Impact of Bank-Specific and Macro-Economic
Variables on Bank Performance
Master’s Thesis
University of Groningen
International Economics and Business
Vladimir Zivkovic (s1497596)
Thesis supervisor Dr. Gerrit Lanjouw
TABLE OF CONTENT
1. Introduction 3
2. Literature review 6
3. Central hypothesis 10
3.1 Hypothesis and variables 11
3.1.1 Dependent variable 11
3.1.2 Independent bank-specific variables 12
3.1.3 Macroeconomic variables 14
4. Research methodology 16
4.1 Multiple regression analysis and assumptions 17
4.2 Econometric model 19
4.3 Measurement of the dependent, independent and control variables 21
5. Data 23
6. Results 24
6.1 Descriptive statistics 24
6.2 Bivariate correlation 25
6.3 Multiple regression analysis 26
6.4 Multiple regression analysis with regional dummy variables 27
6.5 Multiple regression analysis with dummy variable (banks from developing
countries) 29
7. Conclusion 30
1. Introduction
Numerous empirical studies are trying to explain the influence of bank-specific and macro-economic factors on bank profitability. A strong and efficient banking system creates financial stability and reduces the economy’s vulnerability to unfavorable macroeconomic shocks. Many empirical studies state that overall economic achievement of a country is a positive function of its financial sector, mainly of its banking system. Current studies indicate that countries with well-developed financial institutions are likely to experience higher rates of real GDP per capita growth (Levine, 1997; Levine and Zervos, 1998). Changes in the macroeconomic factors influence banks’ performance and financial stability simultaneously. Thus it is of great importance for the governmental authorities to quantify the linkages between macroeconomic factors and the banking sector in order to maintain financial stability and high level of bank performance (Gerlach, Peng and Shu, forthcoming).
During the last decade the banking systems of many developed, developing and transition countries have experienced key transformations (deregulation and privatization processes). The high rates of bank lending have led to an increased number of non-performing credits that have affected the emergence of a banking crisis in many transition and developing countries (the Latin American countries in the middle of 1990s, the East Asian countries 1997-1998, Russia 1998). The common necessity for these countries is a stable and efficient banking system in order to finance both public and private investments (Claeys and Vander Vennet 2004). Many authors have conducted cross-country studies for countries within one region, given that these countries share similar challenges and problems in the financial and banking sector.
further deepening of the banking sector integration, and a consolidation of the European banking system that contributes to increased competition and banking efficiency.
After the collapse of the communist system in 1989, Central and Eastern European countries (CEEC) were faced with new economic, financial, political and social challenges. The Central and Eastern European countries had centrally planned economies and mono-bank systems. In centrally planned economies the government controls all major sectors of the economy. The key characteristics of mono-bank systems are: questionable lending policies and non-performing loans to large state enterprises that make this banking system very fragile and insolvent. Breaking up of the mono-bank system into two-tier system was inevitable for CEE transition countries. This has enabled them to provide a stable and competitive banking system with rational credit allocation, financial market liberalization and economic privatization of the large state banks, which will lead to the establishment of a modern financial system (Long and Rutkowska 1995, Begg, 1997). The considerable foreign banking capital inflows in the CEE countries and the creation of an efficient regulatory framework have contributed to the rapid transformation and development of their banking systems (Hollo and Nagy, 2006).
Following the banking crises of the mid-1990s, caused by the high bank lending, the banking systems in Latin America experienced financial liberalization characterized by a strong credit growth (Gelos, 2006). After the banking crisis in the mid-1990s the banking systems in Latin American countries have started to consolidate slowly with the help of regulatory tightening. This resulted in reduction of the number of banks and stronger competition. Furthermore they have lowered the entry barriers for foreign banks (Gelos, 2006).
rapidly improved since the crisis, both in terms of performance and stability (Laeven , 2005).
The countries in each of the previously mentioned regions share similar banking challenges. Therefore the main goal of this study is to make a comparison of the extent to which bank specific and macroeconomic factors affect bank profitability and, in particular, whether profitability is similar or different across the regions (West Europe, East Europe, North America, Latin America and Far and Central Asia).
A set of bank indicators are going to be used in order to explain bank performance measured by the net interest margin (NIM). This thesis examines whether independent variables (equity, loans, customer and short term funding, earning assets, overhead costs, foreign ownership and bank size) and macro economic variables (inflation and current account balance) could be accepted as explanatory variables for bank performance.
The main research question is:
Is bank profitability related to similar macro-economic and bank-specific variables across regions?
2. Literature review
This section explores the contemporary academic literature from the research field of this study; in particular the correlations between on the one hand, bank performance, and on the other, bank specific and macroeconomic variables.
In the literature, bank performance is usually expressed as a function of internal and external determinants. The internal determinants are the bank specific or micro determinants of bank profitability. The external determinants are not directly related to banks but affect the operation and performance of banks since they reflect the economic and legal environment (macroeconomic variables).
Demirguc-Kunt and Huizinga (2000) investigate the impact of bank-specific variables, financial development and structure variables on bank performance by using data for banks from developed and developing countries in the period 1990-1997. They found that banks have higher profits in underdeveloped financial systems compared to developed financial systems due to inefficiency and less competition among the banks. Higher bank development contributes to tougher competition, higher efficiency and lower profits. (Demirguc-Kunt and Huizinga 2000). Furthermore, as financial systems become developed, bank profits and margins are not significantly different across bank-based and market-based systems. On the other hand financial systems in developing countries tend to be more bank-based compared with financial systems from developed countries. They state that the greater financial and stock market development improve the efficiency of the banking sector in countries with underdeveloped financial systems. This leads to fastening the pace of growth on firm and macro-level. Another finding in their research is that financial structure does not have a significant influence on profits and interest margins.
Demirguc-Kunt et al. (2001) state that inflation on the one hand contributes to accomplish higher realized interest margins and higher profitability, but on the other hand, to increasing the banking costs. Moreover the authors have found that the corporate tax burden is fully passed on bank customers in developing as well as developed countries. Legal and institutional differences are important for bank interest margins and profitability. Improved contract enforcement, efficiency in the legal system, and lack of corruption are related with lower interest margins and profitability.
Furthermore Demirguc-Kunt et al. (2004) investigate the influence of bank regulation, market structure and national institutions on net interest margins and overhead costs. They have used a dataset for more than 1,400 banks across 72 countries. The results show that tighter regulation on bank entry and bank activities increase the cost of financial intermediation. Moreover, inflation has a high influence on bank margins and overhead expenditures, while concentration is positively related to interest margins. The bank regulations became unimportant for national determinants of economic freedom or property rights protection. But on the other hand these institutional determinants strongly explain cross-bank net interest margins and overhead costs.
Panayiotis A., Sophocles B. and Matthaios D. (2005) have observed the influence of bank-specific, industry-specific and macroeconomic determinants on bank profitability, using an empirical construction that incorporates the traditional Structure-Conduct-Performance (SCP) approach. They use a GMM technique for panel data of Greek banks for the period 1985-2001. The results indicate that all bank-specific determinants, with the exception of size, influence bank profitability in the predicted way.
than domestic banks whereas in developed countries foreign banks have smaller margins and profits than domestic banks. Macroeconomic conditions also give an explanation for interest margins and profitability. (Demirguc-Kunt and Huizinga, 1999)
Claeys and Vennet (2004) examine determinants of bank interest margins in the Central and Eastern European countries (CEEC). They evaluate to what extent the relatively high bank margins in CEEC can be explained by low efficiency and non-competitive market conditions controlling for the macroeconomic environment and the impact of foreign and state-owned banks. They analytically compare CEEC banks with Western European banks. The results show that banking in the CEEC is on a virtuous path, particularly in the EU accession countries (Czech Republic, Estonia, Hungary, etc) where increased efficiency contributes to raising the customer benefits, and capital adequacy supports systemic stability. On the other hand in the non-accession countries significant policy measures are still required such as institutional reforms, banking deregulation, etc.
There is a vast literature covering the topic of performance differences between the foreign and domestic banks, and the effect of foreign bank entry on bank performance. One such study is done by Jan Uiboupin (2004) that investigates the short-term effects of foreign banks entry on bank performance in Central and Eastern European countries. The results indicate that foreign bank entry has a negative influence on domestic banks’ revenues from interest earning assets, non-interest income and profitability.
Also Demirguc-Kunt et al (2001) investigate the relationship between foreign bank entry and the performance of the domestic banking sector in 80 countries. They have used panel estimations with observations of 7,900 banks for the period 1988–1995. The core result of the study is that foreign banks have higher profits than domestic banks in the developing countries, while in developed countries foreign banks are less profitable. Foreign bank presence in the market is strongly related with lower costs and lower net interest margins for domestic banks.
have investigated the benefits and costs of foreign ownership by analyzing determinants of profitability for domestic takeover and greenfield banks. The profitability of foreign banks is less influenced by macroeconomic conditions in their host countries. Only domestic banks earn higher profits in more concentrated banking markets, whereas takeover banks suffer from diseconomies of scale due to the fact that they have acquired large institutions.
Hanweck and Ryu (2005) have constructed a model of bank behavior that explains net interest margin changes, as a result of credit, interest-rate, and term-structure shocks, for different groups of banks for the period 1986-2003. Generally they have found that banks with different product-line specializations and asset sizes respond in different ways to these shocks. Large and more diversified banks are likely to be less responsive to interest-rate and term-structure shocks, but more responsive to credit shocks. Moreover, the structure of assets and liabilities helps to moderate the effects of changes and volatility in short-term interest rates on bank interest margins. Finally, the results indicate that banks are not able to prevent interest-rate volatility.
3. Central hypothesis
The research goal of this article is to make a comparison of the extent to which bank specific and macroeconomic factors affect bank profitability and to investigate whether these influences are similar or different across regions.
The central hypothesis in this thesis is the following:
3.1 Hypothesis and variables
3.1.1 Dependent variable
Bank profitability is going to be used as the dependent variable for measuring bank performance. Net interest margin (NIM) is going to be used for the operatioanalization of bank performance.
The dependent variable NIM is computed as interest income minus interest expense divided by the average value of interest-bearing assets for each bank.
NIM= (Interest Income-Interest Expense)/ Average Value of Interest-Bearing Assets
NIM is focused on the conventional borrowing and lending operations of banks (Abreu and Mendes 2005).
3.1.2 Independent bank-specific variables
In this study the bank-specific variables that influence bank profitability are: Equity to Total Assets ratio (E/TA), the Loans to Total Assets ratio (L/TA), Customer and Short Term funding to Total Assets ratio (CST/TA), Earning Assets to Total Assets ratio (EA/TA), Overhead Costs (OV), Foreign Ownership (FO) and Size of the Bank (BS). Based on the review of previous studies that explore the relationship between bank and macro economic variables and bank profitability, the hypotheses below are formulated.
Equity to total assets ratio
Hypothesis 1: The higher equity to total assets ratio (E/TA), the higher the bank’s profitability.
This hypothesis states that there is a positive relationship between the ratio Equity to Total Assets and the financial performance of the banks. The Equity to Total Assets ratio is defined as book equity (assets minus liabilities) divided by total assets. According to Demirguc-Kunt and Huizinga (2000), a higher E/TA lowers the bank’s need for external funding (costs of funding are reduced) and therefore increases the potential for higher profits and lowers the bankruptcy risk.
Loans to total assets ratio
Hypothesis 2: The higher ratio of loans to total assets (L/TA), the higher the bank profitability.
Customer and short term funding to total assets
Hypothesis 3: The higher customer and short term funding (deposits) to total assets ratio (STF/TA), the higher the bank profitability.
Hypothesis three explores the influence of the deposits (customer and short term funding) to bank profitability. According to Demirguc-Kunt and Huizinga (2000) customer and short term funding represent deposits of the bank. The higher amount of deposits in the bank gives the possibilities for issuing additional loans and lowering the interest cost, leading to higher profits.
Earning assets to total assets ratio
Hypothesis 4: The higher earning assets to total assets ratio (EA/TA), the higher the profitability of the bank.
Earning assets are defined as cash, real estate and other non-interest earning assets. The EA/TA ratio measures the non-commercial activities of the banks. According to Abreu and Mendes (2005) earning assets can be used as a proxy for management quality. If a bank managers are able to apply bank asset in a more efficient way (i.e. the composition of assets includes a higher proportion of earning assets), this will lead to higher net interest margin.
Overhead cost to total asset ratio
Hypothesis 5: The lower overhead costs to total assets ratio (OV/TA), the higher the bank’s profitability.
Foreign ownership
Hypothesis 6: Foreign ownership (50 % of the shares or higher) increases the bank profitability.
The foreign ownership variable equals one, if at least fifty percent of the bank’s shares are in foreign hands, and it is zero if more than fifty percent of the bank’s shares are in domestic hands. Mathieson and Roldos (2001) have suggested that foreign banks can offer a more stable supply of credits because they are the branches and subsidiaries of large international banks and have the opportunity to use funds from their parents (on average they have more diversified portfolios). Furthermore, large international banks are expected to have better access to global financial markets and the entry of foreign banks can offer cheaper credits compared to banks with a domestic ownership. According to Demirguc-Kunt et al. (2001) foreign banks accomplish reasonably high net interest margins and profitability in relatively poor countries. This is due to the fact that foreign banks are often exempted from adverse domestic banking regulations, and also they apply superior banking techniques.
Bank size
Hypothesis 7: The larger the size of the bank, the higher the bank profitability.
Size of the bank can be a significant determinant of net interest margins and bank profitability. In general, larger banks are associated with a better performance because of economies of scale and scope. Many empirical studies have generally concluded that slight economies of scale do exist in the banking industry (Demirguç-Kunt et al 2004). According to Levine (2004) larger banks may require lower overhead expenditures as a share of total assets.
3.1.3 Macroeconomic variables
the multiple regression model as macroeconomic variables. These variables are described below.
Inflation
Hypothesis 8: The higher inflation (INF), the higher is bank profitability.
According to Demirguc-Kunt et al. (2001) inflation is related to higher realized interest margins and a higher profitability. The authors have found that the profitability increases with a higher rate then the costs themselves during inflation. Hanson and Rocha (1986) also found a positive relationship between inflation and net interest margin by using data for 29 countries for the period 1975-1983.
Current account balance
Hypothesis 9: The larger the current account balance (CA), the lower the bank profitability.
4. Research methodology
4.1 Multiple regression analysis and assumptions
The multiple regression model relies on five essential underlying assumptions. Four out of five assumptions have to be confirmed in order to realize Best Linear Unbiased Estimators (BLUE). At the same time, to get more reliable statistical results it is required to perform the third assumption of homoskedasticity. In the following section we can see the assumptions.
MR1. The value of y for each value of x is Y = b0+ b1 x +b2 x2....+ εt
The first assumption, MR1, assumes linearity or a linear relationship between the dependent and the independent variables where: b0 is the intercept, and b1, b2 … bnare
slopes of the function. Further, εt is the error term that is consisted from all the factors
that have an impact on Y other than the exogenous variables included in the model.
MR2. The average value of the random error e is E (ε) = 0
The second assumption, MR2 indicates that the error ε is a random variable with a zero mean.
MR3. The variance of the random error e is var (ε) and is constant
The third assumption, MR3 indicates that errors reveal homoskedasticity as a substitute of heteroskedasticity (changeable variances). This shows that value of the variance or the error term is constant on the independent (explanatory) variables. In other terms, this shows that if there is no pattern in the distribution of the error terms around any given X, the assumption of homoskedasticity is correct. On other hand, if the third assumption is violated we still have a BLUE model, but it is no longer reliable and as a result the regression output can not be considered as consistent in conditions of test statistics. The reason for this is that the variance, which is the most important part of these statistics, is not constant anymore. The distribution of the error term (ε) for cross-sectional data should graphically indicate a regular distribution in relation to the function Y = b0 + b1 X
technique is a better measure for heteroskedasticity in a case when the source of heteroskedasticity is known. But in the present model the source of heteroskedasticity is uncertain and as a result the White technique for large samples is applied. In this study the sample is of a satisfactory size. White (1980) has established a technique that approximates the corrections of their population parameters under conditions of heteroskedasticity in the case of large samples. This provides a heteroskedasticity-consistent sample variance estimate of the standard errors. Thus, the robust White covariance estimator is resistant to the heteroskedasticity problem. The White covariance estimator will be used to minimize the heteroskedasticity problem.
MR4. The covariance between any pair of random errors is cov ( εi, εj ) = 0
The fourth assumption, MR4 implies that errors are not autocorrelated. The randomness of the sample indicates that error terms for different observations should be uncorrelated for cross-sectional data. Most often autocorrelation occurred as a problem in the case of time series analysis, but also this assumption could be violated in the case of cross-section data. For that reason Durbin-Watson test will be used to diagnose the autocorrelation.
MR5. The values of xtk are not random and also not exact linear functions of the other independent variables.
4.2 Econometric model
The regression analysis is composed from three regression models that determine the influence of the independent variables (E/TA, L/TA, STF/TA and EA/TA, OV/TA, FO (dummy variable) and BS), macro economic variables (INF and CA/GDP) on dependent variable NIM of the banks, and to identify whether or not the hypotheses mentioned above are supported. Many academics have measured bank profitability with NIM (Demirguç-Kunt et al 2004, Abreu and Mendes 2005, Uiboupin (2004), etc).
The multiple regression model indicating the influence of independent variables on bank profitability can be shown in the following equations:
NIM = β0 + β1 E/TA+ β2L/TA + β3 STF/TA + β4 EA/TA + β5 OV/TA + β6 FO (dummy variable) + β7 BS + β8 INF + β9 CA/GDP
Where: β0is a constant that represents what the net interest margin will be if all the other
variables are 0, NIM is the net interest margin (dependent variables), E/TA is the equity to total assets ratio, L/TA is the loans to total assets ratio, STF/TA is the short term funding to total assets ratio, EA/TA is the earning assets to total assets ratio, OV/TA is the overhead costs to total assets ratio, FO is the foreign ownership(dummy variable where 0 is for domestic and 1 is for foreign ownership) and BS is the bank size (bank-specific variables), INF is the inflation and CA/GDP is the current account balance to GDP ratio (macroeconomic variables).
To determine the impact of foreign ownership (variable six) on bank profitability dummy variable will be used. The number of dummy variables will be created according to dummy rule (k-1) where k is the number of the original variables and k-1 is the number of dummy variables. Because two types of banks are examined (with 50 and less than 50 percent of foreign ownership) one dummy variable is needed (2-1=1). Therefore, the variable is coded 1 for banks that have 50 or more percentage of foreign ownership and 0 for banks that have less than 50 percentage of foreign ownership to which all other banks will be compared.
number of the original variables and k-1 is the number of dummy variables. Because five world regions are examined in this study, it is necessary to create four dummy variables (5-1=4). Dummy variables for regions East Europe, West Europe Far and Central Asia and South and Central America will be created which will be coded with 1, 2, 3 and 4. The decision concerning which region will not be coded is always arbitrary. The region which is not coded is the region to which all other regions will be compared. In this case the North American region is not going to be coded. So North America is 0. Multiple regression analysis, including dummy variables, will be conducted to determine the influence of the bank specific and macroeconomic variables on bank performance and to examine the similarities and differences among the regions. The equation is:
NIM = β0 + β1 E/TA+ β2L/TA + β3 STF/TA + β4 EA/TA + β5 OV/TA + β6 FO (dummy variable) + β7 BS + β8 INF + β9 CA/GDP + β10 (dummy East Europe) + β11 (dummy
West Europe) + β12 (dummy Asia) + β13 (dummy South and Central America)
Additionally, comparison between banks that come from developed (North America and West Europe) and developing regions (East Europe, Asia and South and Central America) are going to be examined in order to determine the impact of bank specific and macroeconomic variables on developed and developing regions and also to observe which banks from which regions (first group developed countries: West Europe, North America and the second group transition and developing countries: East Europe, Asia, South and Central America) are more profitable. Therefore, a dummy variable will be created for banks from developing countries that will be coded 1. Banks from developed countries will be not coded, so they are 0. The equation is:
4.3 Measurement of the dependent, independent and control variables
In order to measure the bank- specific and macro economic variables for the year 2004, and to place them in the regression analysis model the variables are defined as follows:
The variable NIM (dependent variable) is the ratio defined as interest income
minus interest expense divided by average interest-bearing assets.
The variable E/TA (independent variable) is the ratio defined as book equity
divided by total assets.
The variable L/TA (independent variable) is the ratio defined as loans divided by
total assets.
The variable STF/TA (independent variable) is the ratio defined as short term
funding divided by total assets.
The variable EA/TA (independent variable) is the ratio defined as earning assets
divided by total assets
The variable OV/TA (independent variable) is the ratio defined as overhead costs
divided by total assets.
The variable FO (independent variable) is a dummy variable whereas dummy
equals one if bank has at least 50% of foreign ownership.
The variable BS (independent variable) is defined as natural logarithm of total
bank assets.
The variable INF (macroeconomic variable) is defined as nominal value express
in percentage for each country.
The variable CA/GDP (macroeconomic variable) is the ratio defined as the current
account balance divided by GDP for each country.
Abreu and Mendes (2005) have used annual data to examine the impact of banks and macro financial variables on bank margins and performance. For this reason, they use time-series method at the cost of not smoothing the variables that vary over time.
5. Data
6. Results
6.1 Descriptive statistics
Tables 1 to 6 in the appendix show the descriptive statistics for the dependent variable net interest margin (NIM) and independent bank-specific and macro-economic variables. The mean value for the net interest margin (NIM) varies from region to region. South American and East European regions have higher mean values of the net interest margin compared to other world regions. Further, the STF/TA and EA/TA independent variables for the Asian region have a higher mean value in comparison to the other regions. Other
obvious characteristics are that the OV/TA independent variable for the East European region have higher mean values in comparison to the other regions.
6.2 Bivariate correlation
By using a bivariate correlation analysis, multicolliniarity among the variables was examined. The results indicate that highest correlation exists between GDP growth and inflation. The correlation between these two variables is -0,453. The conclusion drawn is that variables are not highly correlated. Furthermore, the results from Durbin-Watson test for autocorrelation show that there is no autocorrelation. Results from the Durbin-Watson test can be found in appendix B, C and D.
Table 1. Bivariate correlation
NIM ETA LTA STFTA EATA OVTA FO lnbanksize inflation CAGDP
NIM 1 ETA ,243** 1 LTA ,198** -,185** 1 STFTA -,081** -,453** ,075** 1 EATA -,178** -,115** ,203** ,205* 1 OVTA ,330** ,231** -,060** -,137* -,378** 1 FO ,054** ,117** -,147** 0,015 -,078** ,113** 1 lnbanksize -,143** -,276** 0,033 ,046* ,044* -,141** -,101** 1 inflation ,380** ,082** -,090** 0,041 -,208** ,189** ,181** -,203** 1 CAGDP -,132** ,081** -,149** -0,47* ,070** 0,026 ,122** -,048** -0,042 1
6.3 Multiple regression analysis
Table 2. Multiple regression analysis
Dependent V ariable NIM B -c oeffic ient t-s tatis tic s P robability
C o n s ts n t 0 .8 4 7 0 .5 8 9 0 .5 5 5 E TA 0 .0 6 2 5 .9 4 9 *** 0 .0 0 0 *** L TA 0 .0 4 0 1 3 .7 9 7 *** 0 .0 0 0 *** S TFTA 0 .0 0 1 0 .2 6 5 0 .7 9 0 E ATA -0 .0 2 1 -1 .4 7 5 0 .1 4 0 OVTA 0 .1 9 1 2 .9 5 5 0 .0 0 3 FO -0 .2 4 0 -1 .3 4 9 0 .1 7 7 2 L N B AN K SIZE -0 .0 2 2 -0 .9 2 3 0 .3 5 5 IN FL ATION 0 .3 5 2 6 .8 1 5 *** 0 .0 0 0 *** C AGD P -0 .0 4 1 -6 .1 1 6 *** 0 .0 0 0 *** R -s q ua re d 0 .3 2 6 D u rb in -W a ts on 1 .6 8 8
Num ber of obs ervations 2291 For calculation White covariance estimator is used
*, **, ***-significant at 10%, 5% and 1%
Source: The Author
6.4 Multiple regression analysis with regional dummy variables
Table 3. Multiple regression analysis with regional dummy variables Dependent V ariable NIM B -c oeffic ient t-s tatis tic s P robability
C o n s ta n t 0 .3 8 9 0 .2 7 2 0 .7 8 5 E TA 0 .0 5 5 5 .3 3 8 * * * 0 .0 0 0 * * * L TA 0 .0 4 0 1 4 .2 4 0 * * * 0 .0 0 0 * * * S TFTA -0 .0 0 0 -0 .2 0 0 0 .8 4 1 E ATA -0 .0 1 5 -1 .0 5 5 0 .2 9 1 OVTA 0 .1 8 0 2 .9 7 2 0 .0 0 3 FO -0 .4 2 7 -2 .4 7 5 0 .0 1 3 L N B AN K S IZE -0 .0 3 1 -1 .2 2 7 0 .2 1 9 IN FL ATIO N 0 .2 9 8 5 .0 2 2 * * * 0 .0 0 0 * * * C AG D P -0 .0 4 2 -5 .5 7 6 * * * 0 .0 0 0 * * * D U ME AS T 0 .5 8 4 1 .9 2 9 * * 0 .0 5 3 * * D U MW E E U R 0 .0 5 5 0 .3 5 9 0 .7 1 9 2 D U MAS IA 1 .2 2 5 4 .6 8 4 * * * 0 .0 0 0 * * * D U MS C AME 2 .3 1 7 6 .3 0 4 * * * 0 .0 0 0 * * * R -s q u a re d 0 .3 6 2 D u rb in -W a ts o n 1 .7 6 6
Num ber of obs ervations 2291 For calculation White covariance estimator is used
*, **, ***-significant at 10%, 5% and 1%
Source: The Author
bank specific and macroeconomic variables are similar to the previous regression model. So hypothesis 1, 2 and 8, 9 are supported at 1% significance level.
The beta-coefficients (0,584) for the dummy variable for the East European region has a positive sign which means that banks from this region have higher net interest margin for 0,584 % compared with banks from West Europe and North America regions, but lower profit than banks from the Asian and South American regions. The dummy variable East for the European region is significant at 10%.
Furthermore, the dummy variable for the West European region is not significant and is inconclusive as individual region on independent bank specific and macroeconomic variables. Significance level for the West European region is 0,719 which mean that the independent dummy variable West Europe is not significant at 5% and 10%.
Moreover, from table 2 we can see that Asian banks have higher net interest margin for 1,225 % compared to banks from East Europe, North America and West Europe, but they have lower profit than South American banks. The independent dummy variable Asian region is supported at 1%.
6.5 Multiple regression analysis with dummy variable (banks from developing countries)
Table 4. Multiple regression analysis with dummy variable banks from developing countries D e p e n d e n t V a ri a b le N IM B -c o e ffic ie n t t -s t a t i s t ic s P ro b a b i lit y C 0 .3 8 1 0 .2 6 2 0 .7 9 2 E T A 0 .0 5 5 5 .4 2 4 * * * 0 .0 0 0 * * * L T A 0 .0 3 9 1 3 .8 9 5 * * * 0 .0 0 0 * * * S T F T A - 0 .0 0 1 - 0 .4 5 9 0 .6 4 6 E A T A - 0 .0 1 4 - 0 .9 4 5 0 .3 4 4 O V T A 0 .1 9 0 3 .0 1 6 0 .0 0 2 F O - 0 .4 8 8 - 2 .8 2 1 0 .0 0 4 L N B A N K S IZ E - 0 .0 2 6 - 1 .1 0 9 0 .2 6 7 IN F L A T IO N 0 .2 9 6 5 .5 8 0 * * * 0 .0 0 0 C A G D P - 0 .0 4 0 - 6 .0 8 3 * * * 0 .0 0 0 D E V E L O P IN G 1 .2 8 0 6 .7 0 0 * * * 0 .0 0 0 R - s q u a r e d 0 .3 4 9 D u r b i n - W a ts o n 1 .7 3 2 N u m b e r o f o b s e r va t i o n s 2 .2 9 1 For calculation White covariance estimator is used *, **, ***-significant at 10%, 5% and 1%
Source: The Author
% than banks from developed countries. The independent dummy variable of developing countries is supported at 1% significance level.
7. Conclusion
This thesis investigates the influence of bank-specific and macro-economic variables on bank performance and whether differences and similarities in bank performance exist among banks from five world regions. A cross-sectional investigation was made for developed countries (West Europe and North America) and developing countries (East Europe, Latin America and Far and Central Asia) for the year 2004.
Asian banks. This can be due to the fact that in recent years a lot of improvements have been made in East European countries in terms of undertaking appropriate bank policy, risk management improvement and market consolidation which contribute to lowering the net interest margin. Furthermore, the foreign bank entry in East European region plays an important role in increasing the levels of efficiency which reduces the net internet margin. Fries et al. (2002) examine the performance of 515 banks in 16 Central and Eastern (transition) economies for the period 1994–1999. Their results show that costs of the banks have decreased over time in the transition countries which contribute to lowering the net interest rate.
On the other hand, banks from developed countries (West European and North American regions) have lower net interest margins due to high cost efficiency, a good management practice, developed financial systems and a market concentration. They are characterized with tougher competition, a well developed regulatory framework, accomplishing economics of scale and lower profits. Additionally, the results indicate that banks from developing countries have higher net interest margin in comparison to banks from developed countries. Banks from developing countries are characterized by cost inefficiency, non existence economies of scale and less-than-competitive pricing behavior as also suggested by their relatively high profitability and net interest margins.
Results from the three regression models confirm that the hypothesis one, two, eight and nine are supported. Equity to total assets ratio has a positive influence on net interest margin due to the fact that banks with a higher equity to total assets ratio have a lower need for external funding and therefore possibilities for higher profits. Higher equity to total assets ratio gives opportunities of the banks to issue more credits.
extensive branch network incurs high costs that results in lowering the bank profitability. The development of the stock exchange gives more attractive opportunities for customers to invest their funds in financial derivatives. In recent years, financial derivatives have become more attractive in comparison to deposit interest rates of the banks, which contribute to lower bank profitability.
The next finding in this thesis is that inflation has a positive influence on bank profitability. One explanation for the positive influence of inflation on bank profitability is that the banks are trying to forecast inflation which implies that interest rates have been appropriately adjusted to accomplish higher profits. This finding is in line with the findings of Barth et al (1997), Denizer (2000). They confirmed that on the one hand inflation entails higher costs for wages, more transactions, etc., but on the other hand inflation contributes bank profitability to increase more than bank costs. Moreover, Abreu and Mendes (2005) argue that increasing current account balance to GDP ratio contributes to declining the bank performance.
Furthermore, the hypotheses four, five and six are not supported. The results show that non interest earning assets (cash, real estate, etc) can be very costly for the banks if they are not used on the proper way. Moreover, the positive relationship between overhead costs to total assets ratio and bank profitability can be explained by the fact that banks pass their overhead costs to their customers (depositors and lenders). Hypothesis six is not supported. The reason for this result might be that foreign bank entry is generally associated with decreasing net interest incomes in the developing countries which reduces the profitability of domestic banks. In developed markets this decrease is lower because the foreign banks do not have such a high competitive advantage as in developing countries. Peria and Mody (2004) have found that foreign ownership dummy has a negative sign, supporting the proposition that foreign banks are able to charge lower interest margins.
market share which gives them possibilities to decrease the net interest margin and to increase their competences on the market.
8. Reference
Abreu M. and Mendes V.(2005) “European Bank Net Interest Margins: Do macrofinancial variables matter?" Revue Bancaire et Financière/Bank en Financiewezen Larcier, Mars 2005, nº2, pp.101-105
Barth J. Nolle E. Phumiwasana T. and Yago G. (2003) “A Cross-CountryAnalysis of the Bank Supervisory Framework and Bank Performance”, Financial Markets, Institutions & Instruments, Vol. 12, pp. 67-120
Barth, J. R., D. E. Nolle, and T. N. Rice (1997) “Commercial Banking Structure, Regulation, and Performance: An International Comparison”, Comptroller of the Currency Economics WP 97-6.
Begg D. (1997)” Monetary policy during transition: progress and pitfalls in central and eastern Europe, 1990-6” Oxford Review of Economic Policy, Vol. 13, pp. 33-47
Claeys, S., Vander Vennet, R. (2004) “Determinants of bank interest margins in Central and Eastern Europe”: A Comparison with the West. Mimeo.
Demirguc-Kunt, A., Claessens, S. and Huizinga, H. (2001) “How Does Foreign Entry Affect Domestic Banking Markets?” Journal of Banking and Finance, Vol. 25, pp. 891-911
Demirguc-Kunt A. and Huizinga H. (1999) "Determinants of Commercial Bank Interest Margins and Profitability: Some International Evidence," World Bank Economic Review, Oxford University Press, vol. 13, pp. 379-408
Demirguç-Kunt, A. and H. Huizinga (2000) “Financial Structure and Bank Profitability”, World Bank Policy Research WP 2430.
Denizer, C. (2000) “Foreign Entry in Turkey’s Banking Sector, 1980-97”, World Bank Policy Research WP 2462.
Fries S., Neven D. and Seabright P. (2002) “Bank performance in transition economies” HEI Working paper No. 7
Gelos G. (2006) “Banking Spreads in Latin America“ International Monetary Fund (IMF), working paper, WP/06/44
Gerlach S. Peng W. and Shu C. (forthcoming) “Macroeconomic conditions and banking performance in Hong Kong: A panel data study,” (with.) forthcoming in Investigating the relationship between the financial and real economy, BIS.
Hanson J. and Rocha R. (1986) “High interest rates, spreads, and the cost of intermediation” Two Studies: World Bank Industry and Finance Series 18.
Hanweck A. and Ryu H. (2005) "The Sensitivity of Bank Net Interest Margins and Profitability to Credit, Interest-Rate, and Term-Structure Shocks Across Bank Product Specializations". FDIC Working Paper No. 05-02
Havrylchyk O. and Jurzyk E. (2006) “Profitability of foreign banks in Central and Eastern Europe: Does the entry mode matter?” CEPII publications, Paris
Holló D. and Nagy M. (2006) “Bank efficiency in Enlarged European Union” Magyar Nemzeti Bank (MNB) Working Paper.
Kaminsky, G. L. and Reinhart C. M. (1998) “Financial Crises in Asia and Latin America: Then and Now.” American Economic Review Vol. 88, pp. 444-448
Laeven, L. (2005) “Banking Sector Performance in East Asian Countries: The Effects of Competition, Diversification, and Ownership.” Mimeo. Washington, DC: World Bank.
Levine R. and Zervos S. (1998) “Stock markets, banks, and economic growth” American Economic Review Vol. 88, pp. 537-558
Levine R., Barth J., and Caprio G. (2004) “Bank Regulation and Supervision: What works best?” Journal of Financial Intermediation Vol. 13, pp. 205-249
Long M., and Rutkowska I. (1995) “The role of commercial banks in enterprize restructuring in Central and Eastern Europe” Policy Research Working Papers 1423 World Bank
Mathieson, D. J. and Roldos, J. (2001) “The Role of Foreign Banks in Emerging Markets.” Foreign participation in financial systems in developing countries 2001, pp. 15-55, World Bank/IMF/Brookings Emerging Market Series. Washington, D.C. Brookings Institution Press
Peria M. and Mody A. (2002) “How Foreign Participation and Market Concentration Impact Bank Spreads”: Evidence from Latin America,” Journal of Money, Credit, and Banking Vol. 36, pp. 511-537
Panayiotis A., Sophocles B. and Matthaios D. (2005) "Bank-Specific, Industry-Specific and Macroeconomic Determinants of Bank Profitability," Working Papers 25, Bank of Greece
Petersen, W. M. (1986). “The Effects of Inflation on Bank Profitability”, in Recent Trends in Commercial Bank Profitability – A Staff Study, Federal Reserve Bank of New York, pp. 89-114
Uiboupin J. (2004). “Effects of foreign banks entry on bank performance in the CEE countries” Tartu University Press, order No. 569
