MSc Business Economics, Finance track
Do changes in macroeconomic conditions in home
and host country differently affect domestic and
foreign banks?
-Master Thesis-
Author: Radoslovescu Ana Supervisor:
(10622810) R.I. Todorov
ABSTRACT
This thesis uses a panel dataset comprising 3,389 banks located in 93 countries during the 2001-2012 period and aims to find the extent to which foreign and domestic banks are affected by the change in macroeconomic conditions in home and host countries. Using two measures to quantify banks’ performance, the paper finds strong evidence that foreign banks are differently affected than domestic banks, but the way they are influenced varies depending on a series of factors such as the level of development and ownership.
CONTENTS
I. Introduction...3
II. Literature review...6
III. Methodology...11 IV. Data...14 a. Data...14 b. Descriptive Statistics...16 V. Results...20 a. Foreign Banks...20 b. Domestic Banks...25
VI. Robustness check...29
a. Weighted Least Squares...29
b. Different measures of profitability...31
c. Fixed Effects...32
d. Crisis Period...33
e. Mergers and Acquisitions...35
VII. Conclusion...36
References...39
Appendix
I. Introduction
In a world where the accent is put on globalization and financial integration, banking activity has grown and become increasingly global. There is no longer independence between the banks operating in two or more different countries of the world, even if they don’t have any activities in foreign countries. Besides that, over the past two decades, foreign banks’ role in financial intermediation has become more and more important (Claessens and van Horen, 2013).
There are multiple factors that influence bank’s performance, but whether the factors are related to the bank’s home country or the host country may differently affect the bank. First, bank profits do not fluctuate significantly over time, but the persistence tend to be sensitive to macroeconomic shocks, especially monetary policy (Berger et al, 2000). Banks that have a high percentage of assets invested in consumer or real-estate loans were strongly affected by macroeconomic expansion. Second, the relation that exists between private credit and presence of foreign banks1 was found to depend on host country and characteristics of banks from that country (Claessens and van Horen, 2013). So the question arises: do macroeconomic conditions (i.e.: monetary policy) in home and host country differently affect domestic and foreign banks?
Unlike domestic banks, foreign banks have various advantages. First of all, they operate in more countries so their clients’ portfolio is wider and their profitability can rise quicker than in the case of domestic banks. Second, risks took by foreign banks are greater than those took by domestic banks. An argument can be the fact that foreign banks usually are supported by the parent bank so their funding sources are more diversified.
On the other hand, foreign banks also face a series of disadvantages. For example, the level of provision for foreign banks is higher than in the case of domestic ones. Moreover, they have less information about the market and even about the country so they will need an adaptation time in which their costs will by higher leading to a lower profit.
The entrance of foreign banks into less-developed markets is the subject of many empirical papers but the majority of them focuses on the influence exerted by foreign banks on the efficiency of banking systems in the host countries. But the literature related to the degree by which foreign and domestic banks are differently affected by changes in macroeconomic conditions is rather poor.
From all factors took into account, it can be said that interest rate changes and the internal market situation counts for the profitability of banks, whether they are foreign or domestic. In developing countries, foreign banks’ stability is positively influenced by the changes in monetary policy, while domestic banks’ stability is negatively but not statistically significant influenced. In developed countries, both foreign and domestic banks’ stability is influenced in the same way.
It can be easily stated that banks which granted loans without correctly assessing the reimbursement capacity of the borrowers will have a lower performance than banks which correctly assessed the creditworthiness of its borrowers. But if we have two banks with a similar set of characteristics (i.e.: same performance level) from two different countries and there is a change in one’s country, which bank will be affected more, the domestic bank or the foreign one? It can be thought that a worsening of the conditions of one’s particular market will lead to a reduction of the interbank rate, thus affecting the foreign bank, but which bank will suffer more? Furthermore, worsening conditions in the home country may force a bank originating from that country to repatriate funds from its subsidiaries in a host
country in order to support its parent bank.
The existing literature is ambiguous when it comes to the performance of foreign banks compared to domestic ones. Some studies (Berger, Hasan, Zhou (2009); Havrylchyk and Jurzyk (2011); Sturm and Williams (2004)) found that regardless the measure of performance chosen, foreign banks will have an advantage, while other papers state that domestic banks perform better (DeYoung and Nolle (1996); Miller and Richards (2002); Chang, Hasan and Hunter (1998)). Due to the ambivalent literature, this research aims to shed light on this topic.
My contribution can be resumed to two major key points. First, I will take into account changes in both home and host countries in order to analyze the impact on foreign and domestic banks. From what I know, in the research conducted so far, the authors put emphasis more on the host country macroeconomic conditions and analyzed the shocks that occurred in host country. Second, it is hard to quantify the performance of a bank by only one indicator. In my research, due to the high importance of banks in the economy, I will quantify the performance at two levels: profitability and stability.
Finally, more emerging countries tend to become more and more developed and in the future banking groups coming from this developing countries could penetrate other developing countries’ markets. The results of the empirical analysis from this thesis combined with previous studies will be useful for banks to know in which countries they will have better performance. The issues discussed in this thesis are also relevant for developing countries, especially the countries from Eastern Europe and Latin America, that have to make a decision regarding the opening of their banking markets to foreign investment.
II.Literature Review
There is a lot of evidence in the existing literature that foreign and domestic banks have different behaviors and respond differently to similar conditions. Berger, DeYoung, Genay and Udell (2000) conducted a research on different banks in countries like US, UK, Germany, France and Spain, showing that domestic banks are more efficient when it comes to profit as compared to foreign banks. So, a bank’s performance seems to depend on the country of origin. Claessens and van Horen (2012) showed that the performance of foreign banks is higher with respect to capital and liquidity. In other words, foreign banks tend to have a higher level of liquidity and they are able to raise a bigger amount of capital if needed. However, they also found out that these banks are less profitable than domestic banks.
The existing literature also emphasized that changes in conditions of host and home countries may differently affect domestic and foreign banks. The simple existence of a multinational banking group2 implies a high level of interdependency between the parent bank and its subsidiaries in other countries and also between each subsidiary and the others from that group belonging to other countries. So the subsidiaries will strongly be affected by the events and changes that occur in the rest of the multinational banking group.
De Haas and van Lelyveld (2010) find that bank subsidiaries are not stand-alone but rather operations influenced by what happens in other subsidiaries in the multinational banking group with activities in several countries as well as what happens in the parent bank in the home country. This is not the result of macroeconomic international linkages, but the result of internal capital markets. The existence of internal capital markets is proved by the
presence of two types of effects which directly influence the lending activity of subsidiaries. First, lending is influenced by substitution effects which mean that the parent bank from the home country trade-off lending in the countries where it has subsidiaries. Second, lending is also affected by support effects which is equal to say that the parent bank supports its weak subsidiaries.
Subsidiaries from host countries were found to expand their lending faster when economic growth in their home country decreased (De Haas and van Lelyveld, 2010). Hernandez and Rudolph (1997) previously conducted a research from which they found out that worsening conditions in the home country led foreign banks to seek more external lending opportunities. This finding is similar to the one of Martinez Peria et al. (2002) who emphasized that during a period in which home country conditions worsened foreign banks increased their lending to Latin American countries. So there is some evidence showing that changing economic conditions in home countries will affect the behavior of bank subsidiaries around the world.
De Haas and Lelyveld (2014) also found that subsidiaries have a more stable lending activity during local crises. In other words, the presence of multinational banks in the host countries reduces the impact of domestic financial shocks. For example, Dages et al (2000) studied a sample of banks from Argentina and Mexico and proved that during times of local crises, foreign banks showed a significant level of credit growth, unlike domestic banks. Goldberg (2002) also showed that foreign banks did not reduce their level of credit supply during domestic shocks in host countries.
However, in the same paper, de Haas and Lelyveld discovered that the presence of multinational banks also allows for the transmission of foreign shocks in the host countries, just like in the case of the recent financial crisis (2007-2008). For example, multinational
bank subsidiaries had to reduce credit growth three times more than domestic banks. This is similar to the findings of Claessens and van Horen (2012) who discovered that during the recent financial crisis, the banks that reduced the credit more were the foreign ones and not the domestic ones.
Peek and Rosengren (1996) also showed that the globalization of banking in particular does have downsides in the sense that a problem that occurs in a domestic market can affect other markets as well. A significant example has been provided during late 1980s and early 1990s by the surge in Japanese stock prices. Even though the fluctuations in Japanese stock prices were Japanese domestic financial shocks, they negatively affected other countries through the Japanese banking system. A direct effect was a significant decline in commercial lending by Japanese banks in the USA. The conclusion that can be withdrawn from this study is that the transmission of economic shocks does occur at the international level as well. It can be seen the importance of the connection between foreign and domestic banks and the existence of foreign banks on the national market, so the subject of the thesis is an important concern for both economists and banks.
De Haas and Lelyveld (2014) also found evidence related to the differences between greenfield subsidiaries and subsidiaries resulted from take-overs by multinational banks. According to their findings, greenfield subsidiaries exhibit a higher level of integration into internal capital markets of multinational banks than subsidiaries resulted from a past take-over. So the extent to which a change in the macroeconomic conditions of the host country affects the foreign banks is stronger for greenfield subsidiaries than for independent subsidiaries resulted from take-overs, as the latters could be expected not to be fully integrated in the multinational banking group.
This result is also similar to the findings of de Haas and Naaborg (2005) who studied a sample of foreign banks from Central Europe and Baltic states taking into account whether they were the result of a greenfield investment or the result of a take-over. They discovered that the local banks that have been taken-over by foreign banks remain quite independent for a certain period of time, since their local managers remain, the risk management and treasury management methodologies and other factors remains the same for a period of time after the acquisition. In contrast, greenfield subsidiaries are usually modeled after their parent bank in the home country, thus being stronger integrated in the internal capital market of the parent bank.
By including control variables to isolate other factors that influence the performance of banks this research will provide more accurate results. Micco, Panizza and Yanez (2007) found that in developing countries, domestic banks owned by the government tend to be less efficient than banks owned by private economic agents. Their research emphasized the impact of the form of property on bank’s performance. A publicly owned bank may respond differently than a privately owned bank to changing economic conditions.
The level of development of the host and home country may also have a significant impact on the analysis of this thesis. Whether a country is developed or emerging may influence the extent to which domestic and foreign banks are affected by the changes in economic conditions in host and home countries. Claessens and van Horen (2012) find that foreign banks perform better than their domestic counterparts if they come from a developed country. They may be able to achieve economies of scale, to better diversify their portfolios and to manage risk better. In their paper, they also use other variables to control for omitted factors determining bank performance (size, market share, funding structure, asset structure, leverage, general level of development of a country, financial depth).
Hermes and Lensink (2004) use data for almost 1000 banks in 48 countries and find two different results depending on the level of development of the financial system in a given country. On one hand, in a country with a low level of financial development (emerging country), foreign banks tend to negatively affect further development of financial system by increasing costs for domestic banks. On the other hand, in a country highly developed from a financial perspective (developed countries), foreign banks tend to do the opposite thing, contributing to the cost reductions and helping to improve financial stability and development in that country. So, the quality and the level of development of the financial system for a given country is also important in order to find out if changing in economic conditions in that country affects differently domestic and foreign banks.
Economic conditions have, of course, a strong impact on the performance of banks. However, to ignore the consequences of natural events is to neglect an important factor that may change the actual behavior of banks. There are several studies which reflect the transmission of natural shocks. For example, Lambert, Noth and Schauwer (2011) studied the impact of hurricane Katrina among the banks in USA. Although the banks located in the US Gulf Coast area were affected by the hurricane with estimated damages of more than $200 billion, other banks located in the north or east of USA also suffered. The reason is the strong interconnection between banks at the national level. Since USA banking system is strongly connected to other banking systems in the world, the hurricane also affected the rest of the world, not only USA.
III. Methodology
The objective of this thesis is to find out if the changing of macroeconomic conditions differently affects domestic and foreign banks. Because of the international activity of foreign banks, I will take into considerations economic conditions from both home country where the parent bank is located and host country where the subsidiary operates.
Banks’ activities imply a wide range of activities which involve all types of economic agents. Therefore, the macroeconomic conditions will be quantified at two levels: monetary policy and the state of the local market. As Claessens and van Horen (2010) showed, the decision of an international bank to enter a market depends on the latter’s general state. The monetary policy will be measured by the basic interest rate established by the Central Bank of a given country and the state of the local market by the exchange rate. Macroeconomic indicators such the real GDP growth will be taken into account as country-level controls.
The main hypotheses of the thesis are the following:
1. In developed countries, foreign banks usually come from other developed countries, so domestic and foreign banks have similar performance characteristics; in other words, the extent to which they will be affected by a change in macroeconomic conditions of host country should not be significantly different if the monetary policy is similar too.
2. In developing countries, foreign banks will be more affected than domestic banks by a change in macroeconomic conditions of host countries due to the fact that foreign banks are not used to the bad conditions of less mature economies
3. Change in economic conditions in host countries: domestic banks (located in the host country) will not be that much affected as compared to the foreign banks; the extent to which these banks are affected will be established by my research
The empirical analysis conducted in this paper will primarily consist of performing a regression on a vector consisting of two measures of bank’s performance against factors that influence the performance of the bank. The empirical model will be based on panel data. The set of data will contain observations for several banks during a period of 11 years.
To quantify the performance of the bank, two measures will be used. First, profitability will be taken into consideration. It is measured by bank’s return on assets, based on previous studies which used this indicator as a proxy for profitability and based on an ECB report (2010) on measures of profitability. Second, the stability, measured by a bank’s z-score, will be used as a proxy.
The performance of the bank is influenced by several factors. The most important for the current analysis is whether a bank is originating from the country where it is located –it will be a domestic bank in the host country– or is from another country – it will be a foreign bank and the country of origin will be further denoted the home country. In this respect, a dummy variable which is equal to 1 if the bank is a foreign one and 0 otherwise will be used.
I will use other variables which will allow me to control for omitted variable bias. Using dummy variables, I will control for the form of ownership (private or public), grade of development of the host country and home country (developed or emerging). To establish if a country is developed or emerging I will use the World Bank’s classification.
The extent to which foreign banks and domestic banks will be differently affected by a change in macroeconomic factors will be quantified using generalized least squares
approach.
𝑉𝑉𝑖𝑖,𝑗𝑗,𝑡𝑡 = 𝜆𝜆0+ 𝜆𝜆1𝐹𝐹𝑖𝑖,𝑗𝑗+ 𝜆𝜆2Δ𝑅𝑅𝑗𝑗,𝑡𝑡+ 𝜆𝜆3Δ𝑅𝑅𝑖𝑖,𝑡𝑡+ 𝜆𝜆4Δ𝑅𝑅𝑖𝑖,𝑗𝑗,𝑡𝑡 + 𝜆𝜆5Δ𝐸𝐸𝐸𝐸𝐸𝐸ℎ𝑖𝑖,𝑗𝑗,𝑡𝑡 + 𝜆𝜆6𝐷𝐷1𝑖𝑖 + 𝜆𝜆7𝐷𝐷2𝑖𝑖
+ 𝜆𝜆8𝐷𝐷3𝑗𝑗+ 𝜑𝜑 + 𝜚𝜚 + 𝜀𝜀𝑖𝑖,𝑗𝑗,𝑡𝑡
Where “i” represents the bank, “j” the country, “t” time, 𝜆𝜆0, 𝜆𝜆1, 𝜆𝜆2, 𝜆𝜆3, 𝜆𝜆4, 𝜆𝜆5, 𝜆𝜆6, 𝜆𝜆7, 𝜆𝜆8 are the regression coefficients and 𝜀𝜀𝑖𝑖,𝑗𝑗,𝑡𝑡 is the error term. Fi,j is
a dummy variable which takes value 1 if bank i is foreign and 0 otherwise. 𝜆𝜆1 is the coefficient of interest and it measures the change in performance between a foreign bank and a domestic one. Vi,j,t is a vector representing performance and is quantified by return on
assets and bank’s z-score. Rj,t is the basic interest rate decided by the Central bank in country
j (host country) and Ri,t is the basic interest rate in the home country. In the regression
model, I will use the change in interest rate both in home and host countries. It is the main proxy used to quantify the macroeconomic shock at monetary level.
Also, I will use two variables to quantify the differences between home and host country: Δ𝑅𝑅𝑖𝑖,𝑗𝑗,𝑡𝑡 is the difference in interest rates and Δ𝐸𝐸𝐸𝐸𝐸𝐸ℎ𝑖𝑖,𝑗𝑗,𝑡𝑡 represents the difference between the foreign exchange rate and the domestic one. D1,i is a dummy variable which
takes value 1 if bank i is private owned and 0 otherwise. D2,i is a dummy variable which
takes value 1 if the home country is developed and 0 if the country is emerging. (the classification of countries will be made according to World Bank). D3,j. is a dummy variable
which takes value 1 if the host country is developed and 0 if the country is emerging. ϕ is a vector consisting of country characteristics (market capitalization, non-performing loans, domestic credit and real GDP growth) and ϱ is a vector consisting of bank characteristics (liquid assets ratio, deposits to assets ratio, deposits to liabilities ratio and loan loss provision).
The Return on Assets will be calculated using the following formula:
𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖,𝑡𝑡 = 𝑇𝑇𝐼𝐼𝑁𝑁𝑇𝑇𝑇𝑇 𝑅𝑅𝐴𝐴𝐴𝐴𝑁𝑁𝑁𝑁𝐴𝐴𝑁𝑁𝑁𝑁𝑁𝑁 𝐼𝐼𝐼𝐼𝐸𝐸𝐼𝐼𝐼𝐼𝑁𝑁𝑖𝑖,𝑡𝑡 𝑖𝑖,𝑡𝑡
The Z-score is a measure of distance from insolvency (Roy, 1952) and is calculated based on this equation (Beck, de Jonghe and Schepens, 2013):
𝑍𝑍 − 𝐴𝐴𝐸𝐸𝐼𝐼𝑠𝑠𝑁𝑁 =𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖,𝑡𝑡+ 𝐸𝐸𝑅𝑅𝑖𝑖,𝑡𝑡 𝜎𝜎𝑅𝑅𝑅𝑅𝑅𝑅𝑖𝑖,𝑡𝑡
Where ROAi,t represents return on assets for bank i in year t, E/Ai,t is the ratio of
equity to total assets and 𝜎𝜎 denotes the variance of return on assets.
IV. Data
In order to gain insights into how banks’ performance is affected by changes in macroeconomic conditions, the sample is constructed on all banks in the world for the period 2001-2012. Years 2000 and 2013 were omitted from the sample because there is a considerable lack of information regarding banks and macroeconomic conditions. Data was collected from several databases. Information about banks was collected from balance sheet and income statements entries from Bankscope, a database compiled by Bureau van Dijk. The banks in the sample belong to one of the following categories: commercial banks (representing 84.7% of the sample), cooperative banks (2.05%), savings banks (8.05%) and bank holdings (5.2%).
Several selection criteria were applied to the database. It contains both consolidated and unconsolidated entries so in order to avoid double counting, I deleted the unconsolidated observations. First, the bank-year observations with missing values on basic variables taken into account in the regression model were dropped. Second, I kept only the banks that had data for at least three consecutive years in order to avoid panel bias. Third, so that I can compare banks from different countries, I deleted the countries for which there were less than 100 bank observations. (DeJonghe and Oztekin (2013)).
My sample contains both foreign banks and domestic banks so bank-year observations with missing information on home country were also deleted. The distinction between foreign and domestic banks was made according to Claessens and van Horen (2012), a bank will be considered foreign if it is owned at least 50% by foreigners. The same reasoning was applied to distinguish if the bank is publicly or privately owned. Bankscope contains information about the banks’ owners. I manually completed the missing data from banks’ web-sites.
In order to have comparable values, because of the fact that in Bankscope bank observations are reported in different measures and different currencies, all data was transformed into millions USD and inflation-adjusted. Because most of the variables used in the regression model were ratios with small values or were percentage points, I transformed all variables in units. To exclude the impact of large outliers, all variables were winsorized at 1% and 99% levels.
Information about country-level variables was obtained from various databases such as the World Development Indicators database compiled by World Bank. I used International Financial Statistics database from IMF to get data about the monetary policy interest rate. Because of the lack of information, several countries such as Bermuda, Cuba or Sudan had to
be deleted. Also, Taiwan was excluded from my sample because of the country’s ambiguity. Several countries recognize Taiwan as an independent country so in some databases such as Bankscope information can be found, but World Bank and IMF do not recognize it so because of the missing information about macroeconomic indicators, it was deleted. For countries that had just some data missing, I manually collected it from Central Bank’s sites.
The bank-specific database was merged with databases on country-level variables yielding in a sample of 34,153 bank-year observations representing 3,389 banks from 93 countries. Detailed description about the source and construction of variables can be found in Table 1.
Summary statistics
The total sample contains 3,389 banks all over the world, 2,850 from developed countries and 539 from developing economies. The table below shows a split on type of banks by the area where they are located. It can be observed that most of the banks are from United States of America while in Europe the sample of foreign banks is similar to the sample of domestic banks.
The average return on assets for all the banks in the sample is 0.8%. Most of the banks in Europe and America from developed countries managed to have a return on assets greater than the 1.5% level. Although in some countries such as Argentina, Belarus or Uganda the interest rate was high and it changed dramatically from one year to another, the average change in interest rate for the entire sample is -0.4 percentage points.
No of banks ROA - average
Z-score - average
Change interest rate home - average
Change interest rate host - average Europe F 542 0.8% 1.53 -0.3% -0.4% D 684 0.8% 29 -0.1% N/A Africa F 53 1.2% 1.94 -0.2% -0.7% D 18 2.9% 25.7 -0.1% N/A Asia F 129 1.3% 2.79 -0.2% -0.7% D 146 0.6% 20 -0.2% N/A North America F 121 0.7% 1.43 -0.3% -0.2% D 1,498 0.8% 27 -0.06% N/A South America F 101 0.9% 1.4 -0.3% -0.4% D 82 1.4% 15 -0.6% N/A Australia F 8 1% 2.6 -0.4% -0.1% D 7 0.8% 27 -0.09% N/A
Table 3 in the Appendix presents the summary statistics for the basic, country- and bank-level characteristics for the subsample composed entirely of foreign banks. Panel A reports information on the dependent variables, return on assets and z-score. Panel B contains information about the main independent variables, namely interest rate and exchange rate. In Panel C are reported the country-level characteristics and Panel D contains information about the bank-level variables. The definition and sources of the variables can be found in Table 1. Column 1 contains the number of observations for each variable. In column 2-5 are reported the mean, standard deviation, minimum and maximum values.
The average ROA for the banks in the sample is 0.9% but its values range from -9.37% for African Banking Corporation Zambia in 2009 to 11.98% for Union Bancaire
Privee SA from France in 2006. Regarding the measures of variation for this sample, the standard deviation is quite large. For example, ROA has a standard deviation of 2.4%.
When comparing the interest rate changes for home and host country, it can be observed that on average, the interest rate in the home country went down by 0.3%, while in the host country it went down by more than 0.4%. Moreover, the average difference between the interest rate in the host and home country was on average 1.6% for the foreign banks subsample, which shows that foreign banks have to deal with larger interest rates when investing abroad. This difference may be related to the level of development of the countries taken into account in the subsample.
Unreported results show that on average, the interest rate in less developed countries is higher (10.87%) than in more developed countries (3.93%). This conclusion is made under the assumption that the percentage of foreign banks in less developed countries is higher than is more developed ones, which is also shown by Stijn, Claessens and van Horen (2013). The difference in exchange rate between the currencies used in the home and host countries is on average 0.16 USD (all currencies are hereby reported in US dollars; in case of banks reporting their financial statements directly in dollars, the exchange rate taken into account to compute the difference between the exchange rates in the two countries - home and host – is set to be equal to 1).
With respect to country macroeconomic characteristics, the GDP in the host countries grew on average by 7.8%, unlike GDP in home countries, where the growth was only 6.8%. This can also be linked to the level of development of host and home countries. Foreign banks tend to come mostly from developed countries and invest in less developed ones. Emerging countries tend to have higher GDP growth rates (in relative values) when compared to developed ones. Unreported results for this subsample show an average growth
of 10.24% for emerging countries in the sample, while this rate was only 6.23% for developed countries.
Table 4 presents the summary statistics for the interest rate change, country- and bank-level characteristics for the subsample consisting in domestic banks. The average return on assets for domestic banks is 0.8% and the standard deviation is 17.7% which means that ROA has a big volatility. Comparing to foreign banks, it can be observed that domestic banks are more susceptible to risks as their return on assets is lower and the volatility bigger. Another remark is that both the minimum value and maximum value for return on assets for domestic banks are smaller than the ones for the foreign banks.
While the minimum ROA for foreign banks is -9.37%, domestic banks have a minimum value of -5.7%. This also shows that foreign banks tend to have a worse performance than domestic banks. The result is in line with the findings of Chang, Hasan and Hunter (1998). In the domestic banks sample out of 25,426 observations, 2,302 have negative return on assets. During the analyzed period, 9% of the domestic banks had a loss in some years. From these banks, even sustainable banks from developed economies such as ABN AMRO from The Netherlands had worse performance than foreign banks. Out of the 2.302 banks which had negative ROA, 1.5470 are from United States of America. It can be argued that the crisis had an impact on them because of the fact that these banks had negative ROA only after the global crisis in 2008.
Regarding the stability of banks an opposite situation can be noticed. From the summary statistics tables for foreign banks and domestic banks (Table 3 and Table 4) it can be withdrawn that domestic banks are more stable than the foreign ones. While the average z-score for foreign banks is 1.7, domestic banks have an average z-z-score of 26.4. An important observation is about the minimum and maximum values. The lowest value for domestic
banks is 0.25 which shows a high probability of insolvency. On the other hand, 130 foreign banks out of a total of 954 have for at least one year in the period a negative z-score. Also, domestic banks have a maximum z-score of 120 which shows a decrease in failure probability and therefore it can be stated that domestic banks are more stable compared to foreign ones.
V. Results
Foreign Banks
Regarding the initial findings I am reporting the regression result in Tables 5 and 6. The four columns represent the evolution of bank performance as follows: column 1 contains the base line regression. Columns 2 and 3 control for country- and respectively bank-level characteristics and in column 4 are reported results controlled for both levels characteristics.
As it can be seen, the interest rate change influences bank’s profitability. The relation is significant at 0.1% level when bank characteristics and country characteristics are included. The coefficient for interest rate in host country remains significant even when I control for both bank- and country- characteristics. An increase in the interest rate in the host country will lead to an increase in return on assets level. This result is enforced by the fact that the difference between the foreign exchange rate and the domestic rate has a positive coefficient significant at the 5% level.
With respect to the relation between the level of development and bank’s profitability, it can be observed from the regression output that the more developed a host country is, the less profitable a bank is when profitability is measured by return on assets. Unreported results show that in developed host countries, interest rate tends to have less influence on banks’ profits, changes in interest rate in home country becoming insignificant. Also, the difference
between interest rate in home country and interest rate in host country begin to be less significant too. On the other hand, in developed countries, due to the advanced markets, the bank’s ownership becomes significant at 0.1%. This shows that privately owned banks have higher profits than publicly owned banks. In developing countries, bank’s profitability is not influenced by the bank’s ownership of home country development level, but only by changes in interest rate in host country and by the difference between home and host interest rates level, which are significant at 0.1% level.
The results shown in Table 5 are consistent with my hypotheses. It can be concluded that the change in economic conditions in the host countries has powerful influence on the evolution of foreign banks.
Giving the fact that in the foreign banks sub-sample, there are four types of banks, I chose to run the regression on each type in order to see if there are differences between banks. The four types are: commercial banks, cooperative banks, saving banks and bank holdings. From the unreported results, several conclusions can be withdrawn. First of all, there are differences between banks when taking into account the bank’s ownership. Cooperative, commercial and savings banks are statistically significantly influenced by the form of ownerhsip, unlike bank holdings which have a small coefficient, almost equal to 0. The majority of bank holdings in the sample are privately owned, only 8 out of a total of 77 are publicy owned. These banks are from China (4), United Kingdom (3) and Portugal (1). Cooperative banks and savings bank are negatively influenced by the type of ownership, while commercial banks have better profits when they are privately owned.
Because of the bank’s type, interest rate changes differently affect the profitability of the bank. Cooperative and savings banks are not significantly influced by the change in interest rate in home country. This is because this type of banks are not very banks which
have many subsidiaries across the world. On the other hand, commercial banks are significantly influenced by changes in interest rate in home country. The parent bank is usually located in developed countries and because of the large amounts traded, even a small change in interest rate in home country will affect its profitability. Therefore, because of the need to support subsidiaries in developing countries, these are also affected by changes in interest rate in home country.
The difference between interest rate in home country and interest rate in host country also significantly influences the profitability of commercial banks, unlike other types of banks. When the parent bank is located in a developed country and the subsidiary in a developing country, the bigger the interest rate in home country compared to the interest rate in host country is, the greater the profits of commercial banks. When talking about commercial banks coming from developed countries and operating also in developed countries, the difference of interest rates becomes insignificant because of the countries’ similarity.
Commercial banks are gaining the most part of the profit by the amount of deposits and the level of loans given. Due to a more commercial and diversified activity than the other types of banks, commercial banks adapt their strategy to the host country conditions and even to the people there. Therefore, changes in economic conditions in host country are influencing the bank’s profitability. When controlling for both country- and bank-level characteristics, the same relation can be observed. Commercial banks’ profitability is also influenced by the macroeconomic factors in host country, while the other types of banks are not significantly affected by the macroeconomic conditions.
The difference between the exchange rates in home and host country is affecting commercial banks’ profitability. When the currency in home country is appreciating, this type of banks will have greater profits than other types of banks.
The regressions concerning the bank’s stability are illustrated in Table 6. It can be observed that the changes in interest rates in both home and host country affect the bank’s solvency. Unlike ROA which is slightly influenced by the changes in interest rates, the bank’s solvency is highly dependent on these changes. It can be argued that a bank is affected by the interest rate in the host country because of the high interconnection of the banking system in a country. Moreover, the solvency of the subsidiary is also influenced by the macroeconomic conditions from the home country where the parent bank is located. The coefficients of the interest rate changes in both home and host country are positive and statistically significant even at 0.1% level. This shows that an increase in the interest rate will lead to a better stability of the bank. When adding country- and bank-level characteristics it can be observed that both the changes in interest rate in home and host country continue to influence the bank’s z-score, but the interest rate in host country is less important than the interest rate in the home country.
The z-score of a bank is also influenced by the difference between the foreign exchange rate and the domestic exchange rate. Foreign banks will have a greater stability when the foreign exchange rate increases compared to the domestic one, that is the currency in home country appreciates. This relation stands when I control for country- and bank-level characteristics, being significant at 1% level.
Regarding the macroeconomic factors took into account as control variables, it can be observed that a bank’s solvency is significantly influenced by the factors in the host country rather than the ones in the home country. Also, bank-level characteristics have an impact on
the stability. Unlike the profitability of banks, the solvency is not significantly influenced by the level of development in either home or host country. Furthermore, the coefficients are negative which implies that the more developed the country is, the more instable a bank will be. The reason is that in developed countries, there is also a higher competition on bank market than in developing countries. Therefore, a bank has more competitors and foreign banks tend to destabilize quicker than domestic banks because of the split of total cash between parent bank and subsidiaries.
Unreported results show that developed countries have an average interest rate of 2.74%, while undeveloped countries have a much higher rate of 11.27% when I control for outliers. The econometrical analysis shows that when the interest rate in developed countries goes up, the stability of foreign banks operating in them increases. On the contrary, stability decreases for foreign banks who operate in undeveloped countries when the interest rates goes up. This could show that a change in interest rates differently affects the performance of banks depending on the current level of interest rate. If the latter is small enough (case of developed countries), an increase will lead to higher stability, but if it is already high, it will negatively impact stability when it goes up.
Another difference consists in the fact that the solvency of foreign banks which operate in developed countries is also influenced by the change in interest rate in home country. The coefficient is statistically significant at 0.1% level which shows that the solvency of a bank is highly correlated with the economic situation in the parent bank’s country. A subsidiary’s profitability depends only on macroeconomic factors from the host country, being separated from the parent bank, but the stability depends more on economic conditions in home country.
When splitting the sample of foreign banks by the type of bank, similar results are obtained for all types. The only difference is that commercial banks are more influenced than other types of banks by the difference between the exchange rates in home and host countries. That is because of the fact that commercial banks have a wider range of products than the rest of the banks, they grant loans and attract deposits in various currencies. Also, this type of banks have large transactions with the parent bank so if there is a big change in the exchange rate in either home or host country, the solvency of the subsidiary will be influenced.
Domestic Banks
When talking about domestic banks, two main differences can be observed regarding the profitability of a bank. The results are reported in Tables 7 and 8. First of all, whether a bank is privately owned or publicly owned influences the return of assets. The coefficient is statistically significant at 0.1% level and positive. Therefore, a conclusion can be withdrawn: privately owned banks tend to be more profitable than publicly owned banks. This may be because of the fact that privately owned banks are profit oriented and more oriented to the clients’ needs unlike publicly owned banks. Banks owned by the state have worse performance because of the lack of interest concerning the profit and because of the fact that they know they are supported by the government.
As is the case for foreign banks, domestic banks are also influenced by the changes in interest rates and the coefficient is significant at 0.1% level. It can be concluded that a change in interest rate affects the profitability of a bank, if this is either foreign or domestic.
Regarding the level of development of a country, unlike foreign banks, this influences the profitability of domestic banks. When controlling for country- and bank-level
characteristics, the coefficient is still positive and statistically significant at 0.1% level. Domestic banks in developed countries have lower profits than banks in developing countries. The development of a country is also an important factor when discussing about the influence of interest rate changes on bank’s profitability.
Domestic banks from developed countries are significantly influenced at 0.5% level by the changes in interest rates, while banks from developing countries are not significantly influenced. Unreported results show that in the case of domestic banks, there are some differences between banks from developed countries and banks from developing countries. In developed countries, a bank’s profitability is highly influenced by the changes in interest rate as well as by the bank’s ownership. Like foreign banks, domestic banks in developed countries will have a higher profit if they are privately owned, rather than publicly owned. Also, all macroeconomic factors took into account in this research have an influence on bank’s profit, the coefficients being statistically significant at 0.1% level. In developing countries, these indicators have a lower influence on profits.
The split on banks’ type for the domestic sub-sample give different results than the foreign sub-sample. Regarding the ownership of banks, from unreported result it can be stated that only the coefficients for commercial banks and bank holdings are statistically significant at 0.1% level. Both of the coefficients are small and positive, that is commercial banks and bank holdings will have a higher profitability if they are privately owned. The argument for this conclusion is the same as it was in the case of foreign banks, meaning they are more client oriented and also their activities are more commercial and diversified. This type of activities will give better results if the firm is privately owned. For cooperative and savings banks, the bank’s profitability will be lower for banks owned by the private sector.
In the case of domestic banks, it can be observed that between the level of development of the country and bank’s profitability there is a negative relation. Domestic banks tend to perform better if the country is less developed. One argument for this remark could be the fact that in developed countries there are also more foreign banks and the level of competition is higher than in developing countries. Therefore, if the competition is higher, is more difficult to attract clients and to grant loans. Less loans granted lead to a lower return on assets. On the other hand, in developing countries where there is also a small ratio of foreign banks, domestic banks perform better in terms of profitability. That is also because of the fact that in developing countries people’s mentality is at a low level and they prefer to get loans from domestic banks rather than foreign ones, which have more complicated products that are not easy to understand by the people. From all types of banks, only commercial banks are significantly influenced by the country’s level of development.
As I said before, domestic banks are highly influenced by the change in interest rates. While commercial banks use a large amount of funds, changes in interest rate will lead to a greater profit. Cooperative banks have less customers and their area of operations is limited and confined to state. Therefore, interest rate changes will have a smaller impact on the profits. Commercial banks in developed countries are positively affected by changes in interest rate and the coefficient is statistically significant at 0.1% level.
There can be observed differences between domestic banks and foreign banks when talking about the stability of a bank. The obtained results concerning banks’ stability are reported in Tables 6 and 8.
While foreign banks’ solvency is not significantly influenced by the level of development in home country, the latter has a bigger impact on domestic banks. The coefficient is greater and positive which shows that the more developed a country, the greater
the stability of banks is. Also, the coefficient is statistically significant at 0.1% level. A developed economy implies a better organization of the banking system, less fluctuations so the probability of insolvency is low. On the other hand, in developing economies where there are big fluctuations of macroeconomic factors and also there is a high interest rate, bank’s z-score is lower.
Another difference concerns the ownership of the bank. Compared to foreign banks, domestic banks are significantly influenced by the type of ownership. While privately owned foreign banks are slightly better performing in terms of solvency than publicly owned banks, privately owned domestic banks are highly better than publicly owned banks. Also, the type of ownership of a bank counts more if the bank is located in a developed country, where privately owned banks have a greater solvency than publicly owned banks. On the other hand, privately owned domestic banks in developing countries are less solvent than publicly owned banks.
Changes in interest rate do not significantly influence the stability of domestic banks, but most of the macroeconomic factors do have an influence unlike foreign banks whose stability is influenced by interest rate changes. When splitting the sample in banks from developed economies and banks from developing countries, there can be observed a difference on the impact of interest rate changes. A change in interest rate will have a negative impact on bank’s solvency in developing countries but a positive influence significant at 5% level in developed countries. Because in developing economies, the banking system is not very advanced, an increase in interest rates will lead to a higher probability of insolvency.
Most of the observations presented before are still holding when we talk about the solvency of a bank, but there are some differences between commercial banks and other types
of banks. For example, while the result for the whole domestic sample regarding banks’ ownership was positive and significant, this relation holds only for non-commercial banks. The coefficient for commercial banks is not significant and almost 0. As I stated before, the country’s level of development affects the domestic banks’ solvency, but this is applicable more for commercial banks rather than other types of banks.
The changes in interest rate positively influence savings banks and bank holdings’ solvency, for commercial banks being insignificant.
VI. Robustness check
Further analyses are conducted on the baseline results (Tables 5-8) consisting in re-running the results using an alternative methodology and various sample splits. These are presented in Tables 9-18.
Weighted least squares method
The first test conducted represents the rerun of results using a different methodology. Because in the sample, most of the banks are from United States of America and the number of observations in each country varies very much, I chose to use weighted least squares so that the sample bias could be avoided. The weight represents the inverse number of observations in each country. In this manner, each country-year combination will have an equal weight. The results are presented in Table 9. The results are largely similar.
For foreign banks, regarding the profitability, it can be observed that the host’s level of development has the same coefficient, but it becomes statistically significant at 0.1% level. This shows that the level of development negatively influence the profitability of foreign banks, as my hypotheses stated. Also, the interest rate changes in home country become more
significant, showing that the high interconnection between the parent bank and subsidiary has an influence on both banks’ profitability. If the interest rate changes in home country the parent bank will be affected, but this impact will be translated to the subsidiary too. Moreover, from the significance of exchange rates difference’ coefficient, the idea stated before is sustained. It can be observed that the difference between the foreign exchange rate and domestic exchange rate becomes more important for the profitability of foreign banks.
Regarding the stability, the same conclusions can be withdrawn. From Table 9, column 2, it can be observed that the baseline results obtained with GLS are similar to the results obtained with WLS method. As the case of banks’ profitability, their solvency is also influenced by the level of development in host country. So that a foreign bank perform better in terms of stability, it is preferred the host country to be less developed than the home country. An important difference between the two methods concerns the difference between the home interest rate and host interest rate, which becomes significant. It can be argued that foreign banks stability is influenced both by macroeconomic conditions in home country and host country. If the economic conditions in home country worsen, foreign banks will have a lower solvency ratio than domestic banks.
Regarding domestic banks’ profitability, same results are obtained. The results are reported in Table 9, column 3. The major difference is that the bank’s ownership becomes more significant so we can conclude that privately owned domestic banks have higher profits than publicly owned domestic banks. The solvency of domestic banks is still significantly influenced by the bank’s ownership and by the country’s level of development. The coefficient of interest rate is not significant, but is smaller and negative which shows that changes in macroeconomic conditions inversely affect the solvency of banks.
Other measure of profitability
A second robustness test is performed regarding the measurement of profitability. Although it is argued that return on assets is the most accurate measure for profitability, as a check I used return on equity. In Tables 10 and 11are reported the results of rerunning the regressions using return on equity instead of return on assets. ROE varies from -77.33% in case of a Ukrainian bank (JSIB UkrSibBank) in 2010 to the maximum value of 50.9% in case of an Austrian bank in 2008 (Volartis Bank Austria). Observing the values for these variables for both this subsample and the total sample of banks studied in this thesis, we can conjecture that there is quite a strong correlation between ROE and ROA, but not a perfect one, since the proportions of equity and total assets varies from country to country and from bank to bank.
For foreign banks, the results are similar, but some remarks need to be done. First of all, the baseline results showed that the profitability is negatively influenced by the level of development in host country. When rerunning the regressions using ROE, it can be seen from Table 10 that the coefficient of host country’s level of development becomes significant at 1% level and is also smaller, that is the bank’s profitability is more affected.
Secondly, the level of development in home country becomes significant at 5% level. Because bank’s shares are traded on stock exchanges in home country, the level of development influences the bank’s profitability. The more developed the home country is, the higher the profits will be. The coefficients of interest rates changes in home country and of the difference between the foreign exchange rate and domestic exchange rate are greater and significant at 0.1% and 1% level respectively.
Regarding domestic banks, the results obtained using ROE are similar to the results obtained using ROA. It is clear that the level of development negatively influence the bank’s
profitability and the interest rate changes positively affects it. By using ROE, it can be observed from Table 11 that the interest rate changes’ coefficient is greater, which means a bigger impact on bank’s profitability. A difference between using the two measures concerns the bank’s ownership. Although the ownership’s coefficient is the same, when ROE is used, it becomes statistically insignificant.
Fixed Effects
Unreported results of Hausman tests show that the random-effects model is not as accurate as the fixed-effects model. As a third robustness check, I rerun the regressions using bank fixed effects and country fixed effects (both home and host country) so that unobserved heterogeneity could be accounted for. The importance of including bank fixed effects is to gain an unbiased estimation and is outlined by papers such as Gropp and Heider (2010) and Huang and Ritter (2009). Tables 12 and 13 report the results obtained from rerunning the regression with fixed effects, which are enforcing the initial findings. When I account for country fixed effects, the coefficient of the interest rate changes in host country is still significant, compared to the interest rate changes in home country which is not. Since many factors like culture, mentality and history can differ from country to country, while not varying in time, they can explain the differences in profitability measured by ROA. From these results, it can be concluded that even when I control for unobserved factors at country and bank level, the interest rate in host country influences the profits. Therefore, a foreign bank’s profitability is influenced by the macroeconomic conditions in host country.
Unreported results show that in developing countries such as Ecuador, Egypt, Jordan changes in interest rate negatively affects foreign banks. Using country and bank fixed effects, it can be noticed that the difference between the interest rate in home country and the interest rate in host country is significant at 1% level. Foreign banks will be positively
influenced by the macroeconomic conditions in home country but negatively affected by changes in economic conditions in host country, lower their profitability.
When rerunning the regression for stability with both country and bank fixed effects, the results are mainly unchanged. The coefficients of interest rate changes and of the difference between the foreign exchange rate and domestic exchange rate become insignificant. The same observation is applicable for domestic banks.
Crisis periods
A factor that may influence the results is the recent financial crisis. The sample contains banks observations from the crisis period so in order to control for system bank crisis I rerun the regression using 3 dummies for each period: precrisis, crisis and postcrisis. Tables 14-17 illustrate the results. Following Berger (2012) the crisis period is considered to be between 2008-2009 although it started in the third quarter of 2007. This is one limitation of the research because the data in the sample is annually.The precrisis period represents the years from 2001 to 2007, the crisis period is considered to be 2008-2009 and bank observations from 2010 will be included in the postcrisis period.
When controlling for the crisis period, several differences can be noticed, although most of the baseline results are still holding.
First of all, the home country’s level of development is positively and significantly influencing the bank’s profitability only in crisis periods. That is, during crisis periods, developed countries tend to be less affected than developing countries so the parent bank will support its subsidiaries. Another difference is that after the crisis, foreign bank’s profitability will also be influenced by the interest rate in home country. Looking at the difference between home and host interest rates coefficient it can be observed that also after crisis, it has
more influence. As the crisis ends, foreign banks coming from developed countries will have higher profits than foreign banks coming from less developed countries.
Regarding domestic banks, most of the initial findings are the same. The ownership coefficient is not changing so we can conclude that domestic banks’ profitability is influenced by the ownership type whether there is a crisis or not. The country’s level of development is negatively influencing the bank’s profitability in crisis periods and outside them. The interest rate changes negatively affects the profits and this influence is stronger during times of crisis. Usually, in developing countries, when they are hit by a financial crisis, the interest rate changes are dramatic. Therefore, because of the macroeconomic conditions’ worsening and because the Central Bank drastically modifies the interest rate, these changes lower the banks’ profits.
From Table 15, it can be seen that when speaking about foreign banks’ stability, the results are almost the same as in the initial findings during normal times. One notable difference is that during crisis periods, the stability is not influenced anymore by the macroeconomic conditions in home country. Also, after the crisis, the influencing factors become more significant.
In the initial results, it is shown that domestic banks’ stability is negatively influenced by changes in interest rate. However, before crisis the impact is not significant. In crisis periods and in the following few years, the influence is greater. This shows the negative impact of changes in interest rate on bank’s stability. An interesting remark is that the country’s level of development has the same influence on bank’s stability when a bank crisis starts. After the crisis ends, banks in developed countries will recover quicker and will have a higher z-score than banks from developing countries.
Mergers and acquisitions
The last robustness check concerns bank observations which had substantial changes in the growth of total assets, in order to exclude possible M&As or divestitures. Therefore, the selection criteria will depend on the assets’ growth. Banks with annual growth of total assets less than -10% or greater than 15% will be deleted. (DeJonghe and Oztekin (2013))
The main results presented in Table 18 are mostly similar. This indicates that the results were not determined by the mergers and acquisitions or by divestures. Although when including country- and bank-level characteristics in the baseline regression, the interest rate changes become insignificant, when I control for M&As and divestures, the coefficient becomes statistically significant at 0.1% level. Therefore, this shows that foreign banks’ profitability is highly influenced by the changes in interest rate in host countries. Also, the difference between the foreign exchange rate and the domestic one becomes statistically significant at 5% level. The better the macroeconomic conditions in home country compared to host country are, the higher the profits of the bank.
Results concerning the foreign banks’ stability are largely unaffected with some small differences. For example, from the initial results, it was clear that banks’ solvency is influenced by the interest rate changes in host country. The same result was obtained when banks that grow or shrink quickly are excluded, but the coefficient becomes significant at 0.1% level.
When controlling for M&As for domestic banks, the results are mostly unchanged. The interest rate changes affects the bank’s profitability and is still significant at 0.1% level. The ownership coefficient is the same, but not statistically significant. Similarly, the results about domestic banks’ stability obtained by excluding large changes in assets growth reflect
the initial findings. One notable difference is that the changes in interest rate negatively affects the banks’ stability, the coefficient being significant at 10% level. The banks with assets growth greater than 15%, because of the large volumes traded, are not substantially influenced by the changes in interest rate. On the other hand, banks that do not have a significant growth of assets are affected in terms of stability when there are changes in interest rates.
VII. Conclusions
Several papers showed that the recent increase of foreign bank presence has an impact on the host countries’ economy, because of the easier transmission of monetary shocks. Due to the high interconnection of the banking system in every country and between countries, nowadays especially international banks are influenced by various factors, from bank-level characteristics to economic conditions in other countries.
The purpose of this research was to evaluate the impact of changes in macroeconomic conditions both in home and host country on the performance of foreign and domestic banks. In the existing literature several measures were used to quantify the performance of banks. For example, Berger (2013) uses return on equity as a measure for banks’ profitability, while other authors use net interest margin. European Central Bank also argues that a more accurate measure for banks’ profitability is return on assets.
My contribution can be resumed to two major key points. First, I will take into account changes in both home and host countries in order to analyze the impact on foreign and domestic banks. From what I know, in the research conducted so far, the authors put emphasis more on the host country macroeconomic conditions and analyzed the shocks that occurred in host country. Second, it is hard to quantify the performance of a bank by only
one indicator. In my research, due to the high importance of banks in the economy, I will quantify the performance at two levels: profitability and stability.
Using data on 3,389 banks located in 93 countries, I illustrate how banks are influenced by several factors. First of all, it has to be noted that the level of development of a given country do affects the bank’s performance. As Claessens and van Horen (2009) found, foreign banks that come from developed countries will perform better in terms of profitability than banks coming from developing countries. The same relation was found regarding the stability of banks.
Moreover, a relevant finding of this research represents added value to the existing literature. When taking into account the host country conditions, it can be seen that these play an important role for banks’ profitability. When the host country is developed, foreign banks tend to have lower profits than in developing countries. Also, in developed host countries, foreign banks will have the same or lower profits than domestic banks, but the opposite relation is found in developing countries.
From all factors took into account, it can be concluded that interest rate changes and the internal market situation counts for the profitability of banks, whether they are foreign or domestic. When talking about foreign banks, the macroeconomic overview of host country conditions influences the banks’ profitability and stability, while the home country conditions do not matter that much. Though, this was a limitation of my research, because of the lack of access to the net cash-flows between the parent bank and the subsidiary.
To conclude, the most important factors influencing banks’ performance are the changes in monetary conditions and country’s level of development. Banks in developed countries are affected by different factors than banks in developing countries. Also, there is a
notable difference of influencing factors when referring to profitability or stability. In terms of stability, domestic banks will be significantly influenced by the ownership type, privately owned banks being more solvent than publicly owned banks.
The existing literature focused on developing countries. For example, Arena, Reinhart and Vazquez (2007) found small differences across domestic and foreign banks in developing countries. When taking into account both developed and developing countries, several findings shed light on this topic. For example, in developing countries, foreign banks’ stability is positively influenced by the changes in monetary policy, while domestic banks’ stability is negatively but not statistically significant influenced. In developed countries, both foreign and domestic banks’ stability is influenced in the same way.
A remark needs to be made regarding the differences between foreign and domestic banks. A foreign bank’s stability is also influenced by monetary conditions in home country, but if the bank is privately or publicly owned does not matter. On the hand, for domestic banks, the ownership is a significant matter, but there can be observed differences between banks operating in developed and developing countries. In developed countries, privately owned banks have a greater stability that publicly owned banks, while in developing countries, the opposite relation can be found.
The findings of this thesis can be improved by further research, also offering notable insights for banks which want to penetrate other markets. The advantages of foreign banks presented in this research may imply a further development of the banking system as more banks may cross the borders. From an academic point of view, further research on banks performance in a macroeconomic model should control for the net cash-flows between the parent bank and its subsidiaries. In this way, the effect of changes in economic conditions in host and especially home country will reflect a more accurate view.
References:
**European Central bank, “Beyond ROE, how to measure bank performance”, 2010
Arena, M., Reinhart, C., & Vázquez, F. (2006). The lending channel in emerging economics: Are foreign banks different? (No. w12340). National Bureau of Economic Research.
Beck, T., De Jonghe, O., & Schepens, G. (2013). Bank competition and stability: cross-country heterogeneity. Journal of financial Intermediation, 22(2), 218-244.
Berger, Allen N., DeYoung, RobertGenay, Hesna & Udell, Gregory F., 2000, “Globalization of Financial Institutions: Evidence from Cross-Border Banking Performance,” Brookings- Wharton Papers on Financial Services, vol. 3, pp. 23-158.
Berger, Allen N. & Bonime, Seth D. & Covitz, Daniel M. & Hancock, Diana, 2000, “Why are bank profits so persistent? The roles of product market competition, informational opacity and regional/macroeconomic shocks”, Journal of Banking & Finance, vol. 24(7), pp. 1203- 1235.
Berger, A. N., & Bouwman, C. H. (2013). How does capital affect bank performance during financial crises?. Journal of Financial Economics, 109(1), 146-176.
Berger, Allen N., Hasan Iftekhar and Mingming Zhou, 2009, “Bank Ownership and Efficiency in China: What Will Happen in the World’s Largest Nation?” Journal of Banking and Finance, vol. 33, pp. 113-130
Chang, Edward C., Iftekhar Hasan, and William C. Hunter, 1998, “Efficiency of Multinational Banks: An Empirical Investigation,” Applied Financial Economics, vol. 8, pp. 689-696
.
Claessens, Stijn & van Horen, Neeltje, 2012. "Being a foreigner among domestic banks: Asset or liability?", Journal of Banking & Finance, Elsevier, vol. 36(5), pp. 1276-1290.
Claessens, Stijn & van Horen,Neeltje, 2013, "Foreign Banks: Trends, Impact and Financial Stability", DNB Working Papers 330, Netherlands Central Bank, Research Department.
Dages, B. Gerard, Goldberg, Linda and Daniel Kinney, 2000. "Foreign and domestic bank participation in emerging markets: lessons from Mexico and Argentina", Economic Policy Review, Federal Reserve Bank of New York, issue Sep, pp. 17-36.
DeYoung, Robert and Daniel E. Nolle, 1996, “Foreign-Owned Banks in the United States: Earning Market Share or Buying It?” Journal of Money, Credit, and Banking, vol. 28, pp. 622–636
de Haas, Ralph and Naaborg, Ilko, 2005, "Does Foreign Bank Entry Reduce Small Firms' Access to Credit? Evidence from European Transition Economies", DNB Working Paper, No.50
de Haas, Ralph and van Lelyveld, Iman, 2010, “Internal capital markets and lending by multinational bank subsidiaries”, Journal of Financial Intermediation, vol. 19, pp. 1-25.
de Haas, Ralph and van Lelyveld, Iman, 2014, "Multinational Banks and the Global Financial Crisis. Weathering the Perfect Storm?", Journal of Money, Credit and Banking, vol.46, pp. 333-364
Goldberg, Linda S., 2002, "When Is U.S. Bank Lending to Emerging Markets Volatile?", Chapter in NBER book Preventing Currency Crises in Emerging Markets (2002), Sebastian Edwards and Jeffrey A. Frankel, editors, pp. 171-196.
