6.09e-08 3.21e-07 9.30e-07 9.92e-07 -1.00e-05 -9.64e-06 -8.79e-06 -8.92e-06
(5.07e-06) (5.09e-06) (5.23e-06) (5.23e-06) (6.41e-06) (6.43e-06) (6.48e-06) (6.48e-06)
ETA 0.00905*** 0.00968*** 0.0103*** 0.0105*** 0.00837*** 0.00932*** 0.00968*** 0.00977*** (0.00260) (0.00265) (0.00270) (0.00274) (0.00226) (0.00235) (0.00235) (0.00238) Constant 1.278*** 1.217*** 1.244*** 1.214*** 0.675*** 0.588*** 0.627*** 0.604*** (0.0662) (0.0665) (0.0691) (0.0686) (0.0704) (0.0723) (0.0728) (0.0731) Observations 6,051 6,051 6,051 6,051 5,795 5,795 5,795 5,795 R-squared 0.044 0.040 0.025 0.024 0.041 0.033 0.013 0.011 Number of ID 408 408 408 408 408 408 408 408
Country FE YES YES YES YES YES YES YES YES
Time FE YES YES YES YES YES YES YES YES
Note: See note to Table 3. ‘FOR*LOWROA’ and ‘FOR*HIGHROA’ are interaction terms for the foreign ownership dummy variable and below/ above median
21
presence of foreign banks with above median size (FOR*LARGESIZE) on bank-level financial stability. The coefficient is positive and significant at a 10% significance level, using the dependent variable ZA. Indicating that a 1% increase in large foreign bank presence increases bank-level financial stability with 1.18%. Therefore, we reject hypothesis 2 stating that presence of large foreign banks has a more negative effect on financial stability than that of small foreign banks. Thus, the results are not in line with the expectations based on the literature review, since most studies found that large banks account for higher financial risk (Adusei, 2015; Van Oordt & Zhou, 2014; Leaven et al., 2014). However, large banks might lower financial risk since they have higher diversification as a consequence risk is decreased and less stable funding and capital is needed (Leaven et al., 2014).
Next, host market heterogeneity is tested by taking into account different host market development levels as divided by Claessens & Van Horen (2015) in developing, emerging and OECD countries. Table 2 provides an overview of the mean and standard deviation of the income groups of the host countries included in the sample. The results of the division on host country income groups is provided in table 7, Appendix VI. The results are however insignificant for the ZA index and only significant for foreign banks from emerging countries when using the dependent variable ETA, as can be seen in column 5. A remarkable observation is that the presence of foreign banks from emerging countries has a positive and significant effect on bank-level financial stability, at a 5% significance level. The results for developing (column 4) and OECD (column 6) countries are insignificant but the coefficients are even higher for OECD countries and negative for developing countries. Despite the insignificant results, the sign of the coefficients might emphasize that the impact of foreign bank presence might be related to the host country financial development.
To conclude, this paper shows using the Z-index that large foreign banks and foreign banks with high profitability have a positive effect on bank-level financial stability. The results are consistent between different financial stability measures.
4.1.1 Robustness Check
22
23
TABLE 5. The effect of foreign bank presence on bank-level financial stability, introducing a lag term, ROE and interaction variable
(1) (2) (3) (4) (5) (6) (7) (8)
VARIABLES ZA+1 ZE+1 ETA+1 NPL+1 ZA ZE ZE ZA
FOR -0.104** -0.107* -0.530 24.06 -0.120*** -0.193*** -0.156*** -0.0817**
(0.0437) (0.0562) (0.521) (37.42) (0.0431) (0.0557) (0.0556) (0.0413)
GDP -3.78e-14*** -9.29e-15 9.59e-13*** 2.66e-11*** -4.12e-14*** -2.82e-14*** -2.37e-14** -2.69e-14***
(8.32e-15) (1.02e-14) (9.97e-14) (5.33e-12) (8.13e-15) (1.01e-14) (1.01e-14) (7.80e-15)
GDPG 0.0132*** 0.00856*** -0.0673*** -6.681*** -0.00135 0.00315 0.00268 -0.00214 (0.00178) (0.00230) (0.0216) (1.909) (0.00175) (0.00221) (0.00220) (0.00167) INF 0.00231 0.00210 0.0168 -2.878* 0.00622*** 0.00463*** 0.00317* 0.00362*** (0.00140) (0.00177) (0.0172) (1.665) (0.00139) (0.00178) (0.00178) (0.00134) REG 0.103** 0.00726 -0.994** -101.3*** 0.164*** 0.0526 0.0480 0.127*** (0.0406) (0.0512) (0.502) (39.12) (0.0405) (0.0519) (0.0517) (0.0387)
FUND 7.97e-06** 6.29e-06 -4.87e-05 -0.00554 -2.38e-07 -9.80e-06** -9.63e-06** -2.05e-06
(3.81e-06) (4.77e-06) (4.65e-05) (0.00353) (3.74e-06) (4.71e-06) (4.69e-06) (3.58e-06)
ROA 0.0253*** 0.0251*** 0.0409 -25.03*** -0.00155 -0.191***
(0.00219) (0.00288) (0.0265) (4.010) (0.0274) (0.0129)
SIZE 0.119*** 0.0345 -4.248*** -30.69 0.153*** 0.109*** 0.0998*** 0.0941***
(0.0221) (0.0276) (0.269) (21.56) (0.0217) (0.0275) (0.0282) (0.0209)
ROE 4.91e-05*** 3.07e-05***
(3.66e-06) (4.74e-06) ROA*SIZE 0.00941* 0.0488*** (0.00528) (0.00261) Constant 0.511*** 0.511*** 37.04*** 538.8*** 0.348*** 0.0460 0.126 0.768*** (0.164) (0.164) (1.598) (137.4) (0.130) (0.164) (0.167) (0.124) Observations 5,786 5,786 6,349 1,480 6,051 5,795 5,795 6,051 R-squared 0.019 0.019 0.051 0.090 0.047 0.015 0.026 0.130 Number of ID 421 421 424 138 408 408 408 408
Country FE YES YES YES YES YES YES YES YES
Time FE YES YES YES YES YES YES YES YES
24
4.2 Country-level Analysis
A cross-country panel data analysis is performed to study the effect of foreign bank presence on country-level financial stability for the period 1999-2013 for 40 countries. An overview of the summary statistics is given in Appendix II and the heteroscedasticity and Hausman test are provided in Appendix IV and V. The Hausman test shows that the fixed effect model is more suitable than the random effects model and the heteroscedasticity test shows evidence of heteroscedastic data, to solve this issue robust standard errors are used. Thus, a fixed-effect model with robust standard errors is used, including country and year fixed effects.
Table 6 provide the results of the regression Eq. 5. The first column shows the results for foreign bank presence in terms of foreign bank share on country-level financial stability using nonperforming loans indicator, the results are statistically insignificant. Column 2 shows the result of foreign bank presence on financial stability using the country-level Z-index. The coefficient is positive and statistically significant at a 5% significance level, indicating a negative relationship between foreign bank presence and country-level financial stability. Where a 1% increase in foreign bank presence negatively impacts country-level financial stability with 0.0337%. Hence, hypothesis 2 stating that foreign bank presence has a negative effect on financial stability cannot be rejected. The negative impact of foreign bank presence is consistent with the results on the bank-level analysis.
The lagged independent variable FBS is given in column 3 and 4 to control for autoregressive tendencies, similar to the robustness check performed in section 4.1.1. The regression results for the effect of the one-year lagged foreign bank presence on nonperforming loans is again statistically insignificant as can be seen in column 3. Column 4 shows the results of the one-year lagged foreign bank presence on the Z-index, the coefficient is statistically significant at a 10% significance level. Hence, similar to the regression results in column 2, the coefficient is also negative when using the one-year lagged foreign bank presence variable.
25
TABLE 6. The effect of foreign banks presence on country-level financial stability
(1) (2) (3) (4)
VARIABLES CNPL CZA CNPL CZA
FBS 0.215 -0.00337**
(0.216) (0.00157)
FBS-1 0.0909 -0.00287*
(0.237) (0.00166)
GDP -1.79e-12** 1.27e-14 -1.79e-12** 1.26e-14
(8.36e-13) (9.33e-15) (8.36e-13) (9.36e-15)
GDPG -0.514*** 0.00749*** -0.511*** 0.00743*** (0.0840) (0.00259) (0.0844) (0.00256) INF -0.0146 -0.00172 -0.0111 -0.00175 (0.0593) (0.00267) (0.0593) (0.00269) REG -0.281 0.0190 -0.457 0.0220 (2.117) (0.0329) (2.173) (0.0342) Constant 9.580*** 0.923*** 9.694*** 0.922*** (1.293) (0.0244) (1.304) (0.0246) Observations 600 583 599 582 R-squared 0.171 0.064 0.168 0.065 Number of ID 40 39 40 39
Country FE YES YES YES YES
Time FE YES YES YES YES
Note: The regression includes country and year fixed effects and clustered robust standard errors at the country-year
level given in parentheses. (*), (**), (***) denote significance levels of 10%, 5% and 1%, respectively. The panel data regression is conducted for the period 1999-2013 and the dependent variables include ‘CNPL’ indicating the country-level nonperforming loans and the ‘CZA’ indicating the country-country-level Z-index using ROA. The independent variable ‘NFB’ states the number of foreign banks as included in the sample as represented in table 1, the other independent variable is the ‘FBS’ foreign bank share. For an explanation of the country specific control variable see note table 3.
26
term of foreign bank entering emerging countries is included (column 5). The results for the third column are statistically insignificant. Based on the first two columns one can conclude that foreign banks presence in developing country influences country-level financial stability positively while foreign bank presence in emerging countries is negatively associated with country-level financial stability. Even though the regression results are insignificant for column 3, the interaction term indicates that foreign banks presence in OECD countries has an even higher negative impact. The interaction term is statistically insignificant when the Z-Score is used as dependent variable.
TABLE 7. The effect of foreign bank presence on country-level financial stability, introducing host country
characteristics
(1) (2) (3) (4) (5) (6)
VARIABLES CNPL CNPL CNPL CZE CZE CZE
FBS 0.272** -19.95** 0.208 -0.00353** 0.0371 -0.00330** (0.118) (8.057) (0.224) (0.00172) (0.209) (0.00153) FBS*DEV -23.48*** 0.0649 (7.425) (0.212) FBS*EM 20.22** -0.0406 (8.051) (0.210) FBS*OECD 40.33* -0.410 (21.54) (1.033)
GDP -1.78e-12** -1.62e-12* -2.11e-12*** 1.27e-14 1.24e-14 1.60e-14
(8.48e-13) (9.59e-13) (5.60e-13) (9.33e-15) (9.53e-15) (9.61e-15)
GDPG -0.542*** -0.543*** -0.504*** 0.00756*** 0.00754*** 0.00739*** (0.0757) (0.0762) (0.0833) (0.00253) (0.00252) (0.00257) INF -0.0188 -0.0173 -0.0164 -0.00170 -0.00171 -0.00170 (0.0556) (0.0562) (0.0587) (0.00268) (0.00268) (0.00266) REG 1.225 0.958 -0.167 0.0148 0.0165 0.0178 (2.000) (2.024) (2.105) (0.0304) (0.0307) (0.0330) Constant 11.71*** 12.42*** 7.582*** 0.917*** 0.917*** 0.944*** (1.568) (1.929) (1.684) (0.0347) (0.0430) (0.0581) Observations 600 600 600 583 583 583 R-squared 0.226 0.214 0.180 0.065 0.065 0.066 Number of ID 40 40 40 39 39 39
Country FE YES YES YES YES YES YES
Time FE YES YES YES YES YES YES
27
5. Discussion & Limitation
This section compares the empirical results provided in section 4.1 and 4.2 with results of previous studies. The main finding of the paper is that foreign bank presence significantly negatively impacts financial stability. Remarkable is the similar outcomes of the bank-level and country-level analysis, since both showed evidence of a negative relationship. This result is in line with the studies of Lee & Hsieh (2014) and Angkinand & Wihlborg (2010) who studied the impact of foreign bank ownership on financial stability as overall bank risk by using the Z-index. They also found a negative relationship between foreign ownership and financial stability, which supports the home field advantage hypothesis. The home field advantage hypothesis predicts that domestic banks have advantages over foreign banks. Furthermore, the negative relationship between foreign bank presence and financial stability might be related to the fact that foreign banks cause profitability and margin reduction for domestic banks which in turn affect financial stability (Claessens et al., 2011). Besides, foreign bank presence causes increased competition that might destabilize the banking sector (Barajas et al., 2000; Demirgüç-Kunt & Detragiache, 2005). Also short-term dedication of foreign banks, excessive borrowing, easing international capital flows and conquering regulators are drivers for the negative impact of foreign bank presence on financial stability. The short-term dedication of foreign banks is of high importance since this might result in contagion if foreign banks decide to exit the host country due to home country difficulties. (Demirgüç-Kunt & Detragiache, 2005).
28
This paper has several limitations. Omitted variables that are not provided in the model but do however influence the results could lead to biased results. Furthermore, the financial stability indicators used in this paper have some limitations as well. The nonperforming loans to total gross loans used in this study is found to be a so called ‘noisy’ indicator since it features a discriminative power on sound and unsound banking systems (Čihák & Schaeck, 2007). The Z-score is solely based on accounting data and is therefore as proper as the accounting and auditing framework (World Bank, 2016c). Another important limitation is the lack of home country characteristics in the model due to lack of data. This, however, turned out to be of importance when studying the impact of foreign bank entry (Claessens & Van Horen, 2012). Only host country characteristics are included while it might be an addition to not only study the host country development but also the home country development. Another aspect that might influence the results of this study is the degree of competition in the home country since this might lead to efficiency, quality and innovation advantages for foreign banks and in turn causes more efficient operations in the host country (Claessens & Van Horen, 2012).
29
6. Conclusion
Foreign bank presence increased rapidly caused by communication and technology development, deregulations and economic integrations in the last decade (Claessens & Van Horen, 2012). Its impact on financial stability raised concerns among policy makers (Demirgüç-Kunt, Levine & Min, 1998) but nevertheless did not receive spacious attention in literature and, on top of that, the available literature is far from univocal (Claessens & Van Horen, 2012).
This study fills the gap in the literature by not only analysing the impact of foreign bank presence on bank-level financial stability but, in addition, studies the impact on country-level financial stability. In fact, this is the first paper to my knowledge that studies the effect of foreign bank presence on country-level financial stability using an extensive database of more than 400 banks from 40 countries in the period 1999-2013. Altogether, this paper is an extension on foreign bank entry literature (Claessens & Van Horen, 2001, 2012) and literature on the influence of foreign bank entry on systemic risk (Demirgüç-Kunt & Detragiache, 2005). The following main findings of this paper answer the research question. First, foreign bank presence has a significantly negative effect on bank-level as well as country-level financial stability, supporting the home field advantage hypothesis. Second, foreign banks with below median profitability decreases bank-level financial stability while foreign banks with above median profitability increase bank-level financial stability. Third, small foreign banks negatively influence financial stability while large foreign banks positively influence bank-level financial stability.
30
References
Acharya, V. V., Pederson, L. H., Philippon, T., & Richardson, M. 2010. Measuring systemic
risk. FRB of Cleveland working paper No. 10-02, New York University, Stern School of
Business.
Adusei, M. 2015. The impact of bank size and funding risk on bank stability. Cogent Economics
& Finance, 3: 1-19.
Akins, B., Li, L., NG, J., & Rosticus, T.O. 2016. Bank competition and financial stability:
Evidence from the financial crisis. Journal of Financial and Quantitative Analysis, 51(1): 1-28. Angkinand, A., & Wihlborg, C. 2010. Deposit insurance coverage, ownership, and banks’ risk-taking in emerging markets. Journal of International Money and Finance, 29(2): 252-274. Banerjee, S. 2012. Basel I and Basel II compliance issues for banks in India. Macroeconomics &
Finance in Emerging Market Economies, 5(2): 228-245.
Barajas, A., Steiner, R., & Salazar, N. 2000. The impact of liberalization and foreign investment in Colombia’s financial sector. Journal of Development Economics, 63: 157-196.
Beck, T. 2008. Bank competition and financial stability: friends or foes? Policy research working paper 4656.
Beck, T., De Jonghe, O., & Schepens, G. 2012. Bank competition and stability: Cross-country heterogeneity. Journal of Financial Intermediation, 22(2): 218-244.
Berger, A. N., Klapper, F. L, & Turk-Ariss, R. 2009. Bank competition and financial stability.
Journal of Financial Service Research, 35(2): 99-118.
Bijlsma, M., Klomp, J., & Duineveld, S. No date. Systemic risk in the financial sector. A review
and synthesis, CPB Netherlands Bureau for Economic Policy Analysis. Retrieved from
https://www.cpb.nl/sites/default/files/publicaties/download/systemic-risk-financial-sector-review-and-synthesis.pdf, accessed October 2016.
31
Claessens, S., Demirgüç-Kunt, A., & Huizinga, H., 2001. How does foreign presence affect domestic banking markets? Journal of Banking & Finance, 25(5): 891-911.
Claessens, S., & Van Horen, N. 2012. Being a foreigner among domestic banks: Asset or liability? Journal of Banking & Finance, 36(5): 1276-1290.
Claessens, S., & Van Horen, N. 2014. Foreign banks: Trends and impact. Journal of Money,
Credit and Banking, 46(s1): 295-326.
Claessens, S., & Van Horen, N. 2015. The impact of the global financial crisis on banking globalization. IMF Economic Review, 63(4): 868-918.
De Haan, J., & Poghosyan, T. 2015. Size and earnings volatility of US bank holding companies.
Journal of Banking & Finance, 36(11): 3008-3016.
Delis, D. M., Tran, K., & Tsionas, E. 2009. Quantifying and explaining parameter
heterogeneity in the capital regulation-bank risk nexus. Munich Persional RePEc Archive.
Retrieved at https://mpra.ub.uni-muenchen.de/18526/1/MPRA_paper_18526.pdf, accessed November 2016.
Demirgüç-Kunt, A., & Detragiache, E. 2000. Does deposit insurance increase banking system stability? An empirical investigation. Journal of Monetary Economics, 47(7): 1373-1406. Demirgüç-Kunt, A., & Detragiache, E. 2005. Cross-country empirical studies of systemic bank
distress: A survey. International Monetary Fund Working Paper 05/96.
Demirgüç-Kunt, A., Levine, R., & Min, H. 1998. Opening to foreign banks: Issues of stability, efficiency and growth. The implications of globalization of world financial markets, ed. by A. Meltzer (Korea: Bank of Korea).
Donaldson, L., & Davis, J. H. 1991. Stewardship theory or agency theory: CEO governance and shareholder returns. Australian Journal of Management, 16(1): 49-65.
Elsinger, H., Lehar, A., & Summer, M. 2006. Risk assessment for banking systems.
Management Science, 52(9): 1301-1314.
Flamini, V., McDonald, C., & Schumacher, L. 2009. The determinants of commercial bank
32
Gang, J., & Qian, Z. 2015. China’s monetary policy and systemic risk. Emerging Markets
Finance & Trade, 51(4): 701-713.
Hasman, A. 2013. A critical review of contagion risk in banking. Journal of Economic Surveys, 27(5): 978-995.
Havrylchyk, O., & Jurzyk, E. 2010. Inherited or earned? Performance of foreign banks in
Central and Eastern Europe. International Monetary Fund working paper. Retrieved from
https://www.imf.org/external/pubs/ft/wp/2010/wp1004.pdf, accessed October 2016.
Hill, R. C., Griffiths, W. E., & Lim, G. C. 2012.Principles of econometrics. Asia: John Wiley &
Sons (Asia) Pte Ltd.
International Monetary Fund (IMF). 2013. Global financial stability report: transition
challenges to stability. Retrieved from
https://www.imf.org/External/Pubs/FT/GFSR/2013/02/pdf/text.pdf, accessed October 2016. Jensen, M. C., & Meckling, W. H. 1976. Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3(4): 305-360.
Klomp, J., & De Haan, J. 2011. Banking risk and regulation: Does one size fit all? Journal of
Banking & Finance, 36 (12): 3197-3213.
Laeven, L., & Levine, R. 2009. Bank governance, regulation and risk taking. Journal of
Financial Economics, 93(2): 259-275.
Laeven, L., Ratnovski, L., & Tong, H. 2014. Bank size and systemic risk. International Monetary Fund, Research Department.
Lee, C., & Hsieh, M. 2014. Bank reforms, foreign ownership, and financial stability. Journal of
International Money and Finance, 40: 204-224.
Lehner, M., & Schnitzer, M. 2008. Presence of foreign banks and their impact on host countries.
Journal of Comparative Economics, 36(3): 430-452.
33
Perry, P. 1992. Do banks gain or lose from inflation? Journal of Retail Banking, 14(2): 25-30. Regehr, K., & Sengupta, R. 2016. Has the relationship between bank size and profitability changed? Economic review, 101(2): 49-72.
Stiglitz, J. E. 1990. Peer monitoring and credit markets. World Bank Economic Review, 4(3): 351-366.
Van Oordt, M., & Zhou, C. 2014. Systemic risk and bank business models. De Nederlandsche Bank Working Paper No. 442, The Netherlands.
Viñals, J., Pazarbasioglu, C., Surti, J., Narain, A., Erbenova, M., & Chow, J. 2013. Creating a
safer financial system: Will the Volcker, Vickers, and liikanen structural measures help? IMF
Staff Discussion Note. Retrieved from
https://www.imf.org/external/pubs/ft/sdn/2013/sdn1304.pdf, accessed October 2016.
Databases
Bankscope. 2016. World banking information source. Retrieved from https://bankscope-bvdinfo-com.proxy-ub.rug.nl/Search.QuickSearch.serv?_CID=1&context=QJ5QD09INB0WQ54, accessed November 2016.
Datamarket. 2016. Bank Z-score from World Bank, citing Bankscope, Bureau van Dijk (BvD). Retrieved from
https://datamarket.com/data/set/28mf/bank-Z-score#!ds=28mf!2rrg=b.21.2s.3i.c.l.2z.2h.1.20.49.32.3u.3a.3p.r.2l.40.1l.4y.f.56.2x.17.4p.53.5g.4 .3v.42.3t.2u.5f.1p.50.2j.n.6.5c.2i&display=line, accessed October 2016.
World Bank. 2016a. Economy & Growth indictors. Retrieved from
http://data.worldbank.org/topic/economy-and-growth?view=chart, accessed November 2016. World Bank. 2016b. Financial sector indicators. Retrieved from
http://data.worldbank.org/topic/financial-sector?view=chart, accessed November 2016. World Bank. 2016c. Financial stability. Retrieved from
http://www.worldbank.org/en/publication/gfdr/background/financial-stability, accessed December, 2016.
World Bank. 2016d. Worldwide Governance Indicators, Regulatory Quality. Retrieved from
34
Appendices
I. Data
Variable Description Source
Dependent variable
Country-level
CNPL Bank nonperforming loans to total gross loans on country-level in percentage, are the value of nonperforming loans divided by the total value of the loan portfolio.
WorldBank
CZE It captures the probability of default of a country's banking system. Z-score compares the buffer of a country's banking system with the volatility of those returns. Calculated as (ROA+(equity/assets))/sd(ROA); where sd(ROA) is the standard deviation of ROA. ROA, equity, and assets are country-level aggregate figures, estimated using bank-by-bank unconsolidated data from Bankscope (Datamarket, 2016). A higher value of the Z-score implies lower bank risk and higher stability (Lee & Hsieh, 2014).
WorldBank & BankScope from Datamarket
Bank-level ETA Bank equity to total assets in percentage. Ratio measures the amount of protection afforded to the bank by the equity they invested in, a higher ratio the higher the protection, hence lower bank risk.
BankScope
ZA Z-index for ROA = ROA+(equity/assets))/sd(ROA), where sd(ROA) is the standard deviation of return on assets as return volatility proxy using a 4-year average.
35
ZE Z-index for ROE = ROE+(equity/assets))/sd(ROE), where sd(ROE) is the standard deviation of return on assets as return volatility proxy using a 4-year average.
Author’s calculation based on Bankscope
NPL Bank nonperforming loans to total gross loans on bank-level. A lower ratio indicates higher bank stability.
Bankscope
Independent variable
Country-level
FBS The number of foreign banks, calculated as number of foreign banks in the sample/ total banks in the sample, per country.
Author’s own calculation based on Claessens & Van Horen database (2015) Bank-level OWN Dummy which is 1 if bank is foreign owned (FOR) and zero if
domestically owned. Foreign if >50% of shares are held by foreigners) and if foreign owned the home country of the majority shareholder.
Claessens & Van Horen database (2015)
Controlled variables
Country specific variable
GDP Gross Domestic Product in current US dollars, at purchaser's prices is the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products.
WorldBank
GDPG Annual percentage growth rate of GDP per capita based on constant local currency.
WorldBank
INF Inflation as measured by the annual growth rate of the GDP implicit deflator shows the rate of price change in the economy as a whole.
WorldBank
REG Regulatory Quality captures perceptions of the ability of the government to formulate and implement sound policies and regulations that permit and
36
promote private sector development. Estimate gives the country's score on the aggregate indicator, in units of a standard normal distribution, i.e. ranging from approximately -2.5 to 2.5.
IR Real interest rate in percentage, include the lending interest rate adjusted for inflation as measured by the GDP deflator.
Worldbank
Bank specific variable
SIZE The LOG of total assets in US dollars. Bankscope
ROA Return on average assets in percentage. This ratio compares the efficiency and operational performance of banks as it measures returns generated from assets that are financed by the bank.
BankScope
ROE Return on average equity in percentage. The return on equity is a measure of the return on shareholder funds.
BankScope
FUND Liquid assets to deposit and short-term funding in percentage. A deposit run off ratio that estimates what percentage of customer and short term funds could be met if withdrawn occurs. A higher ratio indicates less vulnerability to bank runs.
