1
What is the effect of foreign banks on financial stability in developing
and emerging countries?
University of Groningen Faculty of Economics and Business
Master Thesis, M.Sc. Economic Development and Globalization
Student: Hilde van Assen Student ID number: S3455122
Student email: h.s.van.assen@student.rug.nl
Date: June 2019
2
Abstract
The global financial crisis in 2007/2008 demonstrated the importance of financial stability and the effect on economic growth. This crisis changed the attitude towards foreign banks and initiated the debate on the effects of foreign banks on the economy and financial system of the host country (de Haas, 2014). Before, the theoretical models did not incorporate the role of financial stability. Currently, studies that analyse the effect of foreign bank penetration on financial stability are inconclusive. The foreign banks would influence financial stability via different channels. Important channels are a changed level of competition and the attribution to the development of the financial sector. However, the effect of those channels on financial stability is ambiguous in the literature. Research in this topic is important because foreign banks play an important role in the domestic financial system in developing and emerging countries. The research question in this paper is ‘What is the effect of foreign banks on financial stability in developing and emerging countries?’. Foreign banks would have a negative effect on financial stability according to the first hypothesis. The institutional quality would reduce the negative effect of foreign banks on financial stability in the host country according to the second hypothesis. Those hypotheses are test in a dynamic panel data model. The presence of foreign banks is measured as the share of bank assets held by foreign banks in the total number of bank assets and the share of foreign banks to the total number of banks. The former does not have a significant impact on financial stability and the latter has a negative effect on financial stability, which might indicate that the entry mode of foreign banks is relevant. The second hypothesis is not supported by the empirical analysis. The country should develop a set of requirements that have to be fulfilled by the foreign bank that want to become active in the financial sector to ensure that the foreign bank does not harm the economy and stability. For example, the foreign banks might be required to cooperate with a domestic bank.
3
Table of Contents
1. Introduction ... 4 2. Literature review ... 6 3. Methodology ... 13 4. Data ... 18 4.1 Data ... 18 4.2 Descriptive statistics ... 225. Analysis and results ... 25
6. Conclusion ... 31
7. References ... 33
Appendix 1 ... 36
1.1 Box plot of the variable Z-score ... 36
1.2 Histograms for the degree of foreign bank presence in the country... 36
1.3 Box plot for the variable GDP per capita growth ... 37
1.4 Boxplot and scatterplot for the variable domestic credit as % of GDP (financial depth) ... 38
1.5 Box plot and scatterplot for the variable real interest rate ... 39
1.6 Box plot and scatterplot for variable inflation ... 40
1.7 Pie chart for the dummy sovereign default ... 41
1.8 Boxplot for the variable bank size ... 41
1.9 Boxplot for the variable bank liquidity ... 42
1.10 Boxplot and scatterplot for the variable bank capitalization ... 42
1.11 Box plot and scatterplot for the variable non-interest income as percentage of the total income ... 43
1.12 Boxplot and scatterplot for the variable regulatory capital ... 44
Appendix 2 ... 46
2.1 Breusch-Pagan test ... 46
2.2 Test significance of year dummies (time fixed effects) ... 47
2.3 The Arellano-Bond test ... 48
4
1. Introduction
Developing and emerging countries have liberalised their financial system by removing government intervention since 1990 (Masci, 2008:46). Their goal is to develop the financial system, achieve economic growth and catch up with the developed countries. At that time, the neoclassical theory was considered true and liberalisation would lead to growth according to this theory (Claessens & van Horen, 2013). The banking sector would use their capital better, distribute the capital and savings more efficient, make use of a risk-sharing mechanism and have better access to foreign capital (Gaies & Nabi, 2019).
The banking sector in these countries changed drastically after privatisation of banks and relaxation of regulations. Private banks and organisations, which were prohibited to operate earlier, entered the banking sector and increased the level of competition. Chen et al (2017) highlighted that financial liberalisation was symbolized by a rapid increase of foreign banks in emerging and developing countries. The sum of foreign banks expanded by 74% and the market share of foreign banks became twice as large in the period 1995 until 2009 (Wu et al, 2017). De Haas (2014) highlighted that the foreign banks own between 67% and 100% of the bank assets in the majority of the emerging European countries.
The popular theoretical models were based on the neoclassical theory and did not consider which effect foreign banks have on financial stability and in which way this would affect the economic growth and development of the host country. The global financial crisis in 2007/2008 demonstrated the importance of financial stability and the effect on economic growth. This crisis changed the attitude towards foreign banks and initiated the debate on the effects of foreign banks on the economy and financial system of the host country (de Haas, 2014). A stable financial sector would be important for the economy and deterioration of financial stability would harm the economy. Currently, studies that analyse the effect of foreign bank penetration on financial stability are inconclusive. The channels that play an important role in this relationship have an ambiguous effect on financial stability, which might be an important reason for the heterogeneity observed in study results. Besides that, institutions might play an important role according to Claessens & van Horen (2013). However, researchers deal differently with differences in the quality of the institutions and do not always control for these differences.
5 capital to improve the productivity. Besides that, competition lowers interest rates, which would mitigate the moral hazard problem. Furthermore, banks might take less risks as they do not expect the government to bail them out if they fail. They are not regarded as ‘too big to fail (TBTF)’ anymore since new banks entered the financial sector. But, the lower interest rates charged on loans downgrades the bank’s franchise value. Banks could take more risks instead of less risks due to the devaluation of the franchise value, which deteriorates financial stability. Domestic banks that invest in new technologies have to learn how to use them and will not cash the returns immediately. They might take excessive risk in the short run to prevent bankruptcy if the level of competition is high, which deteriorates financial stability. Foreign banks might also pressure the government to improve regulation and supervision in the financial sector, which strengthens the financial stability. Foreign banks might also focus on the most transparent and profitable companies. They them from the domestic banks, which worsens the performance of domestic banks and deteriorates financial stability.
Host countries are more likely to achieve a higher degree of financial stability after foreign banks enter the market if the quality of their rule of law, public services and policies is higher and if they experience political stability. The positive effects of a higher level of competition would be more pronounced in these countries.
Financial stability is not easy to measure and define. Therefore, different definitions for financial stability emerged (Ijtsma 2017; Gadanecz & Jayaram 2009)). There are different elements in the financial system which are important for financial stability. Their interdependence among themselves and the real economy makes the relationship complex (Gadanecz & Jayaram, 2009). However, the literature developed different methods to measure financial stability. Most definitions refer to “the state of the financial system as a whole, which means that financial instability can be interpreted as a state of affairs in which many financial institutions fail or are in distress simultaneously” (Ijtsma 2017:10). This paper will follow this definition of financial stability.
According to the literature, foreign banks can have huge effects on the stability of the financial sector and the living standards of the citizens in the host country. Therefore, policymakers need guidance in their decision to open up the financial sector to foreign parties and in which form they should allow foreign competition. Research in this topic should provide this guidance and is important. Besides that, foreign banks play an important role in the domestic financial system of developing and emerging countries.
This research paper tries to answer the following question:
‘What is the effect of foreign banks on financial stability in developing and emerging countries?’
6
2. Literature review
The global financial crisis in 2007/2008 reinitiated the debate on the effects of foreign banks on financial stability. Researchers were convinced that the entrance of foreign banks always benefits the host country through a reinforcement of the financial stability, but this is not the case anymore. Foreign banks have advantages and disadvantages for the host country. The costs might exceed the benefits. Foreign banks affect the financial stability of the host country via different channels, which are explained in this literature review. Financial development and a higher level of competition in the banking sector are important channels. However, the relationship between those channels and financial stability is not clear since existing studies in this field do not provide an univocal explanation and mechanism. Besides that, foreign banks can stabilize or destabilize the credit supply in the host country.
The effect of financial development on financial stability
The host country might develop their financial sector thanks to the entering of foreign banks. An efficient and well-functioning banking sector facilitates innovation and entrepreneurship (Meierrieks, 2015). Financial development would promote financial stability for different reasons. First, foreign banks inject new capital in the host country. Households and firms get better access to capital and financial services (financial deepening). Firms are able to increase their profit through investing in more efficient production techniques. The firm’s financial position will improve, which positively influences the quality of the bank’s loan portfolio (Chen et al. 2019; Allen et al. 2011; Claessens et al. 2013; de Haas 2014). Second, foreign banks introduce new know-how, technologies, management skills and products. The new knowledge and technology spills over to domestic banks and cause an improvement in the banks’s performance and productivity. The banks become better in gathering information and monitoring the investments. They mobilize more savings and allocate them more efficiently to the best investment opportunities, which is one of their key tasks (The World Bank, 2020). The bank’s profit and the quality of the loan portfolios increase because the chance that borrowers default on their loan declines. The bank’s solvency- and liquidity ratio improves, which strengthens the financial stability in the host country (Chen et al. 2019; Chen et al 2017; Claessens et al. 2013; de Haas 2014).
Thus, financial development can be seen as one of the benefits of the entrance of foreign banks. Nevertheless, the literature acknowledges that there might be a trade-off between financial development and financial stability (Kakes & Nijskens, 2018).
7 after the bubble burst would be more severe if the bubble is fuelled by credit. The popular financial instability hypothesis of Minsky is based on this first point (Mehrling, 1999). In addition, the foreign banks might encourage and learn the domestic banks to take more risks through a change in their funding structure, undertake non-traditional activities and entering a foreign market (Cubillas & Gonzále, 2014). The financial development occurs through a switch from a bank-based to market-based financial system, which increases the opportunities to generate profits, but also increases the systematic risk in the banking sector. Moreover, the financial sector becomes more complicated and banks become more connected with each other (Kakes & Nijskens 2018).
Cubillas & González (2014) argue that foreign banks enlarge the number of opportunities to take risks for the domestic banks in developing countries, which increase the risks in the banking sector. Domestic banks would have higher incentives to undertake risky activities after foreign banks enter the market in the developed countries.
Financial instability hypothesis of Minsky
At the start, the economy and the financial sector are stable. The financial sector allocates capital to good innovative projects and supports the real economy. The borrowers fulfil their obligation to pay interest and the principal through the generated cashflows of the investment (hedge financing). The investments lead to higher asset prices and capital gains for the asset holders. Banks and investors are encouraged to invest in those assets and the financial sector expands. This process repeats itself. The investors gain more leverage through financing these investments with loans. The investment opportunity has diminishing returns, which implies that the credit and capital in the country should at a certain moment be allocated to other innovative projects which generate higher returns. However, the investors still expect the returns and asset price to increase and a bubble is created. The financial sector shifts from hedge financing to speculative and Ponzi finance (cashflows are not enough to fulfil the obligation to pay interest and repay) which is unsustainable since the asset price cannot rise infinitely (Mehrling, 1999). The bubble will burst at a certain point (‘the Minsky moment’). Investors are highly leveraged and have to sell their assets, which decrease the assets prices further and the investors default on their loans. The majority of the banks offered loans to investors or invested directly in these assets and will have solvency problems. The Minsky cycle is visualized in figure 1.
Figure 1, Financial instability hypothesis of Minsky
8 The effect of competition in the banking sector on financial stability
The foreign banks intensify the level of competition, which would have several advantages according to the theoretical models. First, the higher level of competition stimulates banks to invest in the new available technologies and human capital. The banks might be forced to invest and improve their productivity to remain a competitive player in the market (Chen et al. 2019; Allen et al. 2011; Chen et al. 2017; Claessens et al. 2013). Second, the moral hazard problem is mitigated due to the lower interest rates charged on loans (Ijtsma, 2017). Borrowers will undertake less risky activities because the minimum cashflow they have to generate to pay all the costs is lower. Third, banks have fewer opportunities to become ‘too big to fail (TBTF)’. The managers of TBTF banks take excessive risks because they expect that the government will bail them out in periods of distress (Ijtsma, 2017). TBTF banks might also increase the contagion risks since the power is centred around a few banks. These banks are more likely to be connected with each other (Ijtsma, 2017).
Though, the literature is inconclusive with regard to the effect of competition on financial stability. The higher level of competition can also have a destabilizing effect on the economy (Chen et al. 219; Allen et al. 2011). A more competitive financial sector generates a more fragile and unstable banking sector according to the competition-fragility hypothesis. Banks charge lower interest rates, which lowers their net interest margin and franchise value. The net interest margin is, especially in developing and emerging countries, a very important component in the profits. The banks might take more risks to generate a sufficient level of profit. They might relax the terms for loans applications, which increase the risks in their loan portfolio. The quality of the granted loans deteriorates if the borrowers are more likely to default on their loans. The domestic banks might also take more risks in order to increase their profits and finance the necessary investments to remain competitive. Chen et al. (2019) argue that the banks have more difficulties to collect the necessary information about their borrowers in this new environment, which increase the risks in the loan portfolio.
Hence, the empirical literature should determine the sign in the relationship between competition and financial stability. Anyhow, the empirical literature cannot give a conclusive answer and produces different results (Ijtsma, 2017). Some empirical studies report an inverted u-shaped relationship instead of a linear relationship. They support the competition-stability hypothesis till a certain threshold level of competition and switch afterwards to the competition-fragility hypothesis.
Chen et al. (2019) investigated “the interactive role of bank competition and foreign bank entry on the risk-taking behaviour of banks over the globe” (Chen et al, 2019: 1). They concluded that the entrance of foreign banks would affect the risk-taking behaviour via an inverted-u shaped relationship.
9 their research. Besides that, the income status and the measurement method of competition would influence the relationship.
Empirical results in the study of Yudhi et al. (2019)
Event Result
Increase in the number of foreign banks
Emerging countries:
Ratio of non-performing loans decreases after bank concentration surpasses a certain level (or when market power is sufficiently weak) Advanced countries:
Ratio of non-performing loans decreases until a certain level of bank concentration (or when market power is sufficiently weak)
Increase in the share of foreign bank asset
Increase in non-performing loans
Do foreign banks stabilize or destabilize the credit supply?
Foreign banks can stabilize the credit supply in the host country because they are capable to deal with shocks in the domestic economy (Chen et al. 2019; Chen et al 2017 Claessens et al. 2013; de Haas, 2014). They would not contract their lending activity because they can fund these loans with foreign capital. Furthermore, they have a more diversified portfolio. A more diversified portfolio means that the profit is not only dependent on the returns in the host country. The domestic banks also receive more opportunities to diversify their portfolio through buying foreign assets.
10 Other costs and benefits of foreign bank penetration
Foreign banks might encourage and pressure the host country’s government to improve the regulation and supervision in the financial sector (Chen et al; 2017; Claessens et al. 2013), which induce more transparency. The government impose regulation in the financial sector to make financial firms more risk averse and enhance financial stability. A proper regulatory framework with capital requirements minimizes the costs associated with foreign banks that enter the country and prevent excessive risk-taking behaviour (Chen et al. 2019; Chen et al. 2017). Countries should allow foreign banks in the financial sector in such a way that the potential benefits are maximized and the costs minimized (Allen et all, 2011).
There would be two other costs associated with foreign banks, which deteriorate financial stability:
The foreign banks focus on the most profitable and transparent clients, which are in general the largest firms (cherry-picking). They are able to attract these clients with lower prices and a more diverse assortment. The domestic banks lose these customers and have to serve the non-transparent small and medium sized firms. The banks have to use soft data to assess their creditworthiness. Soft information is not public and is more difficult to acquire and assess. The quality of the domestic bank’s loan portfolio will deteriorate if they have difficulties in assessing the creditworthiness of these firms and a higher share of the borrowers is more likely to default (Chen et al. 2019; Yingkai Yin et al 215; Chen et al. 2017; Claessens et al. 2013; de Haas 2014). Besides that, the domestic banks might lose deposit money because depositors consider the foreign banks as a safer place to store their savings. The deposits are used to finance the lending activities. The banks have to accept the deposits at a higher interest rate or have to search for more expensive alternatives, which is more hazardous. The domestic banks might take more risks in their business operation due to the higher costs (Chen et al. 2019; Yingkai Yin et al 215; Chen et al. 2017; Claessens et al. 2013; de Haas 2014).
11 The role of institutions
Institutions are “systems of established and embedded social rules that structure social interactions” (Hodgson, 2006:18). A country has different types of institutions and some of these institutions might be important in this study.
The government implements policies to organize and provide services, which are crucial to all the citizens. The overall infrastructure, educational system and health system are important examples. The quality of the government and the public services would be important to ensure the spill over of know-how from foreign banks to domestic banks. The domestic bank’s employees would not have the required skills, educational level and devices to learn from the foreign banks and implement the more advanced technology effectively if the quality of the public services is too low (Borensztein, De Gregorio & Lee, 1998). Besides that, domestic banks might experience a big gap between the current technology and the technology used by foreign banks. They have to incur high costs to bridge this gap and the returns will be cashed in the long run. They might take high risks to generate enough revenues in the short run to prevent the foreign banks from destroying them. The competition-fragility hypothesis is more pronounced in these countries. Chen et al (2017) acknowledge the importance of institutions and argue that the risk of foreign banks is more pronounced in countries with a less productive banking sector.
Moreover, the foreign banks introduce technologies which are more based on hard data, information gathering and monitoring. The host country must be able to implement the new technologies. A stable environment that provides certainty is important to be able to make predictions and develop hard data. Property rights, a good legal system and political stability should ensure that contracts can be enforced, investments can be monitored and predictions can be made.
Claessend & van Horen (2013) confirm these points and argue that domestic banks are more likely to go bankrupt through the entry of foreign banks if the institutions are weak.
In addition, a proper regulatory framework with capital requirements minimizes the costs associated with foreign bank that enter the country and prevents excessive risk-taking behaviour (Chen et al. 2019; Chen et al. 2017). Though, the regulation would only be effective if the legal system can ensure that banks are treated equal, stick to the regulation and are punished if they violate the rules (Boujelben, Essid & Plihon, 2014). Therefore, an good and independent legal system is important.
12 The hypotheses
Hypothesis 1
Developing countries and emerging countries have in general an underdeveloped financial sector and a low level of competition in the banking sector. Foreign banks would have a positive effect on financial stability. However, the quality of the institutions will probably be too low in the majority of the countries to be able to learn from the foreign banks. Therefore, foreign banks will have a negative effect on financial stability in developing and emerging countries.
Hypothesis 2
13
3. Methodology
The sample
This paper examines the effect of foreign banks on financial stability in developing and emerging countries because the banking sector changed drastically after foreign banks entered the market in these countries. The foreign banks play a more important role in the financial sector in these countries since a high number of foreign banks entered the market after the host country liberalised the financial sector (Wu et al., 2017). The World Bank classifies countries as low-income, lower-middle income, upper-middle income or high-income and this classification is widely used in research. The World Bank regards only countries in the high-income category as developed countries. Therefore, the countries in the first three categories are included in the sample. The institutional quality is an important variable in the analysis and the databases that measure the institutional quality start in 1996 or are more complete after 1996. Therefore, the chosen time frame is 1996 till 2017.
The statistical method
The relationship between foreign banks and financial stability is analysed with a panel data model. There are two different types of panel data models, namely the static and the dynamic model. The dynamic model is used if the desired outcome is not reached immediately because adjustment is expensive and/or requires time. The dependent variable is included as a lagged explanatory variable in the dynamic model.
The dynamic model is used to test the hypothesis because domestic banks would not be able to adjust immediately to the new environment in which they have to compete with foreign banks. The adoption of the new technology is costly and banks need time implement the new technologies in an efficient way, which is discussed in the section other costs and benefits of foreign bank penetration in the literature review. The systematic GMM estimator is chosen because this estimator is suitable for datasets with a few time periods and many different countries. Besides that, this estimator controls for unobserved heterogeneity between countries. Time-fixed effects are included to control for shocks that affect all countries in the sample in the same way.
The main explanatory variable is the degree of foreign bank presence in the country. The model will include several interactions. The degree of foreign bank presence will be interacted separately with the quality of the rule of law, public services/policies and the degree of political stability. These interactions are included because these institutions would influence the relationship between the foreign banks and financial stability according to the second hypothesis.
14 latter is called foreign bank assets in the regression and the results. The methods control for differences in the size of the sector between countries.
The dynamic panel data model has some additional assumptions compared to the static panel data model. The model is not valid if these assumptions are violated. The static panel data model should be used if the lagged dependent variable is not significant and/or the additional assumptions are violated. Therefore, the static panel data model is also included in the analysis. The static panel data model can be estimated with the fixed effect estimator and the random effect estimator. The fixed effect estimator is chosen because this estimator controls for unobserved heterogeneity. Time-fixed effects are included to control for shocks that affect all countries in the sample in the same way.
Control variables
Financial instability refers to a situation in which many firms in the financial sector, which are especially banks in the developing and emerging countries, fail or are in distress. This section discusses the control variables that are used in the regression.
Institutional quality
The regression includes interaction variables between the degree of foreign bank presence and the quality of the rule of law, public services/policies and the degree of political stability, which is explained in the statistical methods. These institutions should than also be included as control variables.
Boujelbene et al. (2014) argue that banks would be able to better manage risks and accomplish higher revenues in countries with more powerful institutions that provide stability and certainty. The rule of law, political stability and voice & accountability would be especially important. Banks take excessive risks in a weak institutional environment, which triggers a financial crisis. Therefore, the voice & accountability is also important to include as control variable.
Foreign banks have to decide in which country they want to settle and in which way they enter the country. They would base this decision on the quality of the institutions in the country (Meesters, 2009). Countries with better institutions would attract the more efficient banks (Meesters, 2009). Besides that, weak institutions might be more disastrous for foreign banks than for domestic banks (Meesters, 2009).
GDP growth
The increase in the Gross Domestic Product (GDP) means that there are more products/services produced in the real economy, which results in higher profits for business owners. Borrowers are less likely to default, which benefits the Bank’s performance and loan portfolio (Alsamara, Barkat, Jarallah & Mrabet, 2019). GDP growth would strengthen financial stability.
Sovereign default
15 devaluates the government bonds and can lead to significant losses and funding problems. The sovereign debt crisis could flow into a bank crisis, which deteriorates financial stability (Grauwe, de. 2018: 115; Chen et al., 2017).
Financial depth
The relationship between the level of financial depth, which reflects the importance of the financial sector in the economy, and financial stability would be quadratic according to the literature (Kakes & Nijskens, 2018). Financial development would benefit financial stability at the start and damage it after the financial depth reached a certain point because the financial sector is inherently unstable. Therefore, the variable is included as a linear term and a quadratic term in the regression.
Expansionary monetary policy
Banks deposit their money at the central bank and have the opportunity to borrow from them at a certain interest rate. The central bank sets this interest rate at a lower level in an expansionary monetary policy. They try to influence the (long term) interest rates in the market. The lower interest rate in the market would make banks less risk averse and/or encourage them to invest in projects with higher potential returns and higher risks (Chen et al. 2017). An expansionary monetary policy would deteriorate financial stability.
Inflation
According to the traditional theory, inflation deteriorates financial stability (Issing, 2003). Inflation leads to more uncertainty about future returns and to a more volatile business environment. Furthermore, inflation exacerbates the asymmetric information issue between the bank and the borrower. Therefore, banks are exposed to more risks. Besides that, a business cycle boom combined with high inflation might trigger over-investment and asset price bubbles. However, the recently developed ‘new environment hypothesis’ argues the opposite (Issing, 2003). Central banks want to achieve price stability through controlling the inflation rate. The central bank can generate a vision that is too optimistic about the future. The predicted asset prices might be too low or high, which deteriorates financial stability in the future.
Financial regulation and supervision
Financial stability is negatively affected if banks start to take high risks in their business model. Regulation in the financial sector can prohibit banks to take those excessive risks . Regulation on the bank activities and capital restrict the opportunities to undertake risky activities, which strengthens financial stability.
Size of banks
16 The liquidity position of the banks
Banks with a lot of liquid assets have a buffer and are less likely to go bankrupt. They are less vulnerable for monetary shocks and shocks in the domestic economy that induce a bank deposit run (Chen et al. 2017). Countries that have banks with a better liquidity rates would have a higher degree of financial stability.
The capital position of banks
Banks with more capital are more solvent and have a lower probability to go bankrupt (Chen et al. 2017). Countries that have banks with more capital will have more financial stability. The degree that banks diversify their sources of income
A bank with a higher level of income diversification generates profit via different channels and is not dependent on one source of income. Therefore, these banks have a lower chance to go bankrupt. However, the contagion effect might be more pronounced when the banks diversify through buying foreign assets (Chen et al. 2017). Therefore, income diversification can benefit or harm financial stability.
The models
The dynamic panel data model is based on the following regression equation:
FinancialStabilityit =
c
i +δ
financialStabilityi,t-1 + β1foreignBanksit +β2
RuleOfLawit +β3
PublicServicesPoliciesit +β4
PoliticalStabilityit + β5foreignBanksit*RuleOfLawit +β6ForeignBanksit*PublicServicesPoliciesit + β7ForeignBanksit*PoliticalStabilityit + β8GDPit + β9SovereignDefaultit + β10FinancialDepthit + β11FinancialDepthit2 + β12MonetaryPolicyit + β13Inflationit + β14FinancialRegulationit + β15Supervisionit + β16BankSizeit +
β17BankLiquidityit + β18BankCapitalit + β19IncomeDiversificationit + dt + uit
The subscript i denotes the specific country and t the specific time period. Ci are the country-specific aspects, dt the time-fixed effects and uit the usual disturbance (error term).
Model 1: Dynamic panel data model with the system GMM estimator. The degree of foreign bank presence measured as the number of foreign banks to the total amount of banks.
Model 2: Dynamic panel data model with the system GMM estimator. The degree of foreign bank presence measured as the number of bank assets held by foreign banks to the total number of bank assets.
The static panel data model (fixed effect estimator) is based on the following regression equation:
FinancialStabilityit =
αi
+ β1foreignBanksit + β2RuleOfLawit + β3PublicServicesPoliciesit + β4PoliticalStabilityit + β5foreignBanksit*RuleOfLawit +17 + β13Inflationit + β14FinancialRegulationit + β15Supervisionit + β16BankSizeit +
β17BankLiquidityit + β18BankCapitalit + β19IncomeDiversificationit + dt + uit
αi
is the fixed effect for each country to control for heterogeneity.Model 3: Static panel data model with fixed effect estimator. The degree of foreign bank presence measured as the number of foreign banks to the total number of banks.
18
4. Data
First, the measurement method of the variables is discussed. Second, the descriptive statistics of the data and the outliers are discussed.
4.1 Data
This section describes which methods are used to measure the variables. Financial stability
Financial instability refers to a situation in which many firms in the financial sector, which are especially banks in the developing and emerging countries, fail or are in distress. The variable that measures the financial stability should focus on bank’s solvency ratio and the systematic risk in the financial sector. However, to my knowledge researchers did not succeed in developing a database that adequately measures the systematic risk in the financial sector and contains a sufficient amount of observations. Therefore, this study measures financial stability with a variable that focusses on the individual banks. The variable measures the bank’s risk to get insolvent and bankrupt. The most popular measurement methods in the empirical literature are the z-score and the ratio of non-performing loans as a percentage of total loans.
The Z-score represents the probability that the banks become insolvent and get bankrupt. A higher Z-score means that the bank has a better financial position and a lower risk of bankruptcy. The Z-score is calculated with the following formula:
Z = ((ROA + E/A)/σ (ROA))
ROA indicates the return on the assets, E/A the equity to asset ratio and σ (ROA) the standard deviation of the ROA.
The Z-score is used to measure the degree of financial stability because the score is based on information about the earnings of banks, their capital buffer and the volatility of the earnings (Ijtsma, 2017). The NPL ratio is not taken into account because it represents the bank’s credit risk and not the risk of insolvency (Ijtsma, 2017). A bank with a high NPL ratio might not have a high insolvency risk if the capital buffer is sufficient.
The Z-score has a few drawbacks. The Z-score assumes that returns are normally distributed, ignores the timing of the returns and is determined with historical data (Ijtsma, 2017). However, to my knowledge researchers did not develop another database with fewer drawbacks that provides enough information to make statements about financial stability.
Quality of institutions
19 The World Bank developed an index with different dimension to measure the quality of the institutions worldwide. This index is widely used in research. The dimensions voice & accountability, political stability & absence of violence, government effectiveness and the rule of law include the aspects of the institutional framework that are important for financial stability. Therefore, this index is used to measure the institutional quality. Each country receives a score which ranges from -2.5 (weak) to 2.5 (strong).
The dimension voice & accountability is defined as:
“Voice and accountability capture perceptions of the extent to which a country's citizens are able to participate in selecting their government, as well as freedom of expression, freedom of association, and a free media” (World Wide Governance Indicators, 2020).
The dimension political stability & absence of violence is defined as:
“Political Stability and Absence of Violence/Terrorism measures perceptions of the likelihood of political instability and/or politically motivated violence, including terrorism” (World Wide Governance Indicators, 2020).
The dimension government effectiveness is defined as:
“Government effectiveness captures the perceptions of the quality of public services, the quality of civil services and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government’s commitment to such policies” (World Wide Governance Indicators, 2020)
The dimension rule of law is defined as:
“Rule of law captures the perceptions of the extent to which agents have confidence in and abide by the rules of society, and in particular the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence” (World Wide Governance Indicators, 2020)
The first dimension will only be included as control variable. The other dimensions will be separately included as interaction term with the degree of foreign bank penetration in the country and as control variables.
GDP growth
The GDP reflects the goods and services produced in a country. The GDP growth is measured as the real GDP per capita growth. This measurement method is used because it controls for price changes and for differences in the size of the population.
Sovereign default
20 Financial depth
The expansion of the financial sector would benefit the financial stability only till a certain level of financial depth. The credit provided by the financial sector as percentage of GDP is used to measure the financial depth since it captures the importance of the financial sector in the economy. The GDP indicates the size of the economy.
Monetary policy
The central bank tried to influence the (long term) interest rate in the market especially through changing the interest rate that they charge on the banks. They decrease the interest rate if they want to conduct an expansionary monetary policy. Therefore, the real interest rate is used as a proxy for the monetary policy. The real interest rate is adjusted for inflation.
Financial regulation and supervision
Financial regulation would reduce the risks in the financial sector and benefit financial stability. The government could obligate banks to hold a certain amount of capital to their assets, which would be an effective approach to influence the bank’s risk behaviour according to the literature. Therefore, the magnitude of the capital banks have to hold (the regulatory capital) to the risk-weighted assets is used as proxy for the financial regulation in the country.
The World Bank warns that comparison of regulatory capital among countries is not perfect because the national accounting system and taxation differs. However, to my knowledge there is no other suitable measurement method.
Supervision and discipline would also be important to make sure that the banks stick to the regulation. However, it is difficult to measure the quality of the supervision and make a reliable comparison among countries. The World Bank conducted a survey to measure this and compared countries in the years 2001, 2003, 2007 and 2019. However, the sample differs and not all the questions are the same in all these years. The question “does the law establish pre-determined levels of solvency deterioration which force automatic actions such as intervention?” is consistent across the surveys. This question indicates the degree of discipline in the countries and is used to measure the degree of discipline/supervision. The survey published in 2019 is based on the period 2011 till 2016 and is included in the year 2017. To my knowledge there is not another survey which is more suitable. A dummy variable is created, which equals 1 if the answer is yes.
Inflation
The study controls for inflation because it creates uncertainty and a more volatile business environment. The GDP deflator is used to control for inflation because this deflator is based on price changes in the whole domestic economy.
Bank size
21 method. The bank assets of the three biggest banks in the country as share of the total bank assets is used to control for the bank size.
Bank liquidity
The bank liquidity is important because it influences the ability of the bank to deal with monetary/domestic shocks which might induce a deposit run. Therefore, the liquid assets of a bank divided by their short-term funding (including the deposits) is used to control for differences in liquidity between banks.
Bank capitalization
A bank with a higher degree of capital has a lower probability to go bankrupt. The capital of the bank divided by their total assets is used to control for the differences in capital between banks.
Income diversification
22
4.2 Descriptive statistics
This chapter discusses the statistics of the included variables, which are reported in table 1. Besides that, outliers in the data are discussed on the basis of boxplots and scatterplots. These boxplots and scatterplots are included in appendix 1. Outliers are removed from the dataset because they can influence the analysis and bias the results in the regression.
Table 1, descriptive statistics
Note: Dummy discipline is the answer on the question: does the law establish pre-determined levels of solvency deterioration which force automatic actions such as intervention?
The majority of the data is quite diversified judged by the high standard deviations. This is not a surprising discovery since the sample contains developing and emerging countries. These countries have in general a more volatile economy.
Variable Observations Mean Standard
deviation Minimum Maximum
Z-score 2783 12.79 8.77 0.04 96.68
Foreign bank assets as % of total bank assets
835 41.96 32.30 0 100
Foreign Banks as % of total banks
1789 38.48 26.35 0 100
Annual GDP per capita growth (%)
2949 2.85 6.40 -62.38 140.37
Financial depth (domestic credit as % of GDP)
2769 40.68 34.16 -114.69 216.91 Real interest rate (%) 2204 8.90 32.36 -93.51 1158.02
Inflation rate 3216 12.27 101.85 -31.57 4800.53
Dummy sovereign default 2668 0.006 0.08 0 1
Size (assets of 3 biggest banks as % of total assets)
2296 69.50 20.44 17.16 100
Liquidity (liquid assets as % of deposits & short-term funding)
2809 39.03 22.91 1.41 240.61
Capitalization (capital as % of
total assets) 1323 11.01 3.97 1.49 30.6
Income diversification (non-interest income as % of total income)
2749 38.80 15.38 0.03 95.34
Regulatory capital to risk weighted assets
1348 17.41 5.85 1.75 48.6
Dummy discipline 432 0.69 0.46 0 1
Government effectiveness 2510 -0.44 0.64 -2.48 1.46
Rule of law 2522 -0.48 0.68 -2.26 1.56
Voice and accountability 2780 -0.30 0.87 -2.26 1.37
23 The Z score, which is the dependent variable, is diversified with a minimum of 0.04 and a maximum of 96.68. The boxplot, which is included in appendix 1.1, demonstrates that the score of 96.68 is an outlier. South-Africa achieved a Z-score of 96.68 in 1998. Their Z-score was 19.69 in 1997 and 43.41 in 1999. The maximum is 63.40% after the outlier is removed. The two main explanatory variables are bank assets held by foreign banks and foreign banks as percentage of the total. The database of the latter contains much more observations compared to the former database. The histograms, which are included in appendix 1.2, confirms the high variation in the observations. The average share of foreign banks is 38.48% and average share of bank assets held by foreign banks is 41.96%. Those averages indicate that the domestic banks still own the majority of the assets in an average country.
The average GDP per capita growth is 2.85%, which is positive. The box plot, which is included in appendix 1.3, confirms the presence of three positive and two negative outliers. Equatorial Gunea reports 140.37% in 1997, Lybia 121.78% in 2012 and Bosnia & Herzegovina 92.20% in 1996. Lybia reports -62.38% in 2011 and South Sudan -47.59% in 2012. Libya had a rough year in 2011 with a NATO intervention and the death of the dictator, which explains the high negative GDP growth in 2011 (Human Right Watch, 2012). A recovery of the economy or aid may explain the high GDP growth in 2012. The independence of South Sudan in 2011 had a huge impact on the economy and explains the high decline of the economy in 2012 (Tran, 2012). Detection of huge oil reserves in 1996 is responsible for the extreme high GDP growth in Equatorial Gunea (OPEC, 2020). America started to support the economy of Bosnia & Herzegovina with aid in 1996 (USAID, 2018). The received aid may explain the high GDP per capita growth. The minimum is -36.56% and the maximum 60.42% after the removal of the outliers. The average GDP per capita growth is still positive and becomes 2.77%.
The credit as % of GDP is negative in some countries according to table 1. The GDP represent the value of the goods and services produced in the economy and can not be negative. Therefore, the numbers should not be negative. The negative numbers are removed and result in a new minimum of 0.21%. The boxplot and scatterplot based on this variable is included in appendix 1.4. These graphs are based on the situation after the removal of the negative values. The graphs indicate that the two highest observations (216.91% and 216.21%) are outliers. These outliers are removed, which result in a maximum of 202.88%.
The highest interest rate in the sample is 1158.02%, which is an unrealistic high number. The boxplot, which is included in appendix 1.5, confirms that there are five outliers. Zimbabwe reported in 2004 until 2008 respectively 252.12%, 219.12%, 508.74%, 572.94% and 1158.02%. The real interest rate in Zimbabwe increased from 81.33% to 252.12% in 2004, which is a significant upsurge. Zimbabwe left the commonwealth in December 2003 (Evanamusoke, 2015). This decision might explain the exceptional increase in the interest rate. The scatterplot in appendix 1.5 includes the observations below 219% and indicates that the interest rate of 130.34% in Bulgaria in 1996 is also an outlier. Therefore, the 6 highest observations are removed, which result in a maximum of 94.73%.
24 indicates that the two highest observations are clear outliers. The scatterplot in appendix 1.6 provides a better image of the observations below the two highest observations. Inflation rates above the 200% are high and the scatterplot confirms that these are also outliers. Therefore, the observations above the 200% are removed in this dataset. Now, the maximum is 196.98%. The mean for the dummy variable for a sovereign default is 0.006, which indicates that the event of a sovereign debt default is rare. The pie chart in appendix 1.7 confirms this finding. The data for the bank size, the market share of the three biggest banks, does not contain outliers according to the boxplot (appendix 1.8). The banking sector is dominated by only three banks in one or more countries according to the maximum of 100%. However, there are big differences between countries as the data ranges from 17.16% to 100%.
The liquid assets to short term funding vary among banks between the 1.41% and 240.61%. However, the box plot, which is included in appendix 1.9, indicates that the three highest observations are outliers. The maximum is 180.77% after the removal of these outliers. The dataset for the capital to the total assets contains less observations compared to the other bank specific variables. The number of observations (1323) is still enough for a reliable analysis. The variation in the data is smaller compared to the other variables and the average of 11.01% is low. The boxplot, which is included in appendix 1.10, indicates that there are three outliers in the data. However, the scatterplot in appendix 1.10 reveals that the three highest observations are not much higher compared to the other points. Therefore, these observations are not removed.
The average non-interest income to the total income is 38.80%, which indicates that the degree of income diversification is not high in this sample. However, the minimum of 0.03 and maximum 95.34 indicates that there are many differences between banks. The box plot and scatterplot, which are included in appendix 1.11, indicates that these observations are not outliers.
The regulatory capital to risk weighted assets ranges from 1.75% to 48.6%. This database contains less observations compared to the other databases. There are still enough observations for the analysis. The data would contain one outlier according to the boxplot in appendix 1.12. However, the scatterplot in this appendix reveals that the highest observation is only 3% higher compared to the second highest observation. Therefore, the maximum is not removed in the dataset. The dummy discipline is collected via a survey that is conducted in 5 specific years in the period 1996 - 20017. Therefore, this variable has much fewer observations.
25
5. Analysis and results
The panel data model makes several assumptions. Multiple tests are performed to ensure that those assumptions are not violated. Therefore, this chapter starts with the discussion of those tests.
The variables should not be correlated with each other to an extend that is problematic. The correlation between the variables is visualized in table 2. A correlation score of 1 implies perfect correlation. The correlation score should not be too close to 1. The percentage foreign banks and bank assets held by foreign banks are used separately in different models.
Therefore, the high degree of correlation between those variables and some interactions which contain both these variables is not a problem. Moreover, the bank capitalization and the regulatory capital have a high correlation of -0.75. Countries are more likely to demand a lower level of capital to risk-weighted assets if the banks have more capital to their total assets. The high degree of correlation between those variables might bias the estimates in the regression. Therefore, the variable regulatory capital is dropped as independent variable. In addition, the variables interest and inflation have a correlation score of 0.53. The interest rate is the real interest rate, which is adjusted for inflation. Therefore, the degree of correlation between the inflation rate and interest rate is not problematic. The rule of law and government effectiveness have a high correlation score of -0.66. Consequently, the correlation between the interaction terms ForeignBanks*RuleOfLaw & ForeignBanks*GovernmentEffectivenes and the interaction terms ForeignBankAssets*RuleOfLaw &
26 Table 2, correlation matrix
27 The F test for the year dummies indicates that the year dummies are insignificant in all four models. Therefore, time-fixed effects are not included in the regressions. The results are included in appendix 2.2.
The GMM estimation, which is used in the first two models, is biased if autocorrelation is present and/or the instrumental variables are correlated with the error term. The Arellano-Bond test (appendix 2.3) detects autocorrelation and the Sargan test (appendix 2.4) the validity of the instrumental variables. The Arellano-Bond test examines if autocorrelation is present in the first-order and/or second-order. The Arellano-bond test only rejects the null hypothesis, which states that there is no autocorrelation, in the first order. Anyhow, first-order autocorrelation is not problematic and will not bias the results. The Sargan test is not robust for heteroskedasticity and tends to over reject the null hypothesis, which states that the instrumental variables are uncorrelated to the error term, in samples with heteroskedasticity. The sargan test is executed after a two-step estimation to solve this problem (Pinzon, 2015). The null hypothesis is not rejected in both models.
The linear relationship is more pronounced if the presence of foreign banks is measured as the share of bank assets held by foreign banks according to the scatterplots in figure 2. Surprisingly, the interaction term between foreign bank presence and political stability has a negative relationship with financial stability. Though, the regression with the control variables will reveal if these effects are significant.
Figure 2, scatterplots of the main explanatory variables
28 Regression analysis
The dummy for bank discipline is dropped as explanatory variable because the variable contains too little observations. The dynamic panel data model can only be estimated if this dummy variable is not included. This chapter discusses the estimated coefficients in the models. The results are shown in table 3.
Table 3, Regression results with robust standard errors (in brackets the p-value) Dependent variable: Z-score (higher z score → higher level of financial stability) Significance levels: * P- value ≤ 0,10 // ** P-value ≤ 0,05
Model 1 (gmm estimator) Model 2 (gmm estimator) Model 3 (Fixed effect estimator) Model 4 (Fixed effect estimator) Z score lagged 1 year 0.75 ** (0.000) 0.86 ** (0.000)
Foreign banks -0.07 ** (0.047) 0.09 * (0.097)
Foreign bank assets 0.01 (0.39) -0.03 (0.25)
Government effectiveness -0.92 (0.80) 1.67 (0.57) -1.94 (0.36) 0.83 (0.74) Voice & accountability -1.57 (0.21) -0.21 (0.78) -0.26 (0.87) -0.79 (0.57) Political stability 0.75 (0.70) -2.72 (0.20) -1.41 (0.36) -1.32 (0.30) Foreign banks * government
effectiveness
0.03 (0.63) 0.04 (0.41)
Foreign banks * political stability
-0.05 (0.15) 0.009 (0.72)
Foreign bank Assets * government effectiveness
0.001 (0.98) -0.01 (0.74) Foreign bank Assets * political
stability
0.02 (0.52) 0.006 (0.73)
GDP growth 0.04 (0.20) 0.06 * (0.08) 0.06 (0.18) 0.04 (0.18) Financial depth (domestic credit
as % of GDP)
-0.03 (0.54) -0.01 (0.75) -0.06 (0.28) -0.10 ** (0.03) Financial depth2 0.0006 (0.20) 0.0002 (0.32) 0.0004 (0.43) 0.0006 * (0.098) Interest -0.01 (0.83) 0.02 (0.68) -0.04 (0.22) 0.02 (0.63) Inflation -0.001 (0.98) 0.03 (0.40) -0.009 (0.22) 0.0099 (0.78) Dummy sovereign default -3.81 ** (0.02) -1.17 (0.30) -1.05 * (0.06) -1.39 * (0.085) Bank size -0.007 (0.73) 0.02 (0.43) -0.02 (0.55) -0.04 ** (0.03) Bank liquidity -0.21 (0.37) -0.03 (0.43) -0.05 ** (0.01) -0.06 ** (0.001) Bank capitalization 0.28 ** (0.05) 0.22 (0.17) 0.62 * (0.000) -0.01 ** (0.006) Banks’ income diversification -0.01 (0.38) 0.03 (0.22) -0.005 (0.77) -0.01 (0.57) Constant 3.93 (0.15) -4.25 (0.26) 5.38 (0.23) 15.06 ** (0.000)
Observations 625 458 627 84
Countries 62 61 62 48
F-statistic (Prob > F) 0.0000 0.0000 0.0000 0.0000
29 The test results, which are discussed at the start of this chapter, demonstrate that the assumptions of the dynamic data panel model are not violated. Besides that, the coefficient for the lagged z-score is significantly different from zero. Therefore, the GMM estimator, which is used in the first and second model, is valid. The third and fourth model omit the significant lagged dependent variable, which is correlated with the main explanatory variable. Therefore, the estimates in those models are biased. The estimates in the first two models are used in this paper. Model 1 has a bit more observations because the database for the share of foreign banks is more complete compared to the dataset of bank assets owned by foreign banks.
The coefficient, which is a δ, for the z-score is significantly different from zero. The z-score adjusts not immediately to the new situation and environment. The coefficient is equal to 1 minus the speed of adjustment. The speed of adjustment is zero if there is no adjustment in the ongoing term and one if the z-score adjusts immediately. The speed of adjustment in model 1 and 2 is respectively 0.25 and 0.14, which is low.
The presence of foreign banks in the country, which is the main explanatory variable, is measured differently in both models. An increase in bank assets held by foreign banks would not affect the z-score according to the second model. An increase in the number of foreign banks relative to the total number of banks has a negative impact on the z-score. This result is in line with the first hypothesis. A ten percent increase in the share of foreign banks in the country will lead to a decrease of 0.70 in the z-score in the short run. This is a small effect since the average z-score is 12.79. However, it will lead to a decline of 2.8 in the z-score in the long run, which is calculated as 0.70/0.25. The competition-fragility hypothesis would be relevant in the sampled countries. The domestic banks are not able to adjust to the new competitive environment and implement the new technologies efficiently. The result does not support the statement in the study of Chen et al. (2019 & 2017), which is that foreign banks deteriorate the financial stability in the short run and strengthen it in the long run.
Host countries are more likely to achieve a higher degree of financial stability after foreign banks enter the market if the quality of their rule of law, public services and policies is higher and if they experience political stability according to the second hypothesis. These countries would be likely to benefit from the new know-how and adapt to the new more competitive environment. Nonetheless, this hypothesis is not supported by the empirical analysis since the coefficient for the interaction terms between the foreign banks and the relevant institutions is insignificant. Changes in the level of competition have the same effect on the degree of financial stability in countries with strong institutions compared to countries with poor and fragile institutions.
30 percent increase in GDP per capita growth will lead to an increase of 0.06 in the z-score in the short-run.
31
6. Conclusion
The research question in this paper is ‘What is the effect of foreign banks on financial stability in developing and emerging countries?’ The literature review developed two different hypotheses. According to the first hypothesis, foreign banks have a negative effect on financial stability in developing and emerging countries. However, according to the second hypothesis the influence of foreign banks on financial stability is dependent on the quality of certain institutions. The quality of the rule of law, public services, policies and the degree of political stability is important. The negative relationship between foreign banks and financial stability declines as the quality of these institutions increase. The dynamic panel data model with the system GMM estimator is used to test the hypotheses. The presence of foreign banks in the host country is measured in two different ways, namely percentage of foreign banks to the total banks (model 1) and the percentage bank assets held by foreign banks to the total number of bank assets (model 2).
Foreign banks only influence the degree of financial stability if they are operating as a bank and increase the host country’s share of foreign banks to the total number of banks according to the empirical analysis. The effect is in line with the first hypothesis. The effect is different in the short run compared to the long run, which is respectively a decline of 0.70 and 2.80 in the z-score if the share of foreign banks increase with 10%. The domestic banks would not be able to adjust to the new competitive environment and remain competitive, which implies that the competition-fragility hypothesis would be relevant in the sampled countries. An increase in the percentage of bank assets held by foreign banks will not influence the degree of financial stability according to the empirical analysis. Both models do not support the second hypothesis since the coefficient for the interaction terms between foreign banks and the relevant institutions are insignificant.
The different results in model 1 and 2 might indicate that the entry mode of the foreign bank is relevant. But, the significance level of the control variables differs in both models, which might point towards another explanation. Currently, the government is probably unable to prevent that foreign banks engage in the domestic financial market if they liberalize the sector. Globalization advances and/or the economy needs the foreign bank’s capital. Nevertheless, the government should restrict the number of foreign banks that engage in the banking sector since foreign banks would deteriorate financial stability. Besides that, the government should decide in which way the foreign bank is able to engage in the domestic financial sector and control their behaviour. They can require the foreign bank to cooperate with a domestic bank. This form of cooperation would have a higher effect on the share of banks assets held by foreign banks than the share of foreign banks in the economy. The domestic banks would also be better able to learn from the foreign banks, which is an important point in the literature review. To conclude, according to this empirical analysis the government has to be aware that the effect of foreign banks on the degree of financial stability differs in the short-run and the long-run. They have to keep this in mind if they evaluate the behaviour of active foreign banks in the country.
33
7. References
Alsamara, M., Barkat, K., Jarallah, S., & Mrabet, Z. 2019. The switching impact of financial stability and economic growth in Qatar: Evidence from an oil-rich country. The Quarterly Review of Economics and Finance, 73: 205-2016
Agoraki, M.-E.K. , Delis, M.D., & Pasiour, F. 2011. Regulations, competition and bank risk-taking in transition countries. Journal of financial stability, 7: 38-48
Allen, F., Beck, T., Carletti, E., Lane, F.R., Schoenmaker, D. & Wagner, W. 2011. Cross-border banking in Europe: implications for financial stability and macroeconomic policies. London: Centre for Economy Policy Research
April Yudhi, W.S., Natzir, M., Soedarmono W., Trinugroho, I. & Warokka, A. 2019. Foreign penetration, competition and credit risk in banking. Borsa Istanbul Review, 19(3): 249-257 Usaid: Fact sheet: Usaid assistance in Bosnia and Herzegovina (1996-present); [last updated at May 22, 2018]. https://www.usaid.gov/bosnia/fact-sheets/usaid-assistance-bosnia-and-herzegovina
Borensztein, E., De Gregorio, J., & Lee, J. W. 1998. How does foreign investment affect growth? Journal of International Economics, 45(1), 115–135
Boujelbene, Y., Essid, Z., & Plihon. 2014. Institutional quality and bank instability: cross-countries evidence in emerging cross-countries. MPRA Paper 56251, University Library of Munich, Germany
Chen, M., Jeon, B.N., Wang, R., & Wu, J. 2017. Does foreign bank penetration affect the risk of domestic banks? Evidence from emerging countries. Journal of financial stability, 31: 45-61
Chen, S.; Nazir, M.I.; Hashmi, S.H.; Shaikh, R. 2019. Bank Competition, Foreign Bank Entry, and Risk-Taking Behavior: Cross Country Evidence. J. Risk Financial Manag, 12(3), 106.
Claessens, S. & Horen van, N. 2013. Impact of Foreign Banks. DNB Working Paper No. 370 Cubillas, E., & González F. 2014. Financial liberalization and bank-risk taking: International evidence. Journal of financial stability, 11(1): 32-48
Evanamusoke; 7th December 2003: Zimbabwe exits the Commonwealth [published at Dec 6, 2015]. https://commonwealthoralhistories.org/2015/7th-december-2003-zimbabwe-exits-the-commonwealth/
Gadanecz, B. & Jayaram, K. 2009. Measures of financial stability – a review. IFC Bulletin, 31: 365-380
34 Haas de, R. 2014. The dark and bright sides of global banking: a (somewhat) cautionary tale from emerging Europe. Working Paper No. 170
Haas de, R., & Lelyveld van, I. 2006. Foreign banks and credit stability in Central and Eastern Europe. A panel data analysis. Journal of banking and finance, 30: 1927 – 1952
Hodgson, G.M. 2006. What are institutions? Journal of economic issues, XL(1): 1-25 Human Right Watch; World Report 2012: Lybia[published in 2012].
https://www.hrw.org/world-report/2012/country-chapters/libya
Ijtsma, P.A. 2017. Financial stability, economic growth, and the banking sector. Groningen: University of Groningen, SOM research school.
Issing, O. 2003. Monetary and Financial Stability: Is there a Trade-off? Speech at Conference on "Monetary Stability, Financial Stability and theBusiness Cycle; Bank for International Settlements, Basle
Kakes, J. & Nijskens, R. 2018. Size of the banking sector: implications for financial stability. DNB Occasional Studies, 16(06): 1-56
Kulwant Rai & Herman Kamil, 2010. "The Global Credit Crunch and Foreign Banks’ Lending to Emerging Markets; Why Did Latin America Fare Better?," IMF Working Papers 10/102, International Monetary Fund.
Masci, P. 2008. Financial Liberalization, Economic Growth, Stability and Financial Market Development in Emerging Markets. Journal of the Washington Institute of China Studies, 3(3): 42-61
Meesters, A..J. 2009. Efficiency of financial institutions: a stochastic frontier analysis approach. Groningen: University of Groningen
Mehrling, P. 1999. The vision of Hyman P. Minsky. Journal of economic behaviour & organization, 39: 129-158
Meierrieks, D. 2015. Financial Development and Innovation: Is There Evidence of a
Schumpeterian Finance-Innovation Nexus? Annals of Economics and Finance, 15(2): 61-81 OPEC: Equatorial Guinea facts and figures; [accessed at 12th May 2020].
https://www.opec.org/opec_web/en/about_us/4319.htm
Pinzon, E. xtabond cheat sheet [published at November 12, 2015]. https://blog.stata.com/2015/11/12/xtabond-cheat-sheet/
The World Bank: Financial development; [accessed at 4th April 2020].
35 World Wide Governance Indicators: Voice and Accountability; [accessed at the 28th May 2020].file:///C:/Users/hilde/AppData/Local/Packages/Microsoft.MicrosoftEdge_8wekyb3d8b bwe/TempState/Downloads/va%20(1).pdf
World Wide Governance Indicators: Political Stability and Absence of Violence/Terrorism; [accessed at the 28th May 2020].
file:///C:/Users/hilde/AppData/Local/Packages/Microsoft.MicrosoftEdge_8wekyb3d8bbwe/Te mpState/Downloads/pv%20(1).pdf
World Wide Governance Indicators: Rule of law; [accessed at the 5th May 2020].
file:///C:/Users/hilde/AppData/Local/Packages/Microsoft.MicrosoftEdge_8wekyb3d8bbwe/Te mpState/Downloads/rl%20(1).pdf
World Wide Governance Indicators: Government effectiveness; [accessed at the 5th may]. file:///C:/Users/hilde/AppData/Local/Packages/Microsoft.MicrosoftEdge_8wekyb3d8bbwe/Te mpState/Downloads/ge%20(1).pdf
36
Appendix 1
1.1 Box plot of the variable Z-score
1.2 Histograms for the degree of foreign bank presence in the country
41 1.7 Pie chart for the dummy sovereign default
42 1.9 Boxplot for the variable bank liquidity
46
Appendix 2
2.1 Breusch-Pagan test
The Breush-Pagan test indicates if there is heteroskedasticity in the data. The null hypothesis state that there is no heteroskedasticity. The null hypothesis is rejected if the p-value is smaller than 0.05
Model 1 & 3: The degree of foreign bank presence measured as number of foreign banks to the total amount of banks.
H0 Constant variance
Chi2(1) 41.95
Prob > chi2 0.00000
Model 2 & 4: The degree of foreign bank presence measured as number of bank assets held by foreign banks to the total amount of bank assets.
H0 Constant variance
Chi2(1) 25.63
