The Impact of Foreign Banks Entry on Domestic Banks Profitability in a Transition Economy.

The Impact of Foreign Banks Entry on Domestic Banks Profitability in a Transition Economy. Dorothea Schäfer DIW Berlin Oleksandr Talavera DIW Berlin February 15, 2007 The usual disclaimer applies. We thank Kirill Mokh for excellent research assistance. Corresponding author: Dorothea Schäfer, DIW Berlin, Department of Innovation, Industry and Services, Königin- Luise-Str. 5, 14195 Berlin, Phone +49-30 89789-162, Fax +49-30 89789-104 Email: dschaefer@diw.de. 1

The Impact of Foreign Banks Entry on Domestic Banks Profitability in a Transition Economy. Abstract This paper aims to analyze the effects of foreign banks entry on the behavior of domestic banking sector. The empirical material is collected from National Bank of Ukraine, which provides detailed statistics about 160 banks during 2003q2-2005q3. Using dynamic panel data estimator, we find that the banks profitability is generally associated with foreign presence. When a foreign bank enters Ukrainian banking sector, the performance of domestic banks decreases. However, there is positive relationship between domestic banks profitability and share of foreign banks assets in Ukraine. This study confirms that foreign banks foster competition, being a source for positive spillover effects as well. Keywords: Ukraine, bank competition, foreign banking. JEL: G21, L10. 2

1 Introduction Ukrainian banking sector has recently attracted considerable attention from foreign investors. In the second quarter of 2005 International Mortgage Bank entered the Ukrainian market, while Vneshtorgbank acquired Mriya Bank. In 2006 BNP Paribas acquired 51 percent of shares of Ukrsibbank and Banca Intesa SpA got control over 88.4 percent of shares of Ukrsozbank. Foreign capital participation in Ukrainian banking system increased to about 25 percent at the beginning of 2006. 1 This evidence has recently stimulated debates in economic and public policy in Ukraine. Continuing trend of acquisitions in the banking sector has lead to the issue of preliminary drafts of laws on restrictions of foreign banks entry and limitation of their activities in Ukraine. This legal barriers definitely give competition benefits to Ukrainian banks, since foreign banks are more efficient (Sturm and Williams (2004), Weill (2003)). Foreign banks bring new product, new strategic and risk-management techniques, and sound corporate governance culture, which are developed in host countries. Hence, domestic banks might benefit from the spill-over effects and improve their economic efficiency. Furthermore, foreign banks entry enhances legislative framework, financial monitoring, reduces corruption and stimulates the development of transparent intermediary operations (de Haas and van Lelyveld (2006), Green, Murinde and Nikolov (2004)). At the bank level, an important stream of literature provides us with information on globalization of financial sector (Berger, Dai, Ongena and Smith (2003), Grosse and Goldberg (1991)), and on differences in performance among banks (Havrylchyk (2006), Sarkar, Sarkar and Bhaumik (1998)). This research is related to these papers and seeks to estimate the competition effects caused by foreign presence. More precisely, we investigate how the number of foreign banks and share of foreign capital affect return on assets of their Ukrainian counterparts. Our results are consistent with recent literature that investigates the effects of foreign 1 This information has been collected from web-site http://www.banker.org.ua, accessed on October 7th, 2006. 3

banks entry on performance of domestic banks. Using 7,900 bank observations from 80 countries, Claessens, Demirguc-Kunt and Huizinga (2001) point out that increased foreign ownership may result in reduction of profitability and margins for domestic banks. Along these lines, Lensink and Hermes (2004) find similar link for developed economies. However, they suggest the opposite relationship for countries with low economic development. Exploiting Philippine data, Unite and Sullivan (2003) provide evidence that an increase in foreign bank entry and penetration is associated with a reduction in interest rate spreads and improvements in operating efficiencies, however, not affecting bank profits. In this paper, we begin empirical investigation of this question for Ukraine. We apply the System GMM estimator (Blundell and Bond, 1998) to a quarterly panel of Ukrainian banks over the 2003q2-2005q3 period. After screening procedures our data include more than 1,200 bank-quarter observations, with 140 banks per quarter. Since the impact of foreign presence may differ across categories of banks, we also consider two sample splits. Our main results can be summarized as follows. We find strong evidence of a negative association between the number of foreign banks and return on assets of Ukrainian banks. However, Ukrainian banks also have higher profits when the foreign share of total assets increases. Results obtained from sample splits confirm findings from earlier research that foreign bank entry is an important determinant of domestic banks performance. The rest of the paper is organized as follows. Next section presents short overview of the banking sector in Ukraine. Section 3 describes data and econometric results. Finally, the last section outlines conclusions. 2 Overview of Ukrainian banking sector The Ukrainian banking system is a two-tier system, which includes Central Bank (NBU) and commercial banks. In December 2005 there were 186 registered commercial banks in Ukraine. Among them, there are only 23 banks with foreign ownership. The dynamics 4

of foreign presence is described by Figure 1. While total number of banks has increased during 2003-2005, the share of foreign capital is relatively stable and fluctuates around 15 percent. The only significant increase is related with acquisition of the second largest Ukrainian bank Aval by Raiffeisen International Bank at the end of 2005. According to Moody s analysts, the quality of management and the level of staff training is improving in the Ukrainian business sector and sound corporate governance standards are gradually taking shape. That goes particularly for the country s banks too. 2 Hence, foreign investors are attracted by relatively low competition and profit prospects. In addition, after Orange Revolution there were expected numerous entries of foreign banks (Barisitz (2006)). Increasing competition in banking sector poses new challenges for the domestic banks to implement sound lending policies and risk management practices. Ukrainian banking sector is significantly underdeveloped in this respect comparing to banking sectors of Eastern European countries. The EBRD s index of banking sector reform is much lower for Ukraine (2.7), while in other transition countries like Bulgaria (3.7) or Poland (3.7) it is significantly higher. 3 However, reliable and strong legislative framework still has to be established to suppress corruption, business and political pressure on domestic banks activities. 3 Empirical Implementation 3.1 Data Set We use quarterly data on the number of licensed banks in Ukraine with and without foreign capital participation as published on the official website of National Bank of Ukraine. The NBU data set has 1,578 observations on each variable from 2003q2 to 2005q3. In order to alleviate the influence of extreme observations, bank level variables are 2 Citation: http://www.die-bank.de/index.asp?issue=122003&art=297 3 The index takes values from 1 to 4+, where 1 denotes little progress, and 4+ is full convergence to BIS standards (EBRD (2005)). 5

winsorized at the most extreme (top and bottom) one percent level of the distribution on an annual basis. In order to work only with long time series for an individual bank, we exclude all banks, which have less than half time points. 4 This reduced the number of available observations to 1,171. We measure bank s equity (K) by total own capital. In addition, credits and accounts receivable, benefit/loss in accounting period to be confirmed, and clients assets are used for loan-to-capital (L/K), profitability (ROE), and deposit-to-capital (D/K) ratios, respectively. The basic descriptive statistics for foreign and domestic banks are available it Table 1. On average, foreign banks are more profitable, larger, and have higher loan to capital ratios. The empirical literature investigating banks capital structure behavior has identified that firm-specific characteristics play an important role. We might expect that a group of banks with similar characteristics (e.g., those banks with high levels of profitability) might behave similarly, and quite differently from those with differing characteristics. We consider two different sample splits in the interest of identifying groups of domestic banks that may have similar characteristics relevant to their sensitivity to foreign banks entry. The splits are based on bank size and profitability and are based on banks average values of the characteristic lying above or below median of the sample. For instance, a bank with equity above median of the distribution will be classed as large, while a bank with equity below the median will be classed as small. As such, the classifications are mutually exhaustive. 3.2 Empirical model specification We aim to contribute to the financial literature on by considering an additional factor which may have important effects on bank behavior: foreign presence in the banking sector. In particular, we test the hypotheses whether number of foreign banks and/or 4 Series can have a maximum of ten time points. All banks that have less than five time points are newly-entered banks. 6

foreign banks share matters for domestic banks performance. 5 We parameterize each bank s ROE ratio as a function of lagged profitability, loans and deposits to total assets ratio in a current period, bank s capital in a current period and proxy measures based on the number of operating foreign, local banks and foreign assets share in local banks. Our empirical specification can be summarized as follows: 6 ROE it = γ 0 + γ 1 ROE i,t 1 + γ 2 L/K it + γ 3 D/K it (1) +γ 4 K it + ξ 1 F OREIGN t + ξ 2 LOCAL t + ξ 3 SHARE t + ε it where i and t refer to bank and time, respectively; L/K it is loans to capital ratio, D/K it is deposits to capital ratio, K it is the natural logarithm of bank s net worth, and LOCAL t is the number of domestic banks. Similarly to Lensink and Hermes (2004) we use two measures of foreign presence, F OREIGN t is the number of foreign banks, and SHARE t is the share of foreign assets to total assets for all banks. The sensitivity of banks profitability measured by the coefficients is the main focus of this paper. The negative signs of ξ 1 and ξ 2, and the positive sign of ξ 3 are expected, because increasing number of banks in a market leads to stronger competition and erosion of market share for each operating bank, while increasing foreign capital share in domestic banks may lead to greater profits due to the introduction of technology innovations and more efficient monitoring. 3.3 Results for All Banks In this section we investigate to which extent profitability depends on the number of operating foreign banks and foreign ownership in domestic banks. Initially, we investigate this issue using analysis of all Ukrainian banks. Afterwards, we move to discussion of the size and profitability. The reliability of our econometric methodology depends crucially on the validity of its instruments. We check it with Sargan s test of overidentifying restrictions, which 5 Similar hypothesis are tested by Uiboupin (2005), Lensink and Hermes (2004). 6 The theoretical modelling could be found in Hainz (2004). 7

is asymptotically distributed as χ 2 in the number of restrictions. The consistency of estimates also depends on the serial correlation in the error terms. We present test statistics for first-order and second-order serial correlation. The results are estimated using (Windmeijer, 2005) finite sample correction. The matrix of instruments for all firms estimation includes L/K t 2 to L/K t 5, D/K t 2 to D/K t 5, ROE t 3 to ROE t 6, K t 2 to K t 5 and L/K t 1 to L/K t 4, D/K t 1 to D/K t 4, ROE t 2 to ROE t 5 and K t 1 to K t 4. The results of estimating equation (1) for all banks are presented in Table 2. The estimation results prove strong negative significant relationship between domestic banks profitability and number of foreign operating banks. This outcome is in line with Claessens and Laeven (2004) They also find that more efficient foreign banks entry is associated with decrease in profitability of domestic banks, because of the increasing competition (see also Shaffer (2004)). The same evidence occurs when new domestic banks enter the market. However, the performance of Ukrainian banks is more sensitive to entry of foreign ones. The statistically significant coefficients vary from -0.0023 to -0.0011 for foreign and domestic measures, respectively. There is also a positive significant relationship between banks profitability and foreign ownership share in local banks. Thus, entrance of both domestic and foreign bank intensifies competition, domestic banks also improve their operations and benefit as share of foreign capital participation increases (Lensink and Hermes (2004)). Another important outcome is that measures based on number of banks are statistically insignificant for foreign banks, implying that competition with local banks does not influence their profitability. There are only significant positive and significant negative coefficients for loan and deposit ratios, respectively. The persistence in ROE ratio (coefficient on ROE t 1 ) is in a range from 0.322 to 0.339 for all banks, which suggests that inertia defines only a third of banks profitability. 8

3.4 Results for the Sub-samples of Banks Having established a negative relationship between banks profitability and number of foreign banks, we next investigate the behavior of different groups of banks. For this purpose we group domestic banks according to size and profitability. Results for large and small banks are presented in Table 3. There is a negative significant relationship between profitability for both groups of banks and number of foreign banks. As in the case of all banks, profitability of the large and small Ukrainian banks decreases, when more efficient foreign banks enter the market. We also confirm positive significant relationship between increasing share of foreign capital participation and banks profitability only for the group of large banks. Hence, in order to benefit from foreign business technologies, Ukrainian banks have to own substantial financial resources. Interestingly, the effect of inertia is slightly higher for the large banks (0.37) compared to the small banks (0.31). This result may imply that former are more flexible to change their behavior and market practices over time. It can be explained by the fact that small banks are often established as pocket banks that provide the majority of financial intermediary services just to one affiliated company. The results for the profitability split are presented in Table 4. As it can be seen, proxies based on number of foreign and local banks are significant only for the most profitable banks, while being statistically insignificant for the least profitable banks. This might be explained by the fact that foreign banks capture the market share of the most successful national-wide banks, while regional banks are left to serve local customers alone. There is also a noticeable significant positive relationship between the foreign capital share and profitability for the most profitable domestic banks. In summary, we may draw several conclusions from the analysis of these two subsamples. Variations in foreign presence have a strong effect on the profitability ratios of large or most profitable banks. The subsample evidence buttresses our findings from 9

the full sample and further strengthens support that foreign banks foster competition, being a source for spillover effect as well. 4 Conclusion This paper explores the link between profitability of domestic banks and foreign presence in Ukrainian banking sector. Using the sample of about 160 banks during 2003q2-2005q3, we find that entry of foreign banks decreases the profitability of local domestic banks. After splitting banks by size and profitability level this effect is particularly strong for large, small and most profitable Ukrainian banks, whereas it is marginally important for the least profitable banks. Thus, we obtained empirical confirmation of our hypothesis that growing number of foreign banks leads to the narrowing of domestic banks profitability. On the other hand, increasing share of foreign ownership in domestic banks capital positively influences profitability of large and most profitable domestic banks. These findings confirm recent trend of acquisition of large Ukrainian banks (e.g. Mriya, AVAL, Ukrsibbank, Ukrsotsbank) by foreign banks. We suggest that acquired domestic banks introduce new services, improve monitoring, decrease share of nonperforming loans and gain from technology know-how brought by parental foreign banks. As a result, profitability of acquired domestic banks may increase, because of gained operating efficiency. Our results contribute to the growing debate on the consequences of foreign entry in Ukrainian banking sector. Despite the evidence that many local banks implement new lending practices, improve corporate governance culture and extend their products range over time, there is still overwhelming number of banks, which can not sustain competition and go bankrupt. This research is the first attempt to study and interpret the possible effects of foreign banks entry on banking sector in Ukraine. We believe that our empirical results can facilitate further in-depth analysis of the issue whether Ukraine has to stay open to foreign investment into its banking sector or close its market 10

for some period of time to protect domestic growing banks. 11

References Barisitz, S. (2006), Booming, but risky: The Ukrainian banking sector - hot spot for foreign strategic investors, Financial Stability Report 12, Oesterreichische Nationalbank. Berger, A. N., Dai, Q., Ongena, S. and Smith, D. C. (2003), To what extent will the banking industry be globalized? a study of bank nationality and reach in 20 European nations, Journal of Banking & Finance 27(3), 383 415. Blundell, R. and Bond, S. (1998), Initial conditions and moment restrictions in dynamic panel data models, Journal of Econometrics 87, 115 143. Claessens, S., Demirguc-Kunt, A. and Huizinga, H. (2001), How does foreign entry affect domestic banking markets?, Journal of Banking & Finance 25(5), 891 911. Claessens, S. and Laeven, L. (2004), What drives bank competition? Some international evidence, Journal of Money, Credit and Banking 36(3), 563 83. de Haas, R. and van Lelyveld, I. (2006), Foreign banks and credit stability in Central and Eastern Europe. A panel data analysis, Journal of Banking & Finance 30(7), 1927 1952. EBRD (2005), Transition report. Green, C., Murinde, V. and Nikolov, I. (2004), The efficiency of foreign and domestic banks in Central and Eastern Europe, Journal of Emerging Market Finance 3. Grosse, R. and Goldberg, L. G. (1991), Foreign bank activity in the United States: An analysis by country of origin, Journal of Banking & Finance 15(6), 1093 1112. Hainz, C. (2004), Are transition countries overbanked? The effect of institutions on bank market entry, German Economic Review 5(2), 237 256. Havrylchyk, O. (2006), Efficiency of the Polish banking industry: Foreign versus domestic banks, Journal of Banking & Finance 30(7), 1975 1996. Lensink, R. and Hermes, N. (2004), The short-term effects of foreign bank entry on domestic bank behaviour: Does economic development matter?, Journal of Banking & Finance 28(3), 553 568. 12

Sarkar, J., Sarkar, S. and Bhaumik, S. K. (1998), Does ownership always matter? evidence from the Indian banking industry, Journal of Comparative Economics 26(2), 262 281. Shaffer, S. (2004), Comment on what drives bank competition? some international evidence by Stijn Claessens and Luc Laeven, Journal of Money, Credit and Banking 36(3), 585 92. Sturm, J.-E. and Williams, B. (2004), Foreign bank entry, deregulation and bank efficiency: Lessons from the Australian experience, Journal of Banking & Finance 28(7), 1775 1799. Uiboupin, J. (2005), Short-term effects of foreign bank entry on bank performance in selected cee countries, Bank of estonia working papers, Bank of Estonia. Unite, A. A. and Sullivan, M. J. (2003), The effect of foreign entry and ownership structure on the Philippine domestic banking market, Journal of Banking & Finance 27(12), 2323 2345. Weill, L. (2003), Banking efficiency in transition economies, The Economics of Transition 11(3), 569 592. Windmeijer, F. (2005), A finite sample correction for the variance of linear efficient two-step GMM estimators, Journal of Econometrics 126, 25 51. 13

Number of banks 175 180 185 190 2003m1 2004m1 2005m1 2006m1 Date....05.15.25 Share of foreign capital Number of banks Share of foreign capital Figure 1: Dynamics of foreign presence in Ukrainian banking 14

Table 1: Descriptive statistics Panel A: Domestic banks Obs Mean Std.Dev Min Max ROE 1,207 0.032 0.037 0.000 0.232 L/K 1,238 3.226 2.106 0.109 9.885 D/K 1,231 3.183 2.446 0.029 11.892 Capital 1,232 10.869 0.801 9.632 14.080 Panel B: Foreign banks Obs Mean Std.Dev. Min Max ROE 178 0.045 0.044 0.000 0.215 L/K 175 4.809 2.002 1.170 9.739 D/K 183 4.429 2.267 1.182 11.785 K 182 11.317 0.798 10.076 14.068 Note: Note: This table reports descriptive statistics for Ukrainian and foreign banks operating in Ukraine. The time span is from 2003q2-2005q3. We denote K as a measure of bank s equity, ROE is bank s profit/loss over equity ratio, L is credits or accounts receivable, and D is clients assets. Banks equity is reported as the logarithm of bank s own capital in domestic currency. 15

Table 2: Determinants of bank s profitability: All banks Domestic Foreign Domestic Foreign (1) (2) (3) (4) ROE i,t 1 0.3306*** 0.2111 0.3512*** 0.3222** (0.0506) (0.1987) (0.0456) (0.1537) L/K it -0.0002 0.0204* 0.0010 0.0143 (0.0036) (0.0106) (0.0033) (0.0093) D/K it 0.0041-0.0099* 0.0039-0.0092 (0.0037) (0.0057) (0.0034) (0.0073) K it 0.0038-0.0132-0.0027-0.0048 (0.0050) (0.0133) (0.0045) (0.0040) F OREIGN t -0.0023*** 0.0005 (0.0006) (0.0020) LOCAL t -0.0011* 0.0017 (0.0006) (0.0019) SHARE t 0.1406* -0.0209 (0.0779) (0.2578) N 999 147 999 147 Sargan 0.358 0.475 0.351 0.324 AR(1) -4.82*** -2.14*** -4.95*** -2.4*** AR(2) -1.51 1.37-1.15 1.67 Note: Every equation includes constant term. Asymptotic robust standard errors are reported in the brackets. Estimation using XTABOND2 module for STATA. Sargan is a Sargan-Hansen test of overidentifying restrictions. AR(k) is the test for k-th order autoregression. *significant at 10%; ** significant at 5%; *** significant at 1%. 16

Table 3: Determinants of total profitability to total assets ratio: Large and Small domestic banks grouped by size Large Small Large Small (1) (2) (3) (4) ROE i,t 1 0.3669*** 0.3137*** 0.3744*** 0.3192*** (0.0528) (0.0714) (0.0483) (0.0691) L/K it 0.0040 0.0099** 0.0053 0.0116*** (0.0032) (0.0044) (0.0034) (0.0039) D/K it -0.0023-0.0055* -0.0034-0.0072** (0.0031) (0.0030) (0.0033) (0.0032) K it 0.0000 0.0062-0.0034 0.0002 (0.0052) (0.0112) (0.0045) (0.0071) F OREIGN t -0.0018** -0.0022** (0.0008) (0.0009) LOCAL t -0.0009-0.0014 (0.0009) (0.0010) SHARE t 0.2144* 0.1173 (0.1233) (0.0989) N 521 533 521 533 Sargan 0.337 0.335 0.340 0.336 AR(1) -4.35-3.46-4.45-3.54 AR(2) -1.09.936 -.908 1.04 Note: Every equation includes constant term. Asymptotic robust standard errors are reported in the brackets. Estimation using XTABOND2 module for STATA. Sargan is a Sargan-Hansen test of overidentifying restrictions. AR(k) is the test for k-th order autoregression. *significant at 10%; ** significant at 5%; *** significant at 1%. 17

Table 4: Determinants of total profitability to total assets ratio: most profitable and least profitable domestic banks Most Profitable Least Profitable Most Profitable Least Profitable (1) (2) (3) (4) ROE i,t 1 0.2673*** 0.4262*** 0.2350*** 0.4271*** (0.0468) (0.0722) (0.0424) (0.0691) L/K it 0.0026 0.0093*** 0.0020 0.0086** (0.0047) (0.0034) (0.0047) (0.0032) D/K it 0.0014-0.0084** 0.0011-0.0084*** (0.0046) (0.0034) (0.0043) (0.0031) K it 0.0017 0.0052 0.0028 0.0046 (0.0066) (0.0052) (0.0065) (0.0047) F OREIGN t -0.0024** -0.0011 (0.0011) (0.0008) LOCAL t -0.0020** -0.0001 (0.0009) (0.0006) SHARE t 0.3109** -0.1437 (0.1244) (0.1018) N 525 524 525 524 Sargan 0.334 0.399 0.382 0.317 AR(1) -4.30-3.32-4.41-3.32 AR(2) -0.43 0.41-0.41 0.35 Note: Every equation includes constant term. Asymptotic robust standard errors are reported in the brackets. Estimation using XTABOND2 module for STATA. Sargan is a Sargan-Hansen test of overidentifying restrictions. AR(k) is the test for k-th order autoregression. *significant at 10%; ** significant at 5%; *** significant at 1%. 18