Economic Determinants and Entry Modes of Foreign Banks into Central Europe Aneta Hryckiewicz Goethe University Frankfurt Oskar Kowalewski Warsaw School of Economics (SGH) WNE Research Seminar Warsaw, 28th, February 2008 1 Agenda Motivation Methodology Results Conclusions 2 Motivation Recent trend towards financial liberalization as well as globalization of the financial markets have resulted in an increase of foreign banking operations. There is a growing literature on merits/pitfalls of extensive foreign-bank presence in the developed as well as developing countries, for example: efficiency of the domestic banking sector: Hasan and Merton (2003); Barajas et al. (2000) or Claessens et al. (2000) implications for stability: Demigrüc-Kunt et al. (1998); Levine (1999); Barth (2001) lending to small and medium-size enterprises: Clarke et al. (2002); Berger et al. (2001) There are also some contributions about motivations of foreign banks to enter other markets: Buch, 2000 Focarelli and Pozzolo, 2001 Buch and DeLong, 2003 Magri et al., 2005 Claessens et al., 2007 3 Motivation However, most studies…. – come from the US or Western European countries; – use linear approaches; small samples; – either time-series or cross-section structure of data; – focus on entries via M&As or branches (mostly US) and treat each type of organizational form in isolation; – many studies neglect the importance of local market opportunities and economics or structural characteristics of the host countries; – especially, the variables measuring profit market opportunities mentioned in the theory is either omitted in empirical analysis due to limited data availability or found to be non significant; → All of these might cause some sort of biasness in the results and underestimate or limit the potential implications for policy makers or in some cases may lead even to the impropriate conclusions for developing countries. 4 Motivation The developing countries differ in many respects from the developed economies: information asymmetries underdeveloped institutional structure different macroeconomic environment weak financial regulatory and supervisory framework → Hence, the motives driving the foreign banks to enter the developed countries must be different from the motives for the developing countries, as well as the entry modes chosen by the banks. Interesting questions: what have driven foreign banks to establish their operations in the developing economies? what factors they have taken into account and have played the most important role in the decision making process? which entry modes they have preferred given they have decided to enter a foreign market? 5 What has driven foreign banks into developing countries? Foreign bank entry occurred in almost all developing countries at the beginning of 1990s, however its pattern was not uniform Venezuela Turkey Peru Poland Mexico 1994 1994 1999 Chile 1999 Hungary Brazil Czech Republic Argentina 0% 20% 40% 60% 80% 0% 100% 20% 40% 60% 80% 100% Thailand 1994 Malaysia 1999 Korea 0% 20% 40% 60% 80% 100% 6 Source: IMF 2000, for 1994 ratio of assets of banks where foreign owners own more than 50% (in case of 1999 – 40%) of total equity to total bank assets What has driven foreign banks into developing countries? There is also no obvious pattern based on the level of development Share of foreign credit institutions as % of total assets of CIs 120 1998 2004 100 80 60 The foreign ownership assets is high in Luxemburg, Finland and United Kingdom, while low in Germany, France or Italy. In developing countries, the foreign banking ownership is high in Slovakia, Czech Republic and Lithuania, while significantly lower in Cyprus or Slovenia. 40 20 Source: ECB nia Lithu a La tv ia bu rg Ne th erlan ds Port uga l Spa in Swe d en Un it ed K ingd on Malt a Cy p ru s Cz e ch R epu b lic Pola nd Hu n gar y Slov a kia Slov e nia Esto nia Ita ly Lu xe m Aust ria Belg ium De n mark Finla nd Fra n ce Ger m any Gre e ce Irela nd 0 7 What has driven foreign banks into developing countries? There is also no obvious pattern based on the level of development The differences are more meaningful when we compare the countries worldwide Share of foreign credit institutions in the developed countries, as % of total assets of CIs 100% 80% 60% 40% 20% 0% Germany Sw eden United States Luxemburg New Zealand Share of foreign credit institutions in the developing countries, as % of total assets of CIs 100% 80% 60% 40% 20% 0% Bangladesh Source: Barth and others 2001; data for 1998 Egypt India Hungary Cambodia Macedonia 8 What has driven foreign banks into developing countries? There are also no uniform entry modes of foreign banks based on the financial development Number of subsidiaries and branches in the EU 200 branches subsidiaries 180 160 140 120 100 80 60 40 20 Lu It xe al y N mb et he urg rl a n Po ds rtu U Sw ga l ni te ed d e Ki n ng do Po m la C ze Hun nd ch ga R ry ep ub l Sl i c ov ak Au st r Be ia lg D iu m en m a Fi rk nl G and er m a G ny re ec e Sp a Fr in an c Ire e la nd 0 There are no uniform entry modes in the banking internationalization. France, Luxemburg have high number of foreign subsidiaries, while Greece, Italy or Germany have more foreign branches. 9 Source: ECB; data at 2004 What has driven foreign banks into developing countries? There are also differences in entry modes of foreign banks between developing countries The differences are more meaningful when we compare developing countries worldwide Number of subsidiaries and branches subsidiary branch ru gu ay U Tu rk ey ep ub li c H un Sl ga ov ry ak R ep ub lic Ec ua do r R Br az il C ze ch ia Bo l iv Ar ge nt in a 70 60 50 40 30 20 10 0 High number of branches in Argentina, Turkey, while in Hungary until 2004 there was none. 10 Source: Barth and others 2001; data for 2002 What has driven foreign banks into developing countries? There are also differences in banking internationalization within a bank PL HU CZ SK SL EE LT LI BG RO HR AL CS CG BH Raiffeisen UniCredit Erste Bank Banca Intensa SocGen KBC GE Money EFG ING Citigroup Swedbank/ Hansabank Volksbank Alpha Bank Commerzbank Source: RZB Group Research Report; majority stakes only; data as at 31 December, reflecting ownership structure as at August 2006 11 UA What has driven foreign banks into developing countries? There are also differences in entry modes within a bank PL HU CZ SK SL Raiffeisen B A B,A A A UniCredit A A A Erste Bank A A A Banca Intensa A A A LT LI B,A B KBC A A A A A A A A Citigroup HR AL CS S S,A S A A A A A A A A A A S S A CG BH UA A AS A A A A A A A B,A B B B S B S B,A S S S S S S Swedbank/ Hansabank A A Volksbank A A Alpha Bank Commerzbank RO A EFG ING BG S A SocGen GE Money EE A A A A B A S B S A A B A B B Source: RZB Group Research Report; ECB, national banks supervision reports; A- Acquisition; B-Branch; S-Subsidiary 12 The Importance of a Foreign Bank‘s Entry Mode may affect competitive structure of local banking systems: threatening profits and market share of domestic banks affecting price and quality of bank services in host country may affect banking structure in general through specialization promoting one entry mode may cause neglecting some banking segments involves different levels of parent bank responsibility and financial support with implications of: the parent banks who care about their exposure (can be affected by a failure or great losses of a parent‘s bank institution in a host country) local regulators who care about domestic stability (may prevent country from a crisis or at least attenuate their effects) local depositors who care about safety of their savings 13 Related Literature on Developing Countries There is a limited literature on the determinants of location choices of foreign banks among developing countries, mostly because of data unavailability: Focarelli and Pozzolo, 2001 van Horen, 2006 Claessens and van Horen, 2007 Some recent papers started focusing on the entry modes of foreign banks in the developing countries; however, again, the number of studies limits to a few papers… Cerlutti et al. (2007) 14 Our Sample based on the four Eastern European markets: Poland, Czech Republic, Hungary and Slovak „a wave“ of foreign bank entries 100% 1995 2006 80% 60% 40% 20% 0% Czech Republic Hungary Poland Slovak different policies towards foreign banking and types of banking operations initially underdeveloped financial systems and poor macroeconomic environment; however there were significant discrepancies between the countries; taking time-series‘ approach we can take into account the changes occurring in these countries, starting from the improvement of the macroeconomic environment, regulatory structure, ending up with the development of financial structure and joining the EU; we have unique dataset for foreign entry modes for each of this country also in the time-series‘ context finally, we take non-linear approach to take advantage from higher number of observations; in totally we have 1212 observations. 15 Our Model (1) Analyzing the entry determinants to analyze entry decisions into CE countries, we consider a Poisson regression where the number of entries of foreign banks at time t from a country i (29 OECD countries) into one of the CEE countries - Poland, Hungary, Czech Republic and Slovak: exp iht iht Pr(Yiht y ) y! y an entry is defined to be followed by three modes: branch, subsidiary or M&As y = number of entering banks from country i to country k at time t and Y 1ht, Y2ht, Y3ht,… Y29ht have independent Poisson distribution with parameters λ1ht, λ2ht, λ3ht,… λ29ht. it is used to assuming that iht are log-linearly dependent on the explanatory variables. Thus: ln iht 0 1Kht 2 H hit 3 Bit iht h = host countries (Hungary, Poland, Czech Republic and Slovak) i = entries from the sample (29 OECD countries) defined together as home countries Kht = a vector of variables specific for the host country Hhit = a vector of variables specific for relationship between host and home country Bit = a vector of variables specific for the home countries 16 Our Model (1) Analyzing the entry determinants moreover, we control for some unobserved host-country and time-specific effects. Hence, our error term has one or two component, depending on the specification: iht iht h or iht iht h t both components are iid distributed with zero mean clustering the standard errors on the home country‘s level Pearson test is higly insignificant suggesting that our data are Poisson distributed and our model specificantion is appriopriate 17 Our Model (2) Analyzing the determinants of entry modes to model the changes between the determinants affecting particular organizational forms we have chosen seemingly unrelated bivariate probit estimation (SURB), which allows us to control for the correlation between entry modes within a one country: Pr(Yiht 1) f ( K ht , H iht , Bit ) Pr(Yiht 1) f ( K ht , H iht , Bit ) where Yiht = 1 when an entry via a particular form (M&A, branch or subsidiary) from country i into country h occurs at time t versus an entry through another form from country i into country h occurs at time t, otherwise 0 h = host countries (Hungary, Poland, Czech Republic and Slovak) i = entries from the sample (29 OECD countries) defined together as home countries Kht = a vector of variables specific for the host country Hhit = a vector of variables specific for relationship between host and home country Bit = a vector of variables specific for the home countries clustering the observations on the home country‘s level 18 Testing Hypothesis – Explanatory Variables Determinants of the Model „Follow the Customer“ Non-financial FDIs to GDP (lag) Trade (export+import) to GDP „Location-Specific Factors“ net interest margin „Economic Crises“ exchange rate of a host country towards EUR „Ownership-Specific Factors“ liquid liabilities to GDP bank deposits to GDP differences in the growth rates concentration ratio distance to the host country bank freedom index legal origin of a home country creditor rights differences in efficiencies between host and home banking markets stock market capitalization country risk inflation (log) differences in overheads between home and host banking markets corporate tax rate EU-dummy 19 Results – Model (1) Exchange rate Inflation Tax rate Bank freedom index Creditor rights Liquid liabilities Bank deposits Concentration ratio Net margin Diff. in growth rates Diff. in overheads Distance (1) (2) (3) (4) -0.0276*** 0.0850 0.1049 0.1178*** (0.0087) (0.0538) (0.0704) (0.0400) 0.3402*** 0.7829*** 1.0510** 1.1431** (0.1203) (0.2712) (0.3446) (0.4703) -0.0453 0.0501 -0.0438 -0.0845 (0.0449) (0.0576) (0.0522) (0.1031) -0.4225 -1.4793** -1.2986 -2.0571 (0.4072) (0.6660) (1.1241) (1.2678) -1.6585*** -1.4742*** -1.3578** -1.4094*** (0.4902) (0.4902) (0.6600) (0.4286) -0.4307** -0.8816*** -0.9110*** -0.8652** (0.1741) (0.2748) (0.3439) (0.4001) 0.3948 0.9087*** 0.9070** 0.9320** (0.2117) (0.3092) (0.3962) (0.4601) -0.0527** -0.0123 -0.0104 -0.0614 (0.0263) (0.0351) (0.0726) (0.0835) -0.4232*** -0.6221** -0.8631** -0.4954 (0.1623) (0.2530) (0.4023) (0.3440) -0.0822 -0.1290*** -0.1260*** -0.1297** (0.0574) (0.0484) (0.0477) (0.0508) -0.0891 -0.1004 -0.0979 -0.1254 (0.1105) (0.1160) (0.1264) (0.1388) -0.7091*** -0.7617*** -0.7525*** -0.7358*** (0.2191) (0.2486) (0.2476) (0.2448) 20 Results – Model (1) Non-financial FDI (lag) English legal origin German legal origin French legal origin Scandinavian legal origin EU dummy (1) (2) (3) (4) - 0.0984 0.1453 0.9345** (0.0940) (0.1063) (0.4605) 2.1870*** 2.2097*** 2.1933*** 2.1283*** (0.7380) (0.6934) (0.7011) (0.6897) 2.5333*** 2.1406*** 2.1377*** 2.0718*** (0.5538) (0.5493) (0.5589) (0.5599) 1.6339*** 1.4131*** 1.4150*** 1.3744*** (0.5448) (0.5358) (0.5378) (0.5301) 0.9189 0.8723 0.8725 0.8262 (0.7941) (0.7701) (0.7801) (0.7705) -0.9172** -0.2890 -1.9321 -1.0557 (0.5098) (0.7489) (1.7070) (9.7948) Fdinonlag*y2000 -0.6200* (0.3355) Fdinonlag*y2001 -0.7548** (0.2976) Fdinonlag*y2003 -0.8561** (0.3846) No. of observations 1212 973 973 973 Log likelihood value -360.0716 -282.1786 -278.2254 -273.9627 0.0535 0.1823 0.0132/0.0010* 0.0007 Wald test 21 Results – Model (2) Bank freedom index Creditor rights Liquid liabilities Bank deposits Net margin Distance Concentration ratio Market Cap. Country risk Tax rate Trade Diff. in growth rates Income per capita Wald test Branch Subsidiary M&A -3.6800** -0.2572 -0.5330** (1.4437) (0.3892) (0.2147) -6.0274*** -3.0948*** -0.8189*** (0.8260) (0.7198) (0.2336) -0.6791** 0.3266*** -0.0304 (0.2918) (0.1182) (0.0611) -2.9354*** -0.4692*** 0.1062* (1.0028) (0.1278) (0.0610) -9.0712*** -0.4191* -0.0234 (3.4256) (0.2339) (0.1205) -0.4266** -0.0083 -0.1292 (0.2006) (0.1410) (0.1337) 3.1794*** 0.0856 -0.1017*** (1.2115) (0.0810) (0.0305) 0.1304** -0.0171 -0.0224 (0.0659) (0.0145) (0.0099) 0.4802*** 0.0711 -0.0714*** (0.1424) (0.0712) (0.0279) 0.3908*** -0.0469*** -0.0233* (0.1505) (0.0168) (0.0136) -0.0045 0.0141*** 0.0109 (0.0062) (0.0024) (0.0021) -0.0499 -0.0595** -0.0093 (0.0344) (0.0258) (0.0307) 0.0050*** 0.0001 -0.0003*** (0.0018) (0.0001) (0.0001) 0.3179 0.2702 0.1752 22 Results – Model (2) (1) Branch/ Subsidiary (2) Branch/ Subsidiary (3) Branch/ M&A -0.2600 -6.1753 (4) Branch/ M&A Bank free index -3.5805*** (1.2670) (0.3906) (6.2767) (0.2146) Creditor rights -5.9824*** -3.0971*** -8.5927*** -0.8269*** (0.9302) (0.7156) (2.4903) (0.2358) -0.6776 0.3299*** -0.8109*** -0.0350 (0.2988) (0.1180) (0.1605) (0.0598) -2.9569*** -0.4737*** -3.7757*** 0.1109* (1.0746) (0.1276) (0.9332) (0.0597) -9.1258** -0.4230* -11.2538*** -0.0238 (3.6072) (0.2319) (1.6624) (0.1202) -0.4231** -0.0082 -0.4877*** -0.1293 (0.1967) (0.1404) (0.1883) (0.1343) 3.2009** 0.0855 3.9711*** -0.0999*** (1.2751) (0.0810) (0.7180) (0.0303) 0.1316 -0.0173 0.1637*** -0.0221** (0.0696) (0.0143) (0.0415) (0.0099) 0.4825*** 0.0704 0.6393*** -0.0699** (0.1506) (0.0715) (0.1315) (0.0276) 0.3943** -0.0468*** 0.4736*** -0.0231* (0.1590) (0.0168) (0.0892) (0.0136) -0.0036 0.0141*** -0.0114 0.0109 (0.0058) (0.0024) (0.0071) (0.0021) -0.0493 -0.0597** -0.0247 -0.0090 (0.0339) (0.0257) (0.0397) (0.0306) 0.0050*** 0.0001 0.0062*** -0.0003*** (0.0019) (0.0001) (0.0010) (0.0001) 0.3651 0.3651 0.0992 0.0992 Liquid liabilities Bank deposits Net margin Distance Concentration ratio Market capitalization Country risk Tax rate Trade -0.5328** Non-financial FDIs Diff. growth rates Income per capita Wald test 23 Results – Model (2) (1) Bank free index Creditor rights Liquid liabilities Bank deposits Net margin Distance Concentration ratio Market Cap. Country risk Tax rate Trade (2) Subsidiary/ M&A Subsidiary/ M&A Income per capita Wald test Subsidiary/ M&A Subsidiary/ M&A Subsidiary/ M&A -0.3376 -0.5491*** 0.4846 -0.8223** 0.8016 -0.4249 (0.3669) (0.2131) (0.7726) (0.3402) (1.5960) (0.4262) -2.9704*** -0.8210*** -6.0261* -0.9175*** -17.8308** -0.9668*** (0.7292) (0.2340) (3.4881) (0.2731) (8.2145) (0.3050) 0.3050** -0.0323 1.2992 -0.3524*** 2.0305 -0.3012*** (0.1197) (0.0611) (1.4516) (0.1334) (2.0066) (0.0973) -0.4326*** 0.1100 -1.3996 0.4141*** -2.1860 0.3480*** (0.1278) (0.0618) (1.4499) (0.1370) (1.8188) (0.1217) -0.3742* -0.0224 -0.1041 -0.0456 -0.2176 -0.0802 (0.2225) (0.1212) (0.2447) (0.1062) (0.4224) (0.0997) -0.0024 -0.1309 -0.0086 -0.1725 -0.0015 -0.1707 (0.1427) (0.1346) (0.1725) (0.1396) (0.1708) (0.1386) 0.0746 -0.1042*** 0.1290 -0.0894** 0.0405 -0.0970* (0.0789) (0.0307) (0.0848) (0.0421) (0.0613) (0.0518) -0.0182 -0.0223** 0.0192 -0.0016 -0.1113 -0.0165 (0.0137) (0.0097) (0.0568) (0.0188) (0.0880) (0.0452) 0.0571 -0.0718*** 0.0246 -0.0591** -0.0443 -0.0537 (0.0644) (0.0276) (0.0822) (0.0301) (0.0487) (0.0318) -0.0443** -0.0237* -0.1488 0.0098 -0.4040 -0.0551 (0.0185) (0.0137) (0.1837) (0.0210) (0.3345) (0.0336) 0.0137*** 0.0110*** 0.0131*** 0.0084*** 0.0156*** 0.0087*** (0.0026) (0.0022) (0.0026) (0.0022) (0.0029) (0.0022) -0.0428 0.0981** -0.1090 0.0774* (0.0895) (0.0423) (0.2275) (0.0447) Non-financial FDIs Diff in growth rates Subsidiary/ M&A (3) -0.0611** -0.0090 -0.1111*** -0.0327 -0.1030*** -0.0311 (0.0260) (0.0311) (0.0348) (0.0336) (0.0389) (0.0346) 0.0001 -0.0003*** -0.0002 -0.0002 -0.0009 -0.0004* (0.0001) (0.0001) (0.0006) (0.0001) (0.0011) (0.0002) 0.0066 0.0066 0.0184 0.0184 0.0293 0.0293 24 Conclusions macroeconomic and institutional determinants had significant influence on the foreign banking our results suggest that the foreign banks were mostly encouraged by potential of the CE banking markets and low degree of financial sophistication our findings present that in the course of time and financial development, the motivations of foreign banks have changed the legal origin of the home country mattered most of the foreign bank entries were cross-border transactions choice of the organizational form mattered besides the economic characteristics of a host country, we showed that the entry mode is also determined by the type of activity a foreign bank is going to render in a host country as well as by a type of client followed finally, we showed that the factors affecting the location choice in the developing country differ from the developed countries our results are robust to several variables used by relevant literature on foreign banking 25