the changing role of international banking in development...
TRANSCRIPT
3The Changing Role of InternationalBanking in Development Finance
THE RELATION BETWEEN THE INTER-national banking industry and the develop-ing world is changing, with implications for
the growth and financial health of both sides. Sig-nificant transformation in the structure of the in-dustry, coupled with rapid economic growth andfinancial liberalization in the developing world,has created a new locus of mutual interest and newdynamics of engagement extending well beyondthe traditional realm of provision of trade creditand financing sovereigns in distress. With over2,027 local offices established in 127 developingcountries, the international banking industry nowhas the operating infrastructure and technologyplatforms to book overseas transactions from alarge network of local agencies, subsidiaries, andbranches located in developing countries. Aided bygrowing cross-border lending activity, interna-tional banks play an increasingly important—insome countries, even dominant—role in the financ-ing structure and growth prospects of developingcountries. In many developing countries, inter-national banks now provide the primary gatewaythrough which corporations, sovereigns, andbanks transfer funds abroad, borrow in short andmedium terms, and conduct foreign exchangeand derivatives operations. Foreign claims ondeveloping-country residents held by major interna-tional banks reporting to the Bank for InternationalSettlements (BIS) currently stand at $3.1 trillion andaccount for 9.5 percent of global foreign claims,up from $1.1 trillion in 2002. As of end-June 2007,developing-country residents’ deposits with interna-tional banks amounted to $917 billion, a threefoldincrease since the end of 2002.
The resilience of the relationship between inter-national banks and developing countries, however,
81
.
is being tested by the current episode of financialturmoil. The realization of how powerfully shocksto a relatively small segment of the U.S. creditmarkets spilled over to capital markets in otherdeveloped countries in the summer of 2007 andonward to emerging markets highlights the typeof new challenges policy makers and market par-ticipants are likely to face in an environment ofsecuritized credit and an increasingly interlinkedinternational banking system. Nine months intothe turmoil, it is evident that conventional policyprescriptions borne out of the experience of thestring of emerging-market financial crises of the1990s and early 2000s offer some, but not defini-tive, guidance. The fact that the primary source ofinstability this time around resides in maturecapital markets with significant global impact callsfor stronger international cooperation in monetarypolicy, banking regulation, and liquidity manage-ment, all of which need to account for the growingfinancial links between emerging and maturemarkets. Although policy coordination to date hasmainly taken the form of collaboration in liquidityprovision, policy makers, regulators, scholars, andmarket participants have begun to focus on alonger-term reassessment of the stringency offinancial regulation and the role of asset marketsin financial stability.
This chapter highlights the growing importanceof international banking activity for developmentfinance, focusing on financial intermediation, eco-nomic benefits, and financial stability conse-quences of increased presence of foreign banks indeveloping countries. It identifies the universe ofinternational banks active in developing countries;examines the characteristics of these banks in termsof country exposure, home country jurisdiction,
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 81
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
and links with global money markets; and consid-ers how international banks may serve as a vehicleof transmission of global financial shocks to devel-oping countries. The chapter also maps out thebroad policy challenges facing developing countriesin dealing with the current turmoil, while under-lining the longer-term benefits of their integrationinto global financial system.
The key messages of this chapter are high-lighted below:
• The participation of foreign banks in develop-ing countries’ financial systems has increasedrapidly in recent years. As of 2006, 897 foreignbanks had established a majority-ownershipstake in developing countries. Foreign-ownedlenders account for a particularly high propor-tion of local banking assets in two regions—70 percent in several Eastern European coun-tries, and approximately 40 percent in someLatin American countries—compared withless than 10 percent in developed economiessuch as France and Italy. The presence of for-eign banks has increased in developing regionsfor different reasons: in Sub-Saharan Africabecause of the limited reach of local bankinginfrastructure; in Europe and Central Asiaalong with regional integration into the Euro-pean Union; and in Latin America as a wayfor governments to increase openness to for-eign competition. In many countries, however,foreign bank presence was permitted after afinancial crisis with local banks suffering frommassive nonperforming loans and was moti-vated by the need to recapitalize and reestab-lish a functioning banking system. On thesupply side, home country legislation hasallowed banks to expand in foreign markets,advances in information technology haveenabled banks to automate and manage largeinformation flows across national borders,and a fundamental shift in business strategyhas brought global banks close to customersthrough local activities.
• The increased presence of foreign banks hasgenerated substantial economic benefits tosome developing countries through efficiencygains in banking systems, increased access tocapital, more sophisticated financial services,and expertise in dealing with ailing banks.Foreign banks operating in regions such as
Europe and Central Asia tend to have loweroverhead costs and net interest margins thantheir privately owned and government-owneddomestic counterparts, although the impactvaries depending on the mode of entry and thepolicy and institutional environment of thehost country. Foreign bank entry can also leadto consolidation of fragmented local bankingsystems and the realization of economies ofscale and scope. These improvements in finan-cial sector development have provided animportant avenue for increasing growth indeveloping countries.
• Like globalization in general, the increasedrole of foreign banks can also expose develop-ing countries to certain macroeconomic risks.During the current episode, such risks haveplayed out in developing countries’ greatervulnerability to foreign shocks. Preliminaryeconometric investigation establishes a statis-tically significant relationship between inter-national bank lending to developing countriesand changes in global liquidity conditions, asmeasured by spreads of interbank interestrates over overnight index swap (OIS) ratesand U.S. Treasury bill rates. A 10 basis-pointincrease in the spread between the LondonInterbank Offered Rate (LIBOR) and the OISsustained for a quarter, for example, is pre-dicted to lead to a decline of up to 3 percent ininternational bank lending to developingcountries. Evidence from the internationalsyndicated loan market already reflects thisprediction: both the number of syndicatedloans signed and the total volume of lendingdeclined considerably in the fourth quarter of2007 and first quarter of 2008 compared withthe same periods in previous years. Countriesparticularly active in interbank markets—Brazil, China, Hungary, India, Kazakhstan,the Russian Federation, South Africa, Turkey,and Ukraine—need to be concerned about thepossibility that their domestic banks will facefunding difficulties in international marketsshould liquidity pressures in interbank marketsremain at elevated levels. Also, several coun-tries in Eastern Europe and Central Asia haveexperienced rapid private credit expansion inrecent years on account of their banks borrow-ing extensively overseas and significant foreignbank presence in their credit markets.
82
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 82
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
• A balanced mix of macroeconomic and regu-latory policy measures are called for to maxi-mize the benefits of increased foreign bankpresence in developing countries. Ultimately,policies must take into account differencesacross countries in the monetary framework(such as inflation targeting), exchange-rateregime, regulatory and supervisory capability,regional integration, level of financial sectordevelopment, and nature of exposure to theinternational banking system. Because the ef-ficiency gains associated with foreign banksdepend on the mode of entry as well as onhost country factors, public policy interven-tions can enhance both competition andbanking sector efficiency. Countries that areespecially vulnerable to foreign monetaryshocks should consider establishing backstopforeign currency lines of credit or foreign cur-rency swaps to be made available to domesticbanks in the case of severe financial distress.In countries where regulatory and financialinstitutions are still developing and possiblyweak, particular attention would need to beplaced on the quality of entry requirements,by relying, for example, on home countries’regulation and prudential supervision ofbanking institutions. A high premium shouldalso be placed on the parent bank’s compli-ance with international norms and standardsregarding capital adequacy, corporate gover-nance, and transparency.
• The high level of uncertainty and anxiety inglobal financial markets calls for greater inter-national policy coordination in the areas offinancial regulation, liquidity provision, andmacroeconomic management. Although un-usual in its scale, the coordinated liquidityprovision by the Federal Reserve, the Euro-pean Central Bank (ECB), and other centralbanks in December 2007 and subsequentmonths is consistent with central banks’ com-mon goal of maintaining financial stability.Tension in global interbank markets has beenmoderated by the moves. The fact that themagnitude of the credit turmoil was not onfinancial regulators’ radar screens, however,reveals a significant shortcoming in the currentframework of financial market supervisionand regulation. This realization has, in turn,prompted a growing consensus on the need to
foster greater transparency about the natureof complex financial instruments and each in-stitution’s exposure to them, as well as theneed to somehow institutionalize market dis-cipline as a complement to regulation, as en-visaged under the third pillar of the Basel IIAccord. Toward this end, the United Stateshas launched a far-reaching rethinking of itsfinancial regulation system. In Europe, grow-ing cross-border banking consolidation isdriving increased recognition of the need forrevised regulation and supervisory arrange-ments. At the international level, lack of botha coherent cross-border banking regulatoryframework between home countries and hostcountries and guidelines surrounding thelender of last resort and crisis managementmechanism is a cause for concern. Given thatforeign bank penetration has been more ex-tensive in developing countries than in high-income countries, developing countries shouldhave a strong stake in the development of acoherent approach to the governance of cross-border banking. And though recent efforts inmacroeconomic stabilization and externaldebt management have contributed to the rel-ative resilience of developing countries duringthe recent financial turmoil, these countriesstill need to intensify efforts to monitor foreignborrowing by their banks and risk manage-ment strategies pursued by their corporationswith access to external debt markets.
Growth and transformationof international banking activity in developing countries
Although foreign banks have operated in devel-oping countries for decades, their presence
has expanded rapidly since the early 1990s. Todayinternational banks are a growing force in shapingthe economic transformation and global competi-tive position of many developing countries. Theirimportance results from the interaction of threesets of structural factors: closer integration ofdeveloping countries into the world economythrough greater trade and foreign direct invest-ment (FDI) flows that raise demand for inter-national banking services; technological advancesallowing banks to book assets, control operations,
83
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 83
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
and automate processes across the global supplychain in an integrated manner; and regulatory re-forms in both developed and developing countriesauthorizing banks based in one country to investand operate in the banking sectors of other coun-tries. These factors have resulted in a number ofimportant changes in international banking activ-ity in developing countries—the secular growthin lending exposure, a shift from cross-border tolocal-market delivery of financial services, and sub-stantial foreign investment through cross-borderacquisitions and establishment of local affiliates.
Demand for international banking services indeveloping countries (defined as services renderedby foreign banks to developing-country residents)has evolved over time in response to the changingposition of developing countries on the globaleconomic and financial stage. Attracted by theprospects of asset growth and risk diversification,foreign banks have responded eagerly in expand-ing their overseas businesses in developing coun-tries through both cross-border and local marketactivity.
Quantitatively, the most comprehensive mea-sure of international banking activity in developingcountries, total foreign claims on developing coun-tries held by banks reporting to the BIS, stood at$3.1 trillion in the third quarter of 2007 (figure 3.1),almost six times larger than in 1992, when bankswere recovering from the Latin American debt crisisof the early 1980s.1 Sixty percent of this exposureis in international claims (claims denominated in
foreign currency), including cross-border loans andloans extended by banks’ foreign offices, mostly toresidents of countries in Latin America, East Asia,and Europe and Central Asia (figure 3.2). Despite asteady shift in international banks’ strategy fromcross-border lending to lending through localaffiliates, their exposures to developing countriesremains mostly denominated in foreign currency, ofwhich about 44 percent are in short-term maturity.
Because foreign-denominated exposures aretypically funded in international markets, theytend to be highly sensitive to movements in globalinterbank rates and conditions. Furthermore,exposure to foreign-currency loans is widespreadacross developing-country borrowers, with a ma-jority of borrowers (77 percent) holding morethan half their foreign bank debt in loans denomi-nated in foreign currency.
The strong overall growth in internationalbanking has been interrupted, however, by severalepisodes of credit contractions and economicdownturns. Scaled by aggregate GDP of developingcountries, a measure that serves as a proxy fordemand-side factors, international bank claimsdeclined sharply in the late 1980s and early 1990s(to 13 percent of GDP in 1992), increased steadilythrough the remainder of the 1990s, paused duringthe global slowdown of 2001–02, and mostly ac-celerated since 2003 (reaching 23 percent of GDPin 2007). The latest expansion—from 2003 untilthe onset of global financial turmoil in mid-2007—coincided with an epoch of excessive global liquid-ity, large-scale securitization, and cross-borderbanking sector consolidation (box 3.1).
84
0
500
1,000
1,500
2,000
2,500
3,500
1983
1986
1989
1992
1995
1998
2001
2004
2007
8
12
16
20
24
Percent
% of GDP(right axis)
$ billions
3,000
Figure 3.1 International bank claims ondeveloping countries
Sources: Bank for International Settlements (BIS); World Bank.
Note: These are the foreign assets of banks reporting to the BIS.GDP is aggregate GDP for developing countries.
Middle East andNorth Africa 4%
South Asia 7%
Sub-SaharanAfrica 6%
Figure 3.2 International claims outstanding, byregion, third quarter, 2007
Latin Americaand the
Caribbean20% Europe and
Central Asia43%
East Asiaand Pacific
20%
Sources: Bank for International Settlements; World Bankstaff calculations.
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 84
The international banking industry has witnessed phe-nomenal growth and financial innovation over the
past two decades, punctuated by episodes of consolidation.The spread of modern international banking is convention-ally traced to the establishment of the Eurocurrency mar-ket in the late 1950s and early 1960s, initially in Londonand then in other European financial centers. As measuredby foreign assets of banks reporting to the BIS, interna-tional banking activity expanded at a very fast pace overthe past decade, reflecting expanding world trade, the riseof multinational firms, growth in financing of globalpayments imbalances, and the assimilation of transitioneconomies into global banking system (figure below).Looking back, international banking has gone throughthree distinct phases in the post–World War II era:
• The establishment of the Eurocurrency market in thelate 1950s and early 1960s, stimulated initially by pre-vailing capital controls and restrictions on internationaltransactions in the United States and Western Europe,which prompted national banks to establish officesabroad to service the overseas business of their clients.
• The growing role of banks in Japan in the 1980s asthe Japanese government attempted to open its mar-kets and promote the international role of yen. Thisphase also coincided with the growth of syndicatedbank lending and the expansion of currency andinterest-rate derivatives markets that enhanced banks’scope to expand their geographical reach in bothfunding and lending.
• The increased securitization of credit in recent years,facilitated by the originate-and-distribute model ofbank lending on the one hand and by rapid growthin the market for asset-backed structured financialproducts (such as collateralized debt obligations) anddevelopment of the credit derivatives market on theother. From a public policy perspective, securitizationhas contributed to a shift in regulatory or oversight
responsibility from official agencies to the privatemarketplace, including credit rating agencies andsecurity underwriters.
A wave of cross-border mergers and acquisitions over thepast decade or so has resulted in a significant consolidation ofthe international banking industry and a concentration ofassets in the hands of a few major banks. As of 2007, the top10 banks held 19 percent of the industry’s assets, and the top100 banks accounted for 75 percent, higher than the corre-sponding values of 13 and 59 percent in 1996 (figure below).
Financial innovation and technological change pio-neered by the banking industry itself has transformed thenature and reach of the international banking business,allowing banks greater market reach and new businessareas, including underwriting, asset management, invest-ment banking, and proprietary trading. Rapid growthof the markets for risk transfer—credit derivatives andvarious types of asset-backed securities—has facilitatedhighly leveraged exposures by banks themselves and bynew players such as hedge funds and private equity firms.
Box 3.1 Rapid expansion of the international banking industry
Top 10 banks Top 50 banks Top 100 banks0
10
30
20
40
50
60
70
80
Banking consolidation has increased over time
% of assets
1996 2007
Source: World Bank staff calculations based on The Banker database.
1970
1990
1986
1978
1982
1974
2006
1994
1998
2002
35
30
0
20
15
10
5
25
International banking expansion, 1970–2007
$ trillions
0.1 in 1970
6.3 in 1990
31.8 in 2007
Source: Bank for International Settlements (BIS).
2001 2002 2003 2004 2005 2006 20070
5
15
10
20
25
30
35
40
45
50
Significant expansion in the credit derivatives market
$ trillions
Source: International Swaps and Derivatives Association 2007.
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 85
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
The regional composition of creditor banks todeveloping countries has also changed since theearly 1990s. Largely reflecting the growing weightof claims by residents of Eastern Europe and CentralAsian countries, the role of Western European bankshas increased, accounting for 73 percent of totalforeign claims on developing countries in 2007,compared with 62 percent in 1999 (figure 3.3). Bycontrast, banks from Japan and the United Stateslost market share during this period as they adopteda more cautious approach to overseas expansion.
International banks service their overseasbusinesses through local market participationForeign banks’ direct investment in developingcountries’ banking sectors accounted for a cumula-tive $250 billion over 1995–2006, fueled by bothgreenfield (new) investments and mergers and ac-quisitions (M&A).2 As of end-2006, the 897 foreign
banks with a presence in developing countries con-trolled combined assets of over $1.2 trillion andaccounted for more than 39 percent of total bankingassets in these countries (figure 3.4), compared with$157 billion 10 years earlier, when they accountedfor approximately 20 percent of total bankingassets. Since 2000 the majority of the increase inassets has resulted from increased banking sectorconsolidation and better economic integrationbetween existing and new EU members. Indeed, thenumber of foreign banks in the countries that joinedthe European Union in 2004 jumped from 121 in1995 to 330 in 2006, and the value of their assetssurged from $41 billion to $528 billion.
The share of banking assets held by foreignbanks with majority foreign ownership stake,however, varies dramatically among developingregions and is to some extent dependent on regula-tory restrictions. Overall, foreign ownership of thebanking sector is substantially higher in Europeand Central Asia, Sub-Saharan Africa, and LatinAmerica than in East Asia, South Asia, and theMiddle East and North Africa (figure 3.5). Foreignownership also varies considerably intraregionally.While many small Sub-Saharan African countrieshave shares exceeding 50 percent, Ethiopia,Nigeria, and South Africa have minimal or no for-eign bank participation with majority foreignownership stake (table 3.1). In Latin America, largeeconomies such as Peru and Mexico have foreignpresence accounting for 95 and 82 percent of the
86
Figure 3.3 Composition of foreign claims ondeveloping countries, by nationality of reportingbanks
European banks73.4%
United States 11.9%
Other countries 8.9%
Other countries 10.8%
Japan 4.3%
Japan8.3%
Canada 1.5%
Canada 2.0%
2007
1999
European banks61.7%
UnitedStates17.2%
Sources: Bank for International Settlements; World Bank staffcalculations.
Note: European banks include those from Austria, Belgium,Denmark, Finland, France, Germany, Greece, Ireland, Italy,the Netherlands, Norway, Portugal, Spain, Sweden, Switzerland,and the United Kingdom.
500
900
850
800
750
700
650
600
550
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
Percentage of total assets
Figure 3.4 Foreign banks’ increasing involvementin developing countries, 1995–2006
Number
0
5
10
15
20
25
30
35
40
45
50
Market sharein assets
(right axis)
No. of foreign banks(left axis)
Source: World Bank staff estimates based on data from Bankscope.
Note: Foreign banks are those in which foreign shareholders hold50 percent or more of total capital. Because asset data for 2006 aremissing for a significant number of banks, asset information ispresented only to 2005.
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 86
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
banking sector, respectively, while in smalleconomies such as Guatemala and Ecuador, theshare is 8 and 5 percent, respectively. WithinEurope and Central Asia, foreign banking pres-ence is low in the two largest regional economies,Russia and Turkey, but extensive in most other
countries. In recent years banks from developingcountries have begun to invest in other (particu-larly low-income) developing countries. And as of2006, 256 of the 897 foreign banks operating indeveloping countries were based in other develop-ing countries. Typically, these foreign banks arefrom middle-income countries such as Hungary,Malaysia, and South Africa, and like their high-income competitors they invest mainly within theirown regions.
International banks tend to seek out marketswhere institutional familiarity provides them witha competitive advantage over other foreign banks(Claessens and Van Horen 2008). As such, foreignbank penetration tends to be particularly high indeveloping countries with similar legal systems,banking regulations, and institutional setups ascertain home countries, presumably because suchsimilarities tend to reduce risk and operationalcosts (Galindo, Micco, and Serra 2003). Foreignbank presence also tends to follow lines of eco-nomic integration, common language, and geo-graphical proximity. In Latin America and theCaribbean, for example, 60 percent of foreignbanks are headquartered in the United States andSpain, whereas in Europe and Central Asia morethan 90 percent of foreign banks are headquar-tered in the European Union (figure 3.6). Even
87
All dev
elopin
g
regio
ns
South
Asia
Midd
le Eas
t and
North
Afri
ca
East A
sia a
nd
Pacific
Latin
Am
erica
and
the
Caribb
ean
Sub-S
ahar
an A
frica
Europ
e an
d
Centra
l Asia
0
10
20
30
40
50
60
Figure 3.5 Share of banking assets held byforeign banks, by region
Percent
1995 2000 2005
Source: World Bank staff estimates based on data from Bankscope.
Note: Foreign banks are those in which foreign shareholders hold50 percent or more of total capital.
Table 3.1 Share of banking assets held by foreign banks with majority ownership, 2006
Country 0%–10% Country 10%–30% Country 30%–50% Country 50%–70% Country 70%–100%
Algeria 9 Moldova 30 Senegal 48 Rwanda 70 Madagascar 100Nepal 9 Honduras 29 Congo, Dem. Rep. of 47 Côte d’Ivoire 66 Mozambique 100Guatemala 8 Ukraine 28 Uruguay 44 Tanzania 66 Swaziland 100Thailand 5 Indonesia 28 Panama 42 Ghana 65 Peru 95India 5 Cambodia 27 Kenya 41 Burkina Faso 65 Hungary 94Ecuador 5 Argentina 25 Benin 40 Serbia and Montenegro 65 Albania 93Azerbaijan 5 Brazil 25 Bolivia 38 Cameroon 63 Lithuania 92Mauritania 5 Kazakhstan 24 Mauritius 37 Romania 60 Croatia 91Nigeria 5 Pakistan 23 Burundi 36 Niger 59 Bosnia-Herzegovina 90Turkey 4 Costa Rica 22 Seychelles 36 Mali 57 Mexico 82Uzbekistan 1 Malawi 22 Lebanon 34 Angola 53 Macedonia 80Philippines 1 Tunisia 22 Nicaragua 34 Latvia 52 Uganda 80South Africa 0 Mongolia 22 Chile 32 Jamaica 51 El Salvador 78China 0 Sudan 20 Venezuela, R. B. de 32 Zimbabwe 51 Zambia 77Vietnam 0 Morocco 18 Georgia 32 Namibia 50 Botswana 77Iran, Islamic Rep. of 0 Colombia 18 Armenia 31 Kyrgyzstan 75Yemen, Rep. of 0 Malaysia 16 Poland 73Bangladesh 0 Jordan 14 Bulgaria 72Sri Lanka 0 Russian Federation 13 Paraguay 71Ethiopia 0 Egypt, Arab Rep. of 12Togo 0
Source: World Bank staff estimates based on data from Bankscope. Note: A bank is defined as foreign owned only if 50 percent or more of its shares in a given year are held directly by foreign nationals. Once foreign ownership isdetermined, the source country is identified as the country of nationality of the largest foreign shareholder(s). The table does not capture the assets of the foreignbanks with minority foreign ownership.
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 87
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
excluding HSBC, which moved its headquartersfrom Hong Kong (China) to the United Kingdomin 1993, Asian banks account for 40 percent offoreign banks in East Asia. In Sub-Saharan Africa,more than 30 percent of foreign banks are fromthe region, and the rest are mainly from countrieswith which Sub-Saharan Africa has had economiclinks since colonial times.
The regional focus of banks investing in devel-oping countries is also evident in data on the 20foreign banks with the largest asset holdings indeveloping countries. For example, all majority-owned foreign banking assets of two Spanishbanks, Santander and BBVA, and Canadian ScotiaBank, are in Latin America. Other Europeanbanks, including Italy’s Unicredito and IntesaSanpaolo and Austria’s Erste Bank, Raiffeisen, andHVB, have a significant presence in the Europeand Central Asia region. On the other hand, top20 banks such as BNP Paribas (France), ING(Netherlands), Deutsche Bank (Germany), andCitibank (United States) are more diversified. Allin all, developing countries still account for a rela-tively small share of these banks’ total assets, rang-ing from 1 to 15 percent.
The mode of foreign bank entry has shiftedfrom greenfield investments to M&A and frombranches to subsidiariesCross-border consolidation has been an importantdriver of recent expansion in the amount of FDIin developing countries’ banking sectors. Availabledata show about 750 cross-border M&A trans-actions in developing countries over 1995–2006,totaling $108 billion.3 Meanwhile, the share ofglobal cross-border M&A transactions involvingbanks based in developing countries rose from 12percent in 1995–2002 to 21 percent in 2003–06.The size of these transactions varied considerably,however. The largest was Citigroup’s acquisitionof Mexico-based Banamex (table 3.2). M&Atransactions resulting in majority ownership ac-counted for 407 of 587 recorded entries of foreignbanks in developing countries during 1995–2006(figure 3.7). The share of M&A in total foreignbank entry has jumped dramatically—to approxi-mately 90 percent—since 2004.
When a foreign bank enters a country throughM&A, it generally operates as a subsidiary—alegally independent entity with powers defined by
88
Figure 3.6 Home countries of foreign banks indeveloping regions, 2000–06
United States 6%
Other 3%
Austria28%
OtherEuropean
Union63%
Latin America and the Caribbean
Europe and Central Asia
Portugal8%
Europe 5%
United States 5%
Sub-Saharan Africa
East Asia and the Pacific and South Asia
Source: World Bank staff estimates based on data from Bankscope.
Other10%
Asia(developingcountries)
16%
UnitedKingdom
21%
UnitedStates17%
Europe15%
Other 7%
Asia(high-income
countries)24%
Europe(ex. Spain)
22%
Spain40%
UnitedStates21%
Canada7%
Other7%
Africa (ex. South Africa) 8%
SouthAfrica24%
UnitedKingdom
26%
France17%
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 88
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
its own charter in the host country. In the case ofa greenfield investment, however, the foreign oper-ation may be either a branch or a subsidiary. Abranch is licensed by the host country but itspowers are defined by its parent bank’s charter(subject to limitations imposed by the host coun-try). Subsidiaries seem to be strongly preferredby the 100 largest foreign banks in Latin Americaand Eastern Europe, where they accounted for65 and 82 percent, respectively, of local operationsin 2002 (Cerrutti, Dell’Ariccia, and Martinez Peria2005; Baudino and others 2004).
The decision to enter a developing countythrough a branch or a subsidiary is found to be
affected by several host country factors and thenature of the foreign bank’s business (Cerrutti,Dell’Ariccia, and Martinez Peria 2005). Regula-tions and institutional factors are of paramountimportance in the decision, as foreign banks areless likely to operate as branches in countries thatlimit their activities. In some cases, the organiza-tional structure is shaped by government policiesfavoring one form over the other, for example, inMalaysia, Mexico, and Russia, where investmentthrough branches is not allowed. When branchesare allowed, they are most common in countrieswith high corporate taxes and in poor countries,perhaps in the latter because of lack of market op-portunities. The bank’s desired business in the hostcountry market is also an important factor:branches are more prevalent than subsidiarieswhen foreign operations are small in size and donot provide retail services. Branches are less com-mon in countries with risky macroeconomic envi-ronments. However, when the risks are mostlyrelated to government intervention or other politi-cal events, foreign banks may prefer to operate asbranches.
The distinction between branches and sub-sidiaries also implies different levels of parent-bank responsibility and financial support. Whilesubsidiaries are legally separate entities from theirparent banks, parent banks are responsible forthe liabilities of their branches under most circum-stances. Parent-bank support can play an impor-tant role during times of financial turmoil. Forexample, following the financial crisis in Argentinain the early 2000s, Citibank increased the capitalof its branch operations in the country but sold itssubsidiary there. This said, special contractual
89
Table 3.2 Major cross-border M&A sales by developing countries, 2001–07
Year Acquired bank Host country Acquiring bank Home country % of the asset bought Value ($ billions)
2001 Banamex Mexico CitiGroup United States 100 12.52007 ICBC China Standard Bank South Africa 20 5.52006 BCR Romania Erste Bank Austria 62 4.82006 Akbank Turkey CitiGroup United States 20 3.12005 Bank of China China Merrill Lynch United States 10 3.12004 Bank of Communications China HSBC United Kingdom 20 2.12005 Disbank Turkey Fortis Belgium 90 1.32001 Banespa Brazil Banco Santander Spain 30 1.22005 Avalbank Ukraine Raiffesen Austria 94 1.1
Source: World Bank, Global Development Finance, various years.
Source: World Bank staff estimates based on data from Bankscope.
Note: Foreign banks are those in which foreign shareholders hold50 percent or more of total capital.
Figure 3.7 Mode of entry of foreign banks withmajority ownership
No. of bank entities
0
20
40
50
60
70
80
10
30
20061996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Greenfield investments
Merger and acquisitions transactions
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 89
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
agreements (such as ring-fencing provisions) andreputational considerations may at times blur dis-tinctions between branches and subsidiaries. Forexample, in recent years, a number of bankinggroups have adopted ring-fencing provisions thatgenerally establish that parent banks are not re-quired to repay the obligations of a foreign branchif the branch faces repayment problems because ofextreme circumstances (such as war or civil con-flict) or because of certain actions by the hostgovernment (such as exchange controls, expropri-ations, and the like).4 However, concerns aboutloss of reputation have in certain instances led par-ent banks to rescue and recapitalize subsidiaries,even if they were not legally forced to do so. Forexample, HSCB injected a significant amount ofcapital into its subsidiary in Argentina followingthe crisis. Portugal’s Banco Espiritu Santo did thesame for its Brazilian subsidiary following thelosses due to the real’s devaluation in 1999(Cerutti, Dell’Ariccia, and Martinez Peria 2005).
Foreign bank expansion has beenfostered by financial liberalizationand deregulationSince the mid-1990s, restrictions facing foreignbanks, including limitations on form of investmentand level of foreign ownership, have been gradu-ally eased through unilateral liberalization policies,bilateral and regional trade and investment agree-ments such as the North American Free TradeAgreement (NAFTA), and World Trade Organiza-tion (WTO) membership requirements. In particular,the General Agreement on Trade in Services (GATS)encourages greater openness among WTO mem-bers in provision of financial services from foreignentities. The agreement addresses 17 specific issuesrelated to foreign bank presence in member coun-tries, including foreign bank entry and licensing
requirements (such as minimum capital entryrequirements), method of entry, expansion afterentry, limitations on share of foreign presence inthe banking sector, and permissible activities andoperations. A close examination of reported prac-tices, however, indicates that some developing-country members of the WTO are more restrictivein practice than they should be according to theirWTO commitments (Barth and others 2008).
In many countries, financial sector liberali-zation came after a financial crisis and was moti-vated by the need to reestablish a functioningbanking system (Cull and Martinez Peria 2007). Ingeneral, though, the driving forces behind and tim-ing of financial sector liberalization—and the levelof allowed foreign ownership (table 3.3)—continueto vary considerably among developing countries.5
In the early 1990s many countries in theEurope and Central Asia region allowed foreignbanks to start operations within their borders onlythrough greenfield investment (through licensing)and through purchase of minority stakes in localbanks. Majority ownership was allowed only afterbanking crises hit many of these economies(Baudino and others 2004). Although foreign bankentry was pervasive in the early 1990s for 2004EU accession countries (in particular Hungary andPoland), it occurred later in the 2007 accessioneconomies, Bulgaria and Romania (Hagmayr, Haiss,and Sümegi 2007). In Turkey foreign banks in-vested significantly only after the start of the coun-try’s official EU accession negotiations in 2005.
Most Latin American countries began open-ing their banking systems to foreign entry follow-ing a series of financial crises in the region in themid-1990s (ECLAC 2002). In Mexico, for exam-ple, all banks (except one foreign bank) werenationalized in 1982 and remained under statecontrol until a progressive easing of restrictions
90
Table 3.3 Foreign ownership restrictions in banking sector, 2004 or latest available year
Percentage allowed Country
Not allowed Ethiopia
1%–49% Algeria, China, India, Indonesia,a Kenya, Pakistan, Sri Lanka, Thailand, Uruguaya
50%–99% Brazil, Arab Republic of Egypt, Malaysia, Mexico, the Philippines, Poland, Romania, Russian Federation
No restrictions Argentina, Bolivia, Chile, Colombia, Costa Rica, Dominican Republic, Ecuador, El Salvador,Guatemala, Hungary, Jamaica, Mauritius, Mongolia, Morocco, Mozambique, Nigeria, Paraguay,Peru, Republic of Korea, Senegal, South Africa, Trinidad and Tobago, Tunisia, Turkey, Uganda,Tanzania, República Bolivariana de Venezuela
Source: UNCTAD 2006.a. Denotes 100 percent minus the government ownership percentage, that is, the share of business held by the private sector.
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 90
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
in the 1990s.6 Similarly, in Argentina foreignbank entry was permitted starting in the early1990s but the privatization of state banks accel-erated in the fallout of Mexico’s Tequila crisis.By contrast, in Brazil, where restrictions wereeased in the late 1990s, foreign bank entry is stillevaluated on a case-by-case basis (Peek andRosengren 2000).
Other regions remain relatively less open toforeign bank entry (figure 3.8), although manyEast Asian countries, including Indonesia, Thai-land, and the Philippines, lowered barriers tobanking sector FDI following their 1997–98 finan-cial crises (Coppel and Davies 2003). In China,where banking sector FDI traditionally has beenlimited, the country has recently taken stepstoward liberalization in order to meet its WTOcommitments.7 Countries in South Asia and theMiddle East and North Africa also tend to haverelatively high restrictions on foreign bank entry.India, for example, provides a limited number oflicenses for opening branches and permits foreignbanks to hold only a 5 or 10 percent equity stakein domestic private banks (and this only since
2005), with a few exceptions for stakes in selecteddomestic banks. Further liberalization for foreignbank acquisitions is expected in 2009. The ArabRepublic of Egypt and Algeria have notable restric-tions on foreign investment, although Moroccoand Tunisia have no restrictions.
Technological progress has facilitated FDIin the banking industryInnovations in data transmission, storage, andprocessing have facilitated the unprecedentedgrowth of FDI in emerging economies’ bankingsectors. Reliable global payment systems and real-time settlement systems across time zones haveallowed intermediaries to increase the efficiencyof back-office operations, thereby freeing up re-sources for front-office activities that permit themto enter new markets. Predictably, however, banksfrom developed countries have a marked advan-tage over local banks in developing countries inadopting new technologies because of easier accessto required expertise and the economies of scaleinvolved in already having absorbed the very highfixed costs of deploying the same technologies intheir home operations.
Commentators have identified four areas inwhich technological progress has been especiallyimportant for the geographic expansion of banks.First, the dawn of market-segment and bank-specific credit-scoring methodologies, combinedwith the collection of borrower-specific informa-tion through credit bureaus, has allowed banks tomore efficiently assess the creditworthiness of cus-tomers in new markets. As a result, banks havebeen able to lend over greater distances in boththeir home and foreign markets. Second, impor-tant innovations in risk management systems,often driven by the Basel II Accords, have allowedbanks to increase the size of their balance sheetsfor a given capital base. Improvements in the quan-tification of expected losses for both individual po-sitions (through credit scoring, for example) andaggregate exposures (through value-at-risk analy-sis, for example) and the analysis of balance-sheetbehavior under alternative market scenarios haveenabled banks to better account for the risks ofmoving into new markets. Third, improved instru-ments for securitization and hedging have helpedbanks better manage their international risk expo-sure (Barth, Caprio, and Levine 2001). Finally,new ways of collecting deposits and interacting
91
Latin
Am
erica
and
the
Caribb
ean
0
0.10
0.40
0.30
0.20
0.50
0.60
Figure 3.8 Restrictions on FDI in the bankingsector, 2005
Index
East A
sia
and
Pacific
South
Asia
Sub-S
ahar
an A
frica
Midd
le Eas
t and
North
Afri
ca
Europ
e an
d
Cen
tral A
sia
Source: World Bank staff estimates based on data from UNCTAD(2006).
Note: Regional averages are the simple average of the index for eachcountry within the region. The country index is measured on a0–1 scale, with 0 representing full openness and 1 a de factoprobation. The index is based on government policies related toforeign bank ownership restrictions, screening and approval, andoperational restrictions (in the order of highest weighted restrictionto the lowest).
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 91
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
with customers—the Internet, automated tellermachines (ATMs), and mobile phones—have im-proved access to finance for unserved or under-served residents of countries such as India, Kenya,the Philippines, South Africa, and Zambia.
Economic benefits of internationalbanking
Developing countries stand to reap substantialgains from their increased engagement with
the international banking industry. Access tointernational banking increases potential sourcesof credit to firms and households, enhances pro-vision of sophisticated financial services, andencourages efficiency improvements in domesticbanks, although the impact of all of these factorsvaries depending on the characteristics of banksand the policy and institutional environment ofhost countries. As a result of these influences, in-creased international banking in developing coun-tries has helped ease credit constraints on firms,thereby contributing to growth and development.
Foreign banks have improved accessto financial servicesThe ability of international banks to frequentlyoffer more sophisticated, higher-quality, andlower-priced services than domestic banks todeveloping-country borrowers derives from severalfactors, including access to the technology, thepresence of skilled personnel, and the ability toseize opportunities of scale in operational systemsalready in place in providing services to their do-mestic clients. For example, Arnold, Javorcik, andMattoo (2007) document that foreign banks in theCzech Republic were the first or leading banks tooffer ATM transactions and remote banking andthat they have greatly sped up the process of loanapplications. Garber (2000) notes the ability of for-eign banks to offer new financial products such asover-the-counter derivatives, structured notes, andequity swaps. Levine (2001) cites a dramatic reduc-tion in fees on letters of credit and letters of guaran-tee in Turkey following liberalization of bank entryrules. And Wooldridge and others (2003) highlightthat foreign banks have also supported the devel-opment of local financial markets in many develop-ing countries, particularly in local securities andderivatives markets by investing considerable
capital and expertise. Foreign banks participate asprimary dealers in some local government bondmarkets, and as pension fund managers and swapdealers in other markets.
Increased foreign bank presence can alsoimprove the soundness of the financial system byencouraging stronger regulation and supervision.Numerous studies have found that investmentsby foreign banks in developing countries spurimprovements in bank supervision, with spillovereffects that improve the structure of regulation(Goldberg 2004). Levine (2001) argues that for-eign banks may encourage the emergence of insti-tutions such as rating agencies, accounting andauditing firms, and credit bureaus, citing theexample of improvements in supervision andaccounting standards in Mexico as a consequenceof opening the banking sector to U.S. institutionsunder NAFTA.8 Foreign bank entrants also canbring more advanced safeguards against fraud,money laundering, and terrorism financing, anddomestic banks may emulate such safeguards togain a competitive advantage in access to interna-tional financial markets.
Foreign banks have improved the efficiencyof domestic financial systemsThe entry of foreign banks may improve the effi-ciency of financial systems in developing countries,either because foreign banks are more efficientthan their domestic counterparts or because com-petition from foreign banks in formerly protectedand oligopolistic markets forces domestic banks toimprove their own efficiency.9 Adequate levels ofcompetition are generally viewed as important toreducing costs and increasing innovation in finan-cial markets, while empirical work confirms thatforeign bank entry has helped maintain competi-tion during a process of banking consolidation inmany developing countries (Gelos and Roldos2004). An evaluation of data comparing the sim-ple efficiency measures for foreign and domesticbanks shows decidedly mixed results (table 3.4).In developing countries as a group, foreign banksaverage significantly higher overheads and costs,but lower loan loss reserves, than domestic banks.These results vary substantially by region, how-ever, with Europe and Central Asia recordingparticularly efficient indicators for foreign banks.In Latin America and the Caribbean, foreignbanks have had smaller net interest margins than
92
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 92
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
domestic banks but no difference in costs, whereasin Sub-Saharan Africa, foreign banks performedbetter compared with domestic banks but only sig-nificantly so in loan loss ratios.
These diverse results reflect the wide range ofboth foreign banks and domestic banking condi-tions in developing countries. Characteristics offoreign banks that might affect their efficiencyinclude the efficiency and origin of the parentbank, the type of operation (such as wholesale ver-sus retail), the motive (following the client versusmarket-seeking), the market share of the foreignbanks, and the mode of entry (Berger and others2008; Sturm and Williams 2005). Factors relatedto the host economy, such as initial financial, eco-nomic, and regulatory conditions, may also affectthe efficiency of foreign banks. One factor affectingthe relationship between efficiency and mode ofentry is the advantage that a greenfield entry offersin allowing investors greater scope and choicein setting up a new facility, compared with an
93
Table 3.4 Average foreign and domestic bank performance indicators in developing regions, 1998–2005
Loan loss Loan loss Pretax Net interest Overhead to Taxes to reserves to reserves to profits to Cost to
Category margin (%) assets ratio (%) assets ratio assets ratio gross loans assets ratio income ratio
Developing countriesDomestic 7.27 5.72 0.53 4.51 8.32 1.69 69.60Foreign 6.86 6.30 0.63 3.63 7.27 1.29 76.52
East Asia and PacificDomestic 3.84 2.68 0.35 3.26 6.01 0.66 63.98Foreign 3.83 3.03 0.57 10.35 11.85 2.04 62.10
Europe and Central AsiaDomestic 7.71 6.55 0.67 5.24 8.13 2.08 67.86Foreign 6.02 5.59 0.41 2.92 5.70 1.43 73.73
Latin America and the CaribbeanDomestic 9.79 7.55 0.44 3.06 7.23 1.84 76.74Foreign 7.83 8.05 0.83 2.74 7.52 0.63 81.30
Middle East and North AfricaDomestic 3.57 2.16 0.25 5.84 12.66 1.08 59.78Foreign 3.71 2.69 0.27 8.25 16.07 0.90 76.09
South AsiaDomestic 2.85 2.52 0.44 2.47 6.35 0.92 64.75Foreign 3.75 2.38 1.02 1.62 7.06 2.46 51.07
Sub-Saharan AfricaDomestic 10.08 7.76 0.79 8.52 12.56 2.55 74.08Foreign 9.07 7.24 0.81 3.31 5.54 1.89 81.40
Developed countriesDomestic 2.63 2.20 0.27 1.92 3.19 1.01 59.78Foreign 1.80 1.74 0.23 1.40 2.69 1.26 55.86
Source: World Bank staff estimates based on data from Bankscope.Note: Pairs in bold indicate difference in means of corresponding indicators for foreign and domestic banks and are statistically significant at the 10 percent level. Net interest margin is net interest income as a percentage of earning assets.
M&A transaction, which is typically burdened byoverhang costs and organizational structure inthe existing business. Entry through M&A mayinvolve higher organizational and operationalcosts, which may delay the improvement in effi-ciency of the foreign banks, although an immedi-ate increase in the market share after acquisitionmay increase efficiency through economies ofscale. The efficiency advantage of the new invest-ment mode of entry is borne out by the experienceof foreign banks entering Europe and Central Asia(as it is in developed countries as a whole), thoughnot by the experience of Sub-Saharan Africa,where foreign banks entering through M&A havesuperior efficiency to those entering throughgreenfield investment (figure 3.9). In other regionsthe difference in efficiency associated with new in-vestment and M&A mode of foreign entry is notsufficiently pronounced to project a clear point ofview, in part because of a lower number of M&Atransactions in South Asia.
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 93
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
Foreign bank presence has helped easedomestic credit constraints onmanufacturing firms Access to international banking, whether cross-border or through foreign banks’ local investments,increases the potential sources of credit availableto developing-country firms. If markets are per-fectly competitive and if all lenders have access tofull information, foreign banks’ increased accessto technology, improved opportunities for riskdiversification, and perhaps better corporate gover-nance should enable them to offer lower interestrates and a higher volume of credit. However, bar-riers to information and limits on competition toprotect safety and soundness are pervasive in finan-cial markets, greatly complicating an analysis ofthe impact of foreign banks.
Most empirical studies conclude that the pres-ence of foreign banks increases access to credit. Forexample, Giannetti and Ongena (2005), in a cross-country study using firm-level data, find that for-eign lending increased growth in firm sales, assets,and leverage in Eastern European countries. (Theeffect was dampened, although still positive, forsmall firms.) A survey of firms operating in 35 de-veloping countries suggests that all firms, includingsmall and medium-size firms, report lower obsta-cles to obtaining finance in countries with higherlevels of bank presence (Clarke, Cull, and Matinez
Peria 2006). Beck, Demirgüç-Kunt, and Maksi-movic (2004) conclude that greater foreign bankpresence tends to alleviate the impact of bank con-centration on setting obstacles to credit access.Even if foreign banking tends to improve accessto credit on average, the impact may vary signifi-cantly among countries or firms. Some studieshave found that foreign banks tend to “cherrypick” the best borrowers, thus limiting credit ex-pansion (Mian 2004; Detragiache, Gupta andTressel 2006). Therefore, given the existing mixedempirical evidence, focusing on the informationalrequirements of banking and on the efficiency andreal benefits of foreign bank presence can thus pro-vide insight into the potentially differentiated im-pact and also help determine whether foreignbanks might help to mitigate connected-lendingproblems and improve capital allocation.
Econometric analysis (detailed in annex 3Aof this chapter) shows that foreign banks are par-ticularly important for industries in developingcountries that rely heavily on external financing.For instance, in a country in which the bankingsector is 20 percent foreign owned, such as Brazil,the difference in growth between companies withlow financial dependence (at the 25th percentileof all companies) and those with high financialdependence (at the 75th percentile) is less than1 percentage point on average (figure 3.10). Thedifference increases exponentially when foreignbank presence is stronger. In countries where for-eign ownership of the banking sector is 40–60 per-cent, such as Bolivia and Romania, companies
94
0 0.20 0.40
Foreign bank assets share
0.60
2.4%
1.6%
0.8%
0.800
1
3
2
4
5
Figure 3.10 Real effects of foreign bank presence
Difference in growth rate, %
Low financial dependence
High financial dependence
Sources: World Bank staff estimates based on Bankscope andWorld Bank data.
Green
field
Mer
gers
and
acqu
isitio
ns
Mer
gers
and
acqu
isitio
ns
Mer
gers
and
acqu
isitio
ns
Green
field
Green
field
0
1
3
2
4
5
6
7
8
9
10
Figure 3.9 Ratio of overhead cost to total assetsin select regions, by mode of foreign bank entry,1998–2005
Percent
Developedcountries
Europe andCentral Asia
Sub-SaharanAfrica
Source: World Bank staff estimates based on data from Bankscope.
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 94
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
with high financial dependence grow 1.6 and 2.4percentage points more, respectively, than thosewith low financial dependence. As a whole, theseresults show not only the importance of foreignbank presence for industry growth in developingcountries but also the crucial role of such banks inparticular industries, namely, those most in needof external financing.
Transmission of financial shocksthrough the internationalbanking system
The international banking industry’s adjustmentto the current global financial turmoil bears
importantly on the prospects of foreign creditsupply to developing countries. A large body ofliterature and empirical evidence indicates thatbanks tend to react to adverse financial condi-tions through balance-sheet adjustments in orderto meet a variety of risk management standards(such as value at risk), performance indicators(return on equity), and regulatory requirements(Basel I or II). The response of Japanese banks to thestock and real estate market collapse of early 1990s,when they pulled back from foreign markets—including the United States—in order to reduce lia-bilities on their balance sheets and thereby meetcapital adequacy ratio requirements, is indicativeof how banks can transmit domestic financialshocks to foreign markets.
Three trends are important in the transmis-sion of financial shocks to developing countries:first, mounting pressure on major banks’ capitalpositions as they recognize balance-sheet losses;second, deteriorating liquidity conditions in inter-bank markets; and third, tightening credit stan-dards in the face of global economic slowdown.The fact that all three transmission channels arecurrently operating simultaneously raises the pos-sibility of a sharp global credit downturn, withparticularly negative implications for developingcountries whose corporate sectors depend onbanks as their primary source of external financing.As of March 2008, credit write-downs and lossesdisclosed by major banking institutions exposedto U.S. subprime-related securities amounted to$206 billion, with roughly one-half attributableto European banks ($98.5 billion) and the rest at-tributable to U.S. banks ($92.3 billion) and others
($15.2 billion). Because it seems too early to evalu-ate the implications of bank-specific balance-sheetproblems on the overall banking sector’s willing-ness to lend to developing countries, the followinganalysis focuses on developments in global inter-bank markets and the downturn in the lendingcycle. A useful start would be to highlight some ofthe key characteristics of the top 200 internationallenders to developing countries (box 3.2).
In the current grouping of the top 200 lendersto developing countries, 18 have experienced con-siderable credit deterioration and asset price lossesfrom exposure to subprime-related securities andstructured investment vehicles. Those not directlyaffected by the subprime turmoil have sufferedfrom tightening liquidity conditions in global in-terbank markets and an associated rise in fundingcosts.
Tightening of global liquidity has heightenedshort-term funding pressuresAlthough bank borrowing in the interbank andcommercial-paper markets has increased steadilysince the early 1990s, short-term funding of lend-ing activities skyrocketed after 2002, as liquidityin global financial markets increased because ofeasy monetary policy responses to the global slow-down in 2001. As a result, global banks have in-creasingly relied on short-term financing sourcesnot only for managing liquidity but also for fund-ing their balance-sheet expansion. In essence,banks have engaged in maturity transformation onan unprecedented scale, taking advantage of rela-tively steep yield curves by borrowing short andlending long.
In recent months, however, this strategy hasexposed banks to interest-rate risk from maturitymismatch (flattening of the yield curve) and liquid-ity risk (the inability to roll over interbank debt).Though the former risk is related to monetary andmacroeconomic conditions, the latter arises fromcounterparty risk (informational asymmetriesamong market participants). When perceivedcounterparty risk increases, as it has during thecurrent financial market turmoil, banks becomemore reluctant to lend to each other. And sincemost interbank lending occurs among a clearly de-fined group of global institutions and leads to in-terrelated claims by the same group of institutions,denial of credit to some market participants islikely to be followed by a chain of denied credit
95
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 95
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
96
The universe of international banks with exposure todeveloping-country-based borrowers (a population of
approximately 2,500) spans a large number of institutionsof diverse size, country of origin, funding structure, balance-sheet health, and access to global interbank markets. Thetop 200 lenders include global banking giants such asABN AMRO, Citigroup, Goldman Sachs, HSBC, MorganStanley, and Standard Chartered, which typically haveexposure in multiple countries and provide a wide range ofunderwriting and investment banking services in additionto bank lending, as well as a multitude of smaller bankswith more limited and focused exposure. By asset size,the top 200 lenders range from $970 million (CIMBInvestment Bank based in Malaysia) to $2 trillion (UBS),as of end-2006.
The market share of the top 200 lenders is substantial:together, they account for about 80 percent of cross-borderlending to developing countries. The top 50 lenders accountfor 50 percent (figure below).
Top lenders to developing countries entered therecent financial turmoil with strong profitability andsound capital positions (figures below), reflecting thestrong performance of the banking industry during
the boom years of 2002–06. Banks’ ability to retain thesepercentages in coming months will reflect the severityof the credit squeeze.
Box 3.2 Profile of the top 200 lenders to developing countries
1 100 200 300
No. of loan providers
500 75050
100
90
0
70
60
50
40
30
20
10
80
Cumulative international bank lending to developingcountries
% of lending
Source: World Bank estimates based on data from Dealogic Loan Analytics.
Top 200 lenders to developing countries
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 20070.2
0.4
0.6
0.8
1.0
1.2
1.4
Percent
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 200710.0
11.5
11.0
10.5
12.0
12.5
13.0
13.5
14.0
Percent
Average BIS capital ratioAverage return on assets
Source: World Bank staff calculations based on data from The Banker.
Note: BIS � Bank for International Settlements.
requests, thereby restricting the availability ofliquidity. Several episodes since 1990 illustratethe mechanics of such liquidity strain in globalinterbank markets (box 3.3).
In the context of the current credit marketturmoil, growing uncertainty about counter-party quality resulted in a significant tighteningof liquidity conditions and a widening of spreads
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 96
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
between three-month LIBOR and three-monthovernight index swap rates (the LIBOR/OISspread) from an average of 8 basis points in thefirst half of 2007 to 95 basis points in mid-September 2007.10 In the ensuing months theLIBOR/OIS spread remained at a level more thansix times as high as its long-term average betweenJanuary 1990 and June 2007, even after centralbanks injected massive amounts of liquidity intointerbank markets (figure 3.11). The persistenceof high LIBOR/OIS spreads suggests that factorsbeyond liquidity, such as counterparty exposureand informational asymmetries regarding marketparticipants’ credit quality, are affecting inter-bank markets.
To further investigate the link between globalmoney-market conditions and international banks’lending to developing countries, we analyze how theavailability of interbank liquidity, as measured bythe LIBOR/OIS spread, affects the supply of creditto developing countries in a multivariate panel re-gression framework controlling for macroeco-nomic, institutional, and regional effects (see annex3B for the underlying methodology and estimation).In general, the results reveal that deterioration in
interbank liquidity adversely affects lending to de-veloping countries. As highly leveraged institutions,banks need to roll over a large proportion of their li-abilities on a very short-term basis, and thus even asmall rise in their cost of funding could translate
97
Historically, the international banking industry hasexperienced periodic episodes of tight liquidity, as
reflected by the peaks in spreads between LIBOR and U.S.Treasury bill or other central bank policy rates (figurebelow). In 1991–92, for example, several large U.S. bankssuffered significant deterioration in the quality of their loanportfolios, causing spreads to peak. Interbank spreads jumpedagain during the Asian and Russian financial crises in1997 and 1998, when the global banking system had accu-mulated large exposures to affected countries. Also reveal-ing is the collapse of Long-Term Capital Management inlate 1998, when 15 of the largest players in the interbankmarket had considerable exposure to the hedge fund. Inthat instance, during which the institutions’ identity andextent of exposure were not known at the outset of the cri-sis, the market reaction was systemic, leading to generalizedwithdrawal of liquidity and a surge in interbank rates.Spreads over U.S. Treasuries jumped to 166 basis points.In August 2007, at the start of the current crisis, spreadsover Treasuries shot up to 242 basis points and haveremained elevated in the months following, despitemassive liquidity injections by major central banks.
Box 3.3 Global funding pressure, 1990–2008
Jan.
199
0
Jan.
199
2
Jan.
199
4
Jan.
199
6
Jan.
199
8
Jan.
200
0
Jan.
200
2
Jan.
200
4
Jan.
200
6
May
200
80
6
3
4
5
1
2
300
250
0
50
100
150
200
Spreads between LIBOR and U.S. Treasury bill rates,January 1990–May 2008
Basis points Percent
Sources: World Bank staff estimates based on data from Datastreamand IMF International Financial Statistics.
Note: TED = Treasury-Eurodollar. The TED spread is the difference betweenthe three-month U.S. Treasury Bill interest rate and the three-month LIBOR.
Inflation, 24-monthmoving average
(right axis)
TED spread
Figure 3.11 Term liquidity spreads: three-monthLIBOR/three-month OIS
Basis points
�20
20
60
80
100
120
40
0
Jan.
200
7
Feb. 2
007
Mar
. 200
7
Apr. 2
007
May
200
7
Jun.
200
7
Jul. 2
007
Aug. 2
007
Sep. 2
007
Oct. 2
007
Nov. 2
007
Dec. 2
007
Jan.
200
8
Feb. 2
008
Mar
. 200
8
May
200
8
Apr. 2
008
Sources: World Bank staff calculations based on data fromBloomberg and Datastream.
US$
Euro
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 97
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
into a relatively large scaling back of lending. Notsurprisingly, our empirical investigations show thatan increase in the LIBOR/OIS spread by 10 basispoints can be expected to lead to a net decrease inlending to developing countries by up to 3 percent.The estimations also show that uncertainty sur-rounding the availability of interbank liquidityhurts emerging-market lending. Thus, a 10 percent-age point increase in the volatility of the LIBOR/OISspread decreases credit to developing countries by1 percent.
The credit cycle channel: tightening of creditstandardsIn general, credit supply moves procyclically over thebusiness cycle. The underlying economic mechanismis straightforward: different phases of the businesscycle provide different incentives for informationcollection (borrower screening), thus leading to vary-ing degrees of competition among lenders and, ulti-mately, to different credit standards.11 Given a poolof borrowers, average repayment probability variesnegatively with the business cycle. Since a larger frac-tion of a borrower pool has access to credit in boomtimes (when lending standards tend to be more lax),loans originating at the height of the business cycleare precisely those with the most likely risk of defaultduring an economic downturn. And because thepool of creditworthy borrowers appears larger dur-ing expansions, banks tend to compete more in-tensely for borrowers’ business during those times,providing loans at lower margins and at softer termsand conditions and thereby reducing credit spreads.
While the procyclical character of bank lend-ing is evident in the Federal Reserve’s Senior LoanOfficer Opinion Survey, the survey also suggeststhat lenders anticipate the competitive dynamics ofcredit cycles. As a result, credit standards typicallyturn earlier than the business cycle. In fact, thecorrelation between the fraction of U.S. banksreporting tightening of credit standards in the Fed-eral Reserve survey and GDP growth is �0.47,highlighting the anticipatory nature of credit stan-dards that gives rise to procyclical lending cycles.As the United States recovered from a downturn inthe early 1990s, lending standards became consid-erably more lenient; since mid-2005, however, stan-dards have been rising (figure 3.12). In theEuropean Union, lending standards began tighteningin mid-2007 (figure 3.13). These observations aboutthe procyclical nature of lending in developed
markets hold important lessons for the availabilityof credit in emerging markets.
Our multivariate regression results, in whichwe relate the (logarithm of) foreign bank claims onemerging economies to the fraction of U.S. banksreporting tighter credit standards in a given quar-ter, its lags, and macroeconomic and institutional
98
Q2 19
90
Q2 19
92
Q2 19
94
Q2 19
96
Q2 19
98
Q2 20
00
Q2 20
02
Q2 20
04
Q2 20
06
Q1 20
08�4
10
6
8
2
4
0
�2
80
60
�40
�20
0
20
40
Figure 3.12 Reported tightening in U.S. lendingstandards for commercial and industrial loans,1990–2008
Percent GDP growth, %
Real GDP growth rate(right axis)
Net percentage of senior loan officerstightening lending standards (left axis)
Sources: World Bank staff calculations based on data fromU.S. Federal Reserve Board (2008) and U.S. Bureau of EconomicAnalysis.
Note: The net percentage of tightening is the percentage of seniorloan officers who reported tightening minus the percentage ofofficers who reported easing in credit standards.
Figure 3.13 Reported tightening in EU lendingstandards, by size of enterprise, 2003–07
Percent
�40
0
40
60
80
20
�20
Q1 2003 Q1 2004
Loans to largeenterprises
Q1 2005 Q1 2006 Q4 2007Q1 2007
Source: ECB 2008.
Note: The net percentage of tightening is the percentage of banksthat reported tightening minus the percentage of banks thatreported easing in credit standards.
Loans to small andmedium-size enterprises
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 98
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
control variables, confirm the predictions in theliterature.12 Based on estimates reported in table3B.2, it can be inferred that a 10 percentage pointincrease in banks’ credit standards decreases lend-ing to emerging economies by up to 0.7 percent.The results are even more pronounced in the first-difference specifications, where emerging-marketlending decreases up to 1.2 percent in the follow-ing quarter for a similar change in credit standardsin the current period.
However, interbank funding pressures andtightening credit standards do not affect develop-ing countries in a uniform manner: country sizeand regional factors seem to matter for their accessto foreign credit. Econometric investigation of theinteraction between country size (as measured byGDP) and regional factors with interbank liquidityand lending standards suggests that, because ofthe frequency and volume of their borrowingneeds, larger countries are more severely affectedthan smaller countries by the tightening of liquidityconditions. By contrast, because large countriestypically offer better economic and financialprospects and are perceived as less risky thansmaller countries, they are not differentially affectedby tightening of credit standards during economicdownturns. By region, it appears that tighteningliquidity conditions tends to affect lending toEurope and Central Asia and Latin America muchmore than elsewhere. Also, because foreign banksdominate lending to borrowers in Europe andCentral Asia, the region seems particularly vulner-able to the procyclical behavior of bank lendingduring periods of global economic downturns.
Taken as a whole, our analysis shows that twooverriding factors shaping the current global lend-ing environment—tight interbank liquidity and ris-ing credit standards—are likely to have tangiblenegative effects for the availability of credit to de-veloping countries. Although successive coordi-nated measures by major central banks, includingthe U.S. Federal Reserve, ECB, Bank of Canada,Bank of England, and Swiss National Bank (SNB),to expand their provision of liquidity through aterm auction facility in the United States and cur-rency swap arrangements managed by the ECB andthe SNB and to provide liquidity in exchange for awidened set of collateral, have helped stabilizemarket conditions to some extent, persistently highinterbank spreads seem to point to high counter-party credit risk and an overall transition in the
international banking system away from high levelsof credit securitization and leverage. The practicalimpact of such developments is already visible inthe market for syndicated bank lending to develop-ing countries, with both the volume of deals signedand total deal value recording a sizable drop in thefourth quarter of 2007 and the first quarter of2008 compared with the same periods in the previ-ous year.13 Also indicative of tighter financing con-ditions are higher spreads asked for some borrow-ers—for example, Sberbank, Russia’s largest bank,paid a margin of 45 basis points on its latest loanin December 2007, 15 basis points more than it didin 2006—and the fact that some deals are failing toattract the necessary traction among investors.Indeed, for a country such as Kazakhstan, where96 percent of total foreign claims on the countryare denominated in foreign currency (and in which65 percent of these claims are on the banking sec-tor), heightened pressures in global interbank mar-kets could translate into severe funding constraintson the country’s banking sector.
In contrast to the current financial market tur-moil, which originated squarely in developed mar-kets and is spreading to developing countries, thecase of Argentina in the early 2000s illustrates thereaction of foreign banks to turmoil that began ina developing country, where they had a significantpresence. On the eve of the crisis, foreign banksaccounted for almost 50 percent of Argentina’sbanking assets, as foreign bank entry had acceler-ated in the second half of the 1990s supported bythe progress in the privatization program. Follow-ing the crisis in 2001, the reaction of foreign banksto the crisis varied significantly. Some banks main-tained their assets, whereas others opted to exit.As a result, there was a sizable decline in foreignbank presence and asset ownership in Argentina.Several of these foreign banks also had a majorpresence in other countries in the region. Whilesome banks reoriented their regional activities,there was limited spillover to other countries in theregion, as detailed in box 3.4.
Macroeconomic consequencesof international banking
Growing foreign bank presence has importantmacroeconomic management and financial
stability implications for developing countries.
99
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 99
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
Two aspects in particular warrant attention at thecurrent juncture: domestic credit booms, andthe diminished ability of monetary authorities toinfluence market lending rates through changesin short-term money-market rates. Regarding theformer, private credit in a sample of 29 develop-ing countries has expanded more than 40 percentper year, on average, over 2003–06, whereas
inflation and economic growth in those same coun-tries have averaged 8.8 percent and 7.1 percentper year, respectively.14 This observation calls forexplanation and caution. Although the underlyingpattern of high domestic economic growth andfinancial deepening (the latter of which is definedas the ratio of private credit to GDP or the ratio ofbroad money supply to GDP) in these countries
100
In response to severe economic and currency distress inArgentina in 2001, the government adopted a policy of
conversion of U.S. dollar–based assets and liabilities intopesos (pesofication) and mandatory rescheduling of termdeposits. The pesofication of highly dollarized bank bal-ance sheets resulted in a disproportionate decline in thevalue of bank assets and corresponding equity losses.Subsequently, the government implemented a sequence ofmeasures in the banking sector, including restrictions ondeposit payouts, capital controls, suspensions of enforce-ment of judicial foreclosure procedures, and restorationof depositors’ rights to the full original dollar value oftheir frozen deposits (de la Torre, Levy-Yeyati, andSchmukler 2002).
The reaction of foreign banks to the crisis and thegovernment’s subsequent measures varied dramatically:some institutions maintained their assets, while others soldoff everything. Of the top five foreign banks in Argentinaat the time, which accounted for 35 percent of bankingsector assets in 2000, two Spanish banks, Banco San-tander, and BBVA, maintained their shares in the country(left-hand figure below). U.S.-based BankBoston and
Citibank and U.K.-based HSBC decreased their interestsignificantly. In all, 10 foreign banks opted to exitArgentina. In 2002 (within a year of the crisis), fourforeign banks shut down either voluntarily or after thecancellation or revocation of their banking licenses.a
In 2003, six more foreign banks left the country.b As aresult, foreign banks’ share of assets fell to 24 percentin 2004—down from 52 percent in 2000—and recoveredonly marginally to 31 percent in 2006.
As a result of the crisis, several foreign banks reori-ented their regional activities in Latin America. HSBC, forexample, entered the Mexican market in 2001. The bankleft Brazil in 2005. BBVA left Bolivia in 2002. Citibankentered Mexico with its record-size acquisition ofBanamex in 2001 (right-hand figure below). All in all,however, foreign banks maintained their share of bankingsector assets in the largest Latin American economies.
a. Banco Exterior de America (Uruguay), Chase Manhattan Bank (UnitedStates), Mercobank (Chile), and Banco do Estado de São Paulo (Brazil).b. Scotiabank Quilmes (Canada), Banco General de Negocios (Switzerland),Banco Velox (Uruguay), Banco Bisel (France), Kookmin Bank (Korea), andCredito Argentino Germánico (Germany).
Box 3.4 Foreign banks’ reaction to the Argentine crisis
Top five foreign banks’ share of Argentina’s banking sector assets
Percent Percent
0
2
4
10
8
6
Source: World Bank staff estimates based on data from Bankscope.
Santander BBVA HSBC BankBoston Citibank0
20
40
90
80
70
60
50
10
30
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Argentina
Mexico
Asset share of foreign banks in selected countries
Chile
Brazil
1999 2000 2001 2002 2003
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 100
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
has increased the scope for banks’ expansion,rapid growth in private credit would inevitablyneed to be funded by foreign sources.
Foreign banks have contributed to domesticcredit creation in developing countriesSome developing countries, especially those inEurope and Central Asia, have generally experi-enced swift private credit expansion in recent years,buttressed by strong economic growth and finan-cial deepening. For some of these countries, though,deposit growth is lagging behind credit growth.In these cases, two other factors seem to havecontributed to fast credit expansion. First, thebanking sector in some countries has borrowedextensively from foreign markets and used externalfunds to finance domestic credit creation, asevidenced in Kazakhstan, Latvia, Romania, Russia,and Ukraine, and, to a lesser extent, India. Second,the foreign bank presence in some countries is sig-nificant. Foreign banks’ strong financial footingand easy access to external funding have facilitatedcredit creation in such countries as Albania, Arme-nia, Bulgaria, the Kyrgyz Republic, and Lithuania.As shown in table 3.5, sometimes the two factorswork in tandem, that is, foreign bank presence mayincrease access to the external funding market.
By further examining the 29 developing coun-tries with the fastest private credit growth over2003–06, we find that growth of private credit andits association with foreign bank presence are gen-erally recent phenomena—between 2000 and 2006,
the average ratio of private credit to GDP in thesecountries grew from 10 percent to 25 percent (leftpanel of figure 3.14). Foreign bank assets as a per-centage of domestic banking sector assets in thesame sample of countries also increased substantiallyover the same time frame—from 36 percent in 2000to 50 percent in 2006 (right panel of figure 3.14).
Econometric analysis of a large sample ofdeveloping countries over 1995–2005 further sup-ports the contention that a positive and statisticallysignificant relationship exists between foreignbank presence and private credit growth aftercontrolling for country-specific macroeconomic,institutional, and financial sector developmentindicators, as well as for foreign borrowing bydomestic banks.15
Foreign bank presence appears to haveweakened the transmission of monetary policyMonetary policy has played an increasingly im-portant role in the macroeconomic manage-ment approach of many developing economiesin recent years. Alongside that trend, the ques-tion of how foreign bank presence affects thetransmission of monetary policy has also gainedprominence. As central banks emphasize themarket orientation of their monetary policythrough open-market operations and the liberal-ization of domestic interest rates, one key mecha-nism of monetary policy transmission is the linkbetween the bank lending rate and the short-termmoney-market rate.
101
Table 3.5 Characteristics of selected developing countries with large private credit growth
Annual private Annual deposit Share of foreign Total overseas borrowing Annual GDP Annual credit growth, % growth, % assets in banking, % by banking sector, $ millions growth, % inflation, %
Country (2003–06) (2003–06) (2003–05) (2003–06) (2003–06) (2003–06)
Venezuela, R. B. de 76.6 61.8 34.1 313 7.8 20.6Kazakhstan 69.0 61.0 25.7 24,193 9.8 7.4Azerbaijan 63.4 52.3 2.9 208 20.6 6.1Latvia 56.1 39.6 48.1 2,011 9.6 5.6Albania 55.4 16.5 76.9 — 5.5 1.9Ukraine 54.4 44.3 27.0 4,620 7.8 9.2Belarus 53.3 43.1 16.0 203 9.5 16.0Romania 49.1 34.6 55.1 2,522 6.4 10.7Lithuania 47.3 31.2 91.7 126 8.2 1.6Kyrgyz Republic 44.1 33.9 79.2 — 4.1 4.2Russian Federation 43.8 37.9 12.1 51,203 6.9 11.7Armenia 36.3 24.6 44.5 — 13.0 3.8Bulgaria 35.0 32.0 72.7 1,179 6.0 5.2Argentina 28.2 19.3 29.0 1,340 8.9 8.5India 28.1 18.5 5.0 12,472 8.8 4.4
Sources: World Bank staff estimates based on data from IMF International Financial Statistics, Bankscope, Dealogic DCM Analytics, andWorld Development Indicators (various years).Note: The mean of annual private credit growth over 2003–06 for all developing countries is 25.6; the median is 22.3; and the standarddeviation is 18.1; — � not available.
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 101
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
The debate on the role of foreign banks in thetransmission of monetary policy in developingcountries centers around two opposing views:first, that higher foreign bank presence strengthenstransmission because it enhances financial sectorefficiency and depth; and second, that foreignbanks are less responsive to domestic monetarypolicy impulses because they have access to a largepool of external funds beyond the control of themonetary authority.
In both cases, the structure of the financialsystem is of utmost importance in the functioningof the monetary transmission mechanism. Specifi-cally, the effectiveness of market-oriented policyinstruments depends critically on the sophistica-tion of and competition in the financial sector. Forthe asset price channel to be operative, changes inthe money-market rate—the interest rate typicallytargeted by central banks—must be passed on tothe asset prices relevant to households’ and firms’decisions about how much to consume, invest, andproduce. In an underdeveloped financial system,however, financial markets other than the moneymarket may not exist and money-market ratesmay be decoupled from the relevant asset prices,undermining the effectiveness of open-market op-erations. Greater competition in the banking sec-tor induces a tighter pass-through between policyinterest rates and lending rates, thus enhancing theefficacy of monetary policy. Noncompetitive pric-ing, on the other hand, potentially including asym-metric responses to increases or decreases in thecost of reserves, creates a gap between money-market rates and lending rates, thus impairing the
ability of the central bank to influence the relativeprices.
Figure 3.15 shows the evolution of averagemoney-market and lending rates for a sample of22 developing countries. Figure 3.16 shows theevolution of interest rates for individual coun-tries, several of which have experienced bankingcrises during the period examined. The fairly con-sistent decline in both rates over the past decadeevident in both aggregate and country experienceis noteworthy, reflecting in part the success thesecountries have achieved in lowering inflation, aswell as in deepening their financial systems. Never-theless, there is still a fairly high pass-through frommoney-market rates to lending rates: on average,the correlation coefficient is 0.84.
102
Figure 3.15 Average money-market and lendingrates in 22 developing countries, 1995–2007
Percent
1995 1997 1999 2003 2005 2007
Sources: World Bank staff estimates based on Bankscope and IMFInternational Financial Statistics.
20010
10
50
70
90
40
30
20
60
80
Average lending rate
Average money-market rate
Figure 3.14 Private credit growth and distribution of foreign bank assets in developing countries
Ratio of private credit to GDP
0
0.20
0.30
Sample countries
Sample countriesAll other countries
All other countries
0.40
0.35
0.15
0.10
0.05
2000 2001 2002 2003 2004 2005 2006 2000 2001 2002 2003 2004 2005 2006
0.25
Sources: World Bank staff estimates based on data from Bankscope and IMF International Financial Statistics.
0
10
40
50
60
30
20
Foreign bank assets ratio, %
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 102
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
103
Annex 3C presents an econometric analysis ofthe pass-through from money-market rates tolending rates in developing countries. The resultssuggest that economies with deeper financial sys-tems are associated with lower lending rates and ahigher pass-through from money-market rates tolending rates. The results also show that although
higher foreign bank presence does not seem to affectlending rates, it reduces the pass-through frommoney-market rates to lending rates. This result isconsistent with the view that foreign banks are lesssensitive than domestic banks to domestic mone-tary conditions because of their ability to accessinternational capital markets.
Figure 3.16 Average money-market and lending rates for a sample of countries
Percent
0
10
50
100
60
70
80
90
20
30
40
Sources: World Bank staff estimates based on data from Bankscope and IMF International Financial Statistics.
1995 1997 1999 2001 2003
Lending rateLending rate
Lending rate
Lending rate
Lending rate
Lending rate
Money-market rate
Money-market rate
Money-market rateMoney-market rate
Money-market rate
Money-market rate
2005 20070
120
60
80
100
20
40
1997 1999 2001 2003 2005 2007
Percent
Argentina Brazil
Percent
0
10
35
15
20
25
30
5
10
35
15
20
25
30
5
1995 1997 1999 2001 2003 2005 20070
40
1995 19991997 20032001 2005 2007
Percent
Hungary Poland
Percent
0
50
250
500
300
350
400
450
100
150
200
1995 1997 1999 20032001 20041996 1998 200220000
25
5
10
15
20
1995 1997 1999 2001 2003 2005 2007
Percent
Russian Federation Thailand
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 103
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
Country experiences with the monetarytransmission mechanismSuccessful implementation of monetary policy inany country requires a solid framework that con-ditions the monetary transmission mechanism.Under an inflation-targeting regime, the centralbank typically has direct influence on overnightinterbank lending rates and thus indirectly influ-ences interest rates across the entire term structure.In the case of Brazil, which adopted an inflation-targeting regime in June 1999 following a currencycrisis, and in which there is a moderate degree offoreign bank presence, the pass-through frommoney-market rates to longer-term lending rateshas been strong, with an estimated correlationcoefficient of 0.90 over 1999–2007. In the SlovakRepublic, which adopted an inflation-targetingregime in January 2005 following accession to theEuropean Union in 2004, and in which there isvery high foreign bank presence, the correlationcoefficient of money-market and lending ratesover the same period is lower, at 0.82, suggesting aweaker pass-through than in Brazil.
In the Slovak Republic, the government had un-dertaken widespread banking sector privatizationand restructuring starting in 1998. The reformsallowed foreign institutions to behave more compet-itively and within a few years, they dominated thebanking sector. Between 2000 and 2005, the share ofbanking sector assets held by foreign banks soaredfrom 26 percent to 91 percent. (Of those foreign-held assets, the vast majority are currently held by
just a few banks.) Concurrently, the percentageof foreign-owned banks in the total number of banksincreased dramatically, from approximately 48 per-cent in 2000 to 94 percent in 2005.
There has also been a consistent increase inbanking sector assets held by foreign banks in Brazil,from less than 5 percent in 1995 to more than 25 per-cent in 2005. Over the same years, the percentage offoreign-owned banks in the total number of banksincreased from roughly 22 percent to 35 percent.These trends reflect the fact that a large number ofsmall foreign banks was already present in Brazil in1995 and that in the following decade a small num-ber of very large foreign banks entered the country.Indeed, of the current 12 largest private banks, 5 arebased in Europe and 2 are based in the United States.
To more rigorously test the hypotheses that anincrease in foreign bank presence reduces the pass-through of money-market rates to lending rates andthat an increase in financial depth, as measured bythe ratios of domestic credit to GDP and broadmoney (M2) to GDP, increases the pass-through,we constructed a measure of the pass-through frommoney-market rates to lending rates in Brazil andthe Slovak Republic based on the regression resultsreported in annex 3C (figure 3.17). Specifically, thepass-through is defined as the sensitivity of the av-erage lending rate to a unit change in the money-market rate. For Brazil, the solid line in the figureshows the estimated long-run pass-through, whilethe dashed line shows the pass-through if foreignbank presence had remained constant at the 1995
104
0.95
1.05
1.00
1.10
1.15
1997 1998 1999 2000 2001 2002 2003 2004 2005
Slovak RepublicBrazil
0.90
0.95
1.00
1.05
2000 2001 2002 2003 2004 2005
Counterfactual pass-through
Counterfactual pass-through
Pass-through
Pass-through
Pass-through coefficient
Figure 3.17 Evolution of the pass-through of money-market rates to lending rates
Pass-through coefficient
Source: World Bank staff estimates based on data from Bankscope and IMF International Financial Statistics.
Note: The pass-through coefficient measures the long-run elasticity of lending rates with respect to changes in money-market rates. A value higherthan 1 means that a 1 percent increase in the money-market rate leads to an increase of more than 1 percent in lending rates in the long run.
gdf_ch03_081-120.qxd 5/26/08 3:17 PM Page 104
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
level and all other variables were allowed to taketheir observed values. For the Slovak Republic, thedashed line shows the pass-through if foreign bankpresence had remained at the 2000 level.
A number of results follow from the analysis ofBrazil. First, the estimated pass-through coefficientis higher than 1.00, meaning that each percentagepoint increase in money-market rates translates toan increase in lending rates of more than 1 percent.Second, the pass-through decreases as foreign bankpresence increases. However, notice that the level ofthe pass-through is approaching 1.00, which is con-sistent with the view that foreign banks increasecompetition in developing countries. Indeed, in aperfectly competitive financial market, the pass-through should be 1.00. Finally, even thoughM2/GDP in Brazil increased in the observed period,the counterfactual pass-through is roughly constantbecause of the very small coefficient that the ratioof M2/GDP has in the pass-through regressionequation. It should be stressed, however, thatM2/GDP helps explain the reduction in the gap be-tween lending rates and money-market rates.
In the Slovak Republic, the large increase in for-eign ownership of the banking sector in 2000–02 isreflected in a significant decrease of the pass-through coefficient, which dropped from 1.02 to0.93. The slight recovery of the pass-through coeffi-cient starting in 2003, however, mirrors the smalldecline in foreign ownership over the same years.Overall, the figure suggests that monetary policycould have become less effective as foreign presenceincreased in the Slovak Republic’s banking sector.
Policy lessons and agenda
The broad contour of public policy challengescurrently facing developing countries can
generally be divided into two categories: urgentmeasures geared toward enhancing resilienceand minimizing adverse consequences in the face ofongoing global turmoil; and longer-term actions andinitiatives intended to maximize the potential of theincreasing globalization of the international bankingindustry. Given the considerable diversity acrossdeveloping countries regarding the vulnerability oftheir banking sectors to global shocks (or, morebroadly, vulnerability of their economies to a down-turn in global growth), as well as the range of policyoptions available for capitalizing on banking indus-try globalization, a tailor-made approach is needed.
Policy makers should strengthen their capacity todetect risks and calibrate their policy responsesThe nexus of global slowdown and financial tur-moil is most daunting for two groups of countries:those with large external imbalances financedlargely through financial intermediaries that them-selves depend on international markets for fund-ing; and those in which foreign banks dominatethe domestic banking sector. At the same time, alldeveloping countries, however, are being affectedby heightened risk aversion and financial anxiety.As such, the cost of default protection on emerging-market sovereign debt, a key indicator of investorrisk aversion and sentiment, has increased for vir-tually all developing countries active in interna-tional capital markets. As shown in figure 3.18,emerging-market sovereign five-year credit defaultswaps in a sample of 20 countries traded at an av-erage of 73 basis points in June 2007, with a rela-tively low dispersion among countries. By March2008, spreads had escalated to an average of 267basis points, and dispersion among countries hadwidened significantly.
It is crucial that policy makers in emerging-market countries renew their commitment to thesound policies of the recent past and recognizethe implications of changes in the financial cli-mate. Sustaining and extending the structuralchanges and institution-building efforts that havemade emerging markets’ continued integrationinto global capital markets possible should com-mand high priority, as should strengthening regu-lation and supervision aimed at limiting currencyand maturity mismatches. Although past effortstoward macroeconomic stabilization and externaldebt management have contributed to the relativeresilience of emerging economies during the re-cent financial turmoil, these countries still needto intensify efforts to monitor foreign borrowingby their banks and risk management strategiespursued by their corporations with access to ex-ternal debt markets. Policy makers in developingcountries need also to come to terms with thelikelihood of a higher cost of credit in interna-tional markets in the medium term as global mar-kets find a post-subprime-crisis equilibrium. Thefact that LIBOR rates in all currencies and matu-rities have spiked on several occasions sinceAugust 2007 indicates that heightened fundingpressure is not likely to unwind soon unless theunderlying structural factors—high counterparty
105
gdf_ch03_081-120.qxd 5/26/08 7:30 PM Page 105
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
risk, banks’ reluctance to lend to each other, anduncertainty about valuation of structured financeproducts—are addressed.
The fact that foreign banks involved in devel-oping countries tend to have a significant regionalfocus, multiple-country exposure, and dominantmarket share in several countries highlights theneed for a customized policy response. So too doesthe fact that developing countries have diversedegrees of international versus local claims andthat they hold varying shares of their foreign debt inshort-term maturity (figure 3.19). When foreignbanks lend to multiple countries, they can serveas a source of financial contagion in those coun-tries through common-lender effects. Ten majorinternational banks, including Citibank, Com-merzbank, ING, Natixis, and Société Générale,have lending exposure to at least 50 developingcountries, and 47 banks have exposure to at least30 developing countries (figure 3.20). In severaldeveloping countries, just one or two foreignbanks have a dominant position in the bankingsector, posing the risk of serious macroeconomicconsequences from the failure of a single bank. InAlbania, for example, Austria’s Raiffiesen Bankholds nearly half of banking sector assets; inMexico, almost 50 percent of banking sector assetsare held by two foreign banks (table 3.6).
Global approach to cross-border bankingregulation, transparency, and soundnessis called for With its capacity for straddling multiple jurisdic-tions and its role as the primary conduit for fundtransfer across national borders, the internationalbanking industry inspires policy debate not onlywithin the international financial community butoccasionally also within the international politicalarena. In many respects, international banking in-stitutions are the most powerful private transna-tional actors on the global financial stage, linkingeconomies through their lending, deposit-taking,and foreign exchange operations. However, thereality that the international banking industry stillfalls well short of a fully integrated system and thatbilateral investment treaties constitute the dominantinternational legal mechanism for the promotionand governance of FDI in the banking sector meansthat foreign bank operations in developing coun-tries will continue to be the focus of intense publicpolicy attention regarding matters such as com-petition, monetary policy autonomy, credit to thecorporate sector, asset bubbles, capital flight, andcompliance with anti-money-laundering standards.
Credit market turmoil in developed marketsin recent months has exposed weaknesses in theprevailing regulatory framework and in market
106
Source: Bloomberg.
a. Export-Import Bank of China.b. As of July 19, 2007.
Chile
Mala
ysia
Thaila
nd
Russia
n Fe
dera
tion
Chinaa
Mex
icoPer
u
Bulgar
ia
Brazil
Panam
a
Vietna
m
South
Afri
ca
Colom
bia
Philipp
ines
Indo
nesia
Kazak
hsta
n
Turk
eyb
Pakist
an
Argen
tina
Vene
zuela
, R. B
. de
Figure 3.18 Risk premiums have increased across emerging economies, as shown by spreads onfive-year credit default swaps
Basis points
0
400
100
300
200
500
600
700Jun. 1, 2007 Mar. 17, 2008
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 106
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
incentives that have promoted a high degree ofcredit securitization, complex investment vehicles,and global competition among banks. Lack oftransparency in financial markets severely ham-pered the ability of investors to identify exposures.In the lead-up to the crisis, regulatory pressuresprompted major commercial banks to minimizebalance-sheet exposures by developing off-balance-sheet investment vehicles (such as conduits andstructured investment vehicles). Moreover, creditrating agencies greatly understated default risk inthe subprime mortgage market, which has since
prompted serious discussion of how best to improvethe quality of the rating process, while recognizingthe important role that credit rating agencies playin evaluating risk and disseminating information toinvestors and other market participants.
International policy coordination needs to be enhanced among developed countriesGiven the extent of cross-border exposures, coor-dination of financial regulation is also necessaryin the present environment, as inadequate regula-tion in one country can have major repercussions
107
0
20
40
100
60
80
Sources: Bank for International Settlements; World Bank staff calculations.
Figure 3.19 Composition of foreign claims in select developing countries as of third quarter 2007
Percent
0
40
20
60
80
100
Percent
International claims as a share of total foreign claims Local claims as a share of total foreign claims
Kazak
hsta
n
Panam
a
Russia
n Fed
erat
ion
China
Philipp
ines
Latvi
a
Turk
ey
Indo
nesia
Bulgar
ia
Ukrain
e
Roman
iaPer
u
Colom
bia
Poland
Chile
Thaila
nd
Mala
ysia
Brazil
Venez
uela,
R. B
. de
Slovak
Rep
ublic
South
Afri
ca
Mor
occo
Mex
ico
More international claims More local claims
0
20
40
60
Short-term claims as a share of international claims, %
Kazak
hsta
n
Roman
iaChil
e
Indo
nesia
China
Argen
tina
India
Brazil
Russia
n Fed
erat
ion
South
Afri
ca
Thaila
nd
Turk
ey
Mala
ysia
Croat
ia
Ukrain
e
Panam
a
Slovak
Rep
ublic
Latvi
a
Mex
ico
Hunga
ry
Poland
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 107
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
108
Citiban
k
Sociét
é Gén
érale
ING B
ank
BNP Par
ibas
Natixi
s
Wes
tLB
Comm
erzb
ankKBC
ABN AM
RO
Stand
ard
Charte
red
Barcla
ys
Credit
Lyon
nais
Raiffei
sen
Dresd
ner
Calyon
Deutsc
he B
ank
BOT-Mits
ubish
i
Credit
Agr
icole
Indo
suez
HSBC
Sumito
mo
Mits
ui
Bayer
. Lan
desb
ank
Erste
Ban
k
Royal
Bank o
f Sco
tland
SG Cor
p. & In
v. Ban
k
Credit
Suis
se
Land
esba
nk R
heinl
and
Bayer
nLB
Bayer
. Hyp
o- u
nd Ve
reins
bank
Mizu
ho C
orpo
rate
Ban
k
Wac
hovia
NordL
BW
GZ
HSH Nor
dban
k
Bank A
ustri
a Cre
ditan
stalt
Citigro
up
Mizu
ho
Banca
Naz
ionale
del
Lavo
ro
Forti
s Ban
k
Rabob
ank
DZ Ban
k
Baden
-Wue
rttem
berg
ische
Bank o
f Am
erica
Figure 3.20 International banks with cross-border lending exposure to at least 30 developing countries,1993–2007
No. of countries
0
20
10
70
60
50
40
30
Source: World Bank estimates based on data from Dealogic Loan Analytics.
Table 3.6 Developing countries with highly concentrated foreign banking assets, 2005–06
Host country banking sector
Number of assets held by the Host country banks Foreign bank Home country foreign bank (%)
Albania 13 Raiffeisen International Bank Austria 44.6Lithuania 9 SEB AB Sweden 33.7Angola 11 Banco BPI Portugal 29El Salvador 13 Bancolombia Colombia 26.4Botswana 6 Barclays Bank United Kingdom 26Mozambique 10 Banco Comercial Portugues—Millenium Portugal 25.9Swaziland 5 Standard Bank South Africa 24.6
Nedbank South Africa 17.8Uganda 16 Standard Bank South Africa 24.6Mexico 35 Banco Bilbao Vizcaya Argentaria—BBVA Spain 24.3
Citibank United States 20.8Slovak Republic 17 Erste Bank Austria 22.2Croatia 37 Unicredito Italiano Italy 21.4Zambia 9 Barclays Bank United Kingdom 21.3
Standard Chartered Bank United Kingdom 14.9Ghana 16 Standard Chartered Bank United Kingdom 21.2
Barclays Bank United Kingdom 20.4Bosnia-Herzegovina 29 Raiffeisen International Bank Austria 20.7Romania 28 Erste Bank Austria 20.1Côte d’Ivoire 13 Société Générale France 19.2Macedonia 17 National Bank of Greece Greece 18Madagascar 6 Calyon France 17.7
Bank of Africa Benin 10.4Paraguay 13 Unibanco Brazil 16.4Bulgaria 29 OTP Bank Hungary 15
Unicredito Italiano Italy 9.9Poland 49 Unicredito Italiano Italy 13.9Serbia and Montenegro 47 Raiffeisen International Bank Austria 12.9Cameroon 12 Banque Fédérale de Banques Populaires France 11.9
Source: World Bank staff estimates based on data from Bankscope.
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 108
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
109
on others. To this end, at their April 2008 meetingin Washington, G-7 finance ministers discussed aFinancial Stability Forum (2008) report that rec-ommended steps to tighten regulation and boosttransparency of the international financial system.Of particular note were calls to raise capital re-quirements for certain structured credit products;improve oversight of banks’ risk managementpractices (including for off-balance-sheet expo-sures); toughen requirements governing financialinstitutions’ disclosure of risks and provision ofinformation on securitized products; and requirecredit rating agencies to better manage conflicts ofinterest surrounding rating structured financeproducts and to differentiate ratings of such prod-ucts from bond ratings.
Other international financial oversight bodies,including the BIS, are reconsidering the role of creditrating agencies and credit risk insurance providers.U.K. Prime Minister Gordon Brown has called onthe IMF to cooperate with the Financial StabilityForum in establishing an early warning systemfor global financial crises. At the end of March2008, the United States and the United Kingdomset up a working group to develop proposals formonitoring and regulating the banking system.Shortly thereafter, the U.S. government announceda plan for widespread reform of its financialregulation system, including provisions for theFederal Reserve to regulate investment banks. TheFederal Reserve’s extension of liquidity supportto nonbank financial institutions through two newchannels, the Term Securities Lending Facility andthe Primary Dealer Credit Facility, is also animportant step toward opening a new era in theregulation of financial markets.
Vulnerable developing countries need to focuson the quality of openness to foreign banksPreserving the great benefits of increased access tointernational banking requires safeguarding againstpotential risks. Developing countries should there-fore develop their prudential and oversight policiescarefully. A fundamental strengthening of the insti-tutions responsible for regulation and supervisionof the banking system, for example, should improvethe efficiency of all banks (although countries withstrong financial institutions and deep financial mar-kets should have relatively less concern about therisks posed by international banks). But develop-ing countries with weak institutions and limited
financial depth face a serious dilemma: while theylikely have a lot to gain from attracting foreignbanks, they are subject to adverse financial sectorand macroeconomic consequences if foreign banksimport instability. Many of these developing coun-tries also face considerable difficulty effectively reg-ulating banks, underlining the importance of focus-ing scarce resources on ensuring quality of entry. Aselaborated in standards governing anti-money-laundering efforts, a robust licensing system for for-eign banks should include ensuring that criminalsor their associates are not involved in ownership ormanagement of entering foreign banks. The WorldBank contributes to strengthening safeguardsagainst financial abuse through targeted technicalsupport to countries with weak regulatory regimes.
Often, though, developing countries can relyon the determinations of dependable foreign au-thorities concerning the soundness of foreignbanks.16 For example, host country authoritiesoften require entering banks to seek approval fromhome country supervisors. A complementary strat-egy for safeguarding against the risks of unsoundforeign bank presence is to encourage entry from avariety of jurisdictions, and placing a high premiumon parent banks’ compliance with internationalnorms and standards relating to capital adequacy,corporate governance, and transparency. Despitethe potentially high resource costs involved, coor-dination of foreign bank supervision remains animportant goal. The Basel Committee on BankingSupervision has set out a series of recommendationson the effective coordination of supervisory activi-ties by home and host country governments forinternational banks.17 These include ensuringeffective sharing of information among authori-ties, confidentiality of information, and facilita-tion of on-site bank inspections. Whereas homecountry authorities should undertake consolidatedsupervision of international banks, host countryauthorities have the right to impose restrictions ontheir activities if the foreign bank fails to meet pru-dential standards.
Access to timely and high-quality informationabout bank operations is at the heart of effective su-pervision. While foreign banks should comply withdisclosure requirements imposed by host-countryregulators, supervisors could also make greater useof existing frameworks for the cross-border sharingof information with home-country authorities (BIS2004). Developing-country regulators also need to
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 109
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
110
consider potential international financial instabil-ity, as the failure of a bank with extensive sub-sidiaries or branches in developing countries haspotential macroeconomic implications and posesconsiderable challenges to regulators. In the eventof a major bank failure, determining the level ofliquidity assistance (if any), the recapitalization ofbanks affected, and the management of liquidation
or reconstruction is complicated by the significantpresence of foreign banks in multiple developingcountries and the negotiation of burden sharingwith home-country governments.18 Given thesedifficulties, there is considerable value in workingout a multilateral framework for these arrange-ments before the next financial storm jolts themarkets.
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 110
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
111
Annex 3A: Foreign bank presence has helped ease domestic creditconstraint on firms
slower growth, as well as country and industryfixed effects. Estimating the equation over theperiod 1995–2003 for a sample of 59 developingcountries (a total of 6,527 observations) yields apositive and statistically significant estimate of thecoefficient of interaction, � � 0.11 (p-value � 0.02).Our results are robust to various alternativeeconometric specifications, inclusion of countrycharacteristics variables, and use of alternativegrowth measures.
Data sources: The analysis is based on Brunoand Hauswald (2007). Value-added data comefrom UNIDO (2005) and are measured as thevalue of census output less the value of censusinput, which covers value of materials and sup-plies for production (including cost of all fuel andpurchased electricity) and cost of industrial ser-vices received (mainly payments for contract andcommission work and repair and maintenancework). Data on foreign bank presence are fromClaessens and others (2008).
To gauge the extent to which foreign bankpresence in developing countries enhances the
access of firms to credit, we estimate a growthmodel of firms at industry level, allowing for dif-ferences in financing structure across industries.We use the index of financial dependence devel-oped by Rajan and Zingales (1998), defined as theshare of a firm’s total capital expenditure not fi-nanced with cash flows from operations, and com-puted at industry level as the median of firms inthe industry. The basic model is summarized as:
VAi,j,t � �i,j,t � � FINDEPj � FOBANKi,t
� fixed effects � �i,j,t ,
where VA is the growth rate of value added andFOBANK refers to the share of foreign bankassets to total assets. We also include the share ofindustry to account for “convergence” effects andthe tendency of larger industries to experience
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 111
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
112
Table 3B.1 Multivariate analysis of credit supply to emerging economies
Log(foreign claims) 1st difference log(foreign claims)
Dependent variable (1) Fixed effects (2) Region (3) Fixed effects (4) Region (5) Region (6) Region
Lagged log(fc) 0.73 0.965 0.84 0.975(0.000)*** (0.000)*** (0.000)*** (0.000)***
Log(GDP) 0.197 0.034 0.225 0.027 �0.003 0(0.000)*** (0.000)*** (0.000)*** (0.000)*** �0.395 �0.943
Inflation 0.017 �0.073 �0.037 0.005 �0.037 �0.024�0.895 �0.333 �0.68 �0.932 �0.631 �0.668
Growth 0.051 0.032 �0.138 �0.015 0.201 0.157�0.801 �0.844 �0.37 �0.908 �0.222 �0.194
OIS spread 0.004 0.006 0.009 0.007 0.006 0.007�0.208 (0.049)** (0.000)*** (0.000)*** (0.047)** (0.001)***
Lagged OIS �0.012 �0.006 �0.012 �0.01 �0.005 �0.009(0.000)*** (0.027)** (0.000)*** (0.000)*** (0.054)* (0.000)***
Volatility of OIS �0.002 �0.001 �0.001 �0.001 �0.001 �0.001(0.001)*** (0.008)*** (0.000)*** (0.000)*** (0.011)** (0.000)***
Lagged volatility 0 0 0.001 0.001 0 0.001�0.921 �0.686 (0.042)** (0.054)* �0.69 (0.056)*
ICRG composite �0.003 0.002 0.001�0.228 (0.014)** �0.473
Europe and Central Asia 0.029 0.058 0.02 0.056�0.214 (0.001)*** �0.396 (0.001)***
Latin America and the Caribbean �0.028 �0.013 �0.051 �0.02
�0.216 �0.451 (0.022)** �0.241Middle East and North Africa �0.048 �0.03 �0.043 �0.016
(0.089)* �0.135 �0.136 �0.409South Asia �0.022 0.002 �0.012 0.007
�0.468 �0.944 �0.695 �0.793Sub-Saharan Africa �0.06 �0.005 �0.053 �0.009
(0.004)*** �0.787 (0.012)** �0.593Constant 0.242 0.037 �0.736 �0.148 0.113 0.032
�0.59 �0.442 (0.039)** (0.051)* (0.016)** �0.628
Observations 2,112 2,112 1,622 1,622 2,109 1,621Countries 114 87R2 0.681 0.986 0.822 0.994 0.017 0.054
Source: World Bank staff.Note: ICRG � International Country Risk Guide; OIS � contemporaneous and lagged three-month policy spread. * significant at the 10%level; ** significant at the 5% level; *** significant at the 1% level.
Annex 3B: International banks’funding strategy and lendingto developing countries
(whenever appropriate); and a host of macroeco-nomic, institutional, and regional control variables.Our dependent variables are the (log of the) BISquarterly foreign bank claims on up to 124 emerg-ing economies and their first differences, that is,growth rates in foreign bank claims on emergingeconomies. Table 3B.1 reports the results of our
To more carefully investigate the relationshipbetween global liquidity conditions and inter-
national banks’ lending behavior toward develop-ing countries, we specify a linear model of credit toemerging economies as a function of the contempo-raneous and lagged three-month policy spread(OIS); its volatility; a lagged dependent variable
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 112
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
113
estimation, with country fixed effects and clusteredstandard errors or regional dummy variables.
The lagged OIS spread as an indicator of theavailability (low) or tightness (high) of interbankliquidity persistently comes out negative and sta-tistically significant (p-values in parentheses)across all specifications, whereas the contempora-neous policy spread is statistically less significantand positive but the (steady state) net effect isgenerally negative. This result reflects banks’ oper-ational policies that will offer credit only afterhaving secured the necessary funding on their partin advance so that past access to liquidity mattersmore than current access.
To examine the impact of tightening creditstandards in developed countries on lending todeveloping countries, we looked at another set ofmultivariate regressions with country fixed effectsand clustered standard errors or regional dummyvariables, in which we related the (logarithm of)foreign bank claims on emerging economies to thefraction of U.S. banks reporting tighter credit stan-dards in a given quarter, its lags, and macroeco-nomic and institutional control variables. As shownin table 3B.2, the results confirm that there is astatistically significant negative impact of tightenedlending standards in the United States on lendingto developing countries.
Table 3B.2 Multivariate analysis of credit to emerging economies
Log(foreign claims) 1st diff log(foreign claims)
Dependent variable (1) Fixed effects (2) Fixed effects (3) Fixed effects (4) Region (5) Region (6) Region
Lagged log(fc) 0.811 0.81 0.88(0.000)*** (0.000)*** (0.000)***
Log(GDP) 0.233 0.225 0.212 0.001 0.001 0.001(0.000)*** (0.000)*** (0.000)*** �0.761 �0.776 �0.604
Inflation 0.024 0.017 0.01 �0.044 �0.044 �0.005�0.786 �0.85 �0.863 �0.468 �0.473 �0.902
Growth 0.182 0.169 �0.045 0.184 0.175 0.164�0.266 �0.303 �0.702 �0.174 �0.197 (0.083)*
Tighter U.S. credit �0.054 0.079 �0.067 0.056standards (0.068)* (0.065)* (0.005)*** �0.194
Lag1 tightening �0.066 �0.117 �0.07 �0.115(0.035)** (0.057)* (0.003)*** (0.007)***
Lag2 tightening 0.036�0.395
ICRG composite �0.002 0.001�0.283 (0.050)*
Europe and Central Asia 0.034 0.034 0.054(0.087)* (0.085)* (0.000)***
Latin America and the Caribbean �0.023 �0.024 �0.006�0.219 �0.215 �0.661
Middle East and North Africa �0.027 �0.027 �0.013�0.267 �0.269 �0.426
South Asia �0.01 �0.01 0.019�0.705 �0.705 �0.354
Sub-Saharan Africa �0.024 �0.024 0.006�0.189 �0.186 �0.665
Constant �0.807 �0.724 �1.04 0.031 0.033 �0.079(0.001)*** (0.006)*** (0.000)*** �0.323 �0.298 (0.082)*
Observations 2,999 2,999 2,301 2,991 2,991 2,296Countries 114 114 87R2 0.743 0.743 0.865 0.011 0.011 0.038
Source: World Bank staff.Note: The data on the fraction of U.S. banks reporting tighter credit standards in any given quarter is from the U.S. Federal Reserve’s “SeniorLoan Officer Opinion Survey.” ICRG � International Country Risk Guide. * significant at the 10% level; ** significant at the 5% level; *** significant at the 1% level.
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 113
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
114
Annex 3C:The impact of foreignbank presence on the transmission of monetary policy
The data used to estimate the model consistof quarterly observations from 22 developingcountries, whose selection was based on data avail-ability.19 We used quarterly observations from thefirst quarter of 1995 to the third quarter of 2007,with some missing observations. The data containseries of money-market interest rates, lending in-terest rates, GDP, M2 (broad money), domesticcredit, and the fraction of total assets in the bankingsector owned by foreign banks. The series camefrom the IMF’s International Financial Statisticsdatabase, except for the foreign bank data, whichwere obtained from Bankscope and other officialsources, and the nominal GDP series for Mexico,Russia, Uruguay, and República Bolivariana deVenezuela, which were downloaded from officialsources in these countries.20 Table 3C.1 presents
To study how foreign bank presence affectsthe transmission of monetary policy, we
specify a linear model of lending rates as a func-tion of the money-market rate and control vari-ables that capture the degree of financial deepen-ing. The interaction term between money-marketrate and control variables is added to measurehow the financial deepening variables, includingthe degree of foreign bank presence, affect thesensitivity of lending rates to money-market rates.The model constrains the slope coefficients to beidentical across countries but allows for a coun-try-specific intercept. We use the error correctionframework developed by Pesaran, Shin, andSmith (2000) to allow for more flexibility acrosscountries, especially in terms of different short-run dynamics.
Table 3C.1 Lending rate estimates
Lending rates Estimate 1 Estimate 2 Estimate 3 Estimate 4 Estimate 5
Money market 1.04 1.02 0.92 0.91 0.94[0.02]*** [0.02]*** [0.02]*** [0.02]*** [0.02]***
M2/GDP �0.05 �0.05[0.01]*** [0.01]***
Credit/GDP �0.04 �0.04[0.01]*** [0.01]***
Foreign banks 0.17 �0.24 �0.85[0.54] [0.55] [0.52]
Money market � M2/GDP 0.0005 0[0.0001]*** [0.0002]***
Money market � credit/GDP 0.0005 0[0.0002]** [0.0002]***
Money market � foreign banks �0.09 �0.08 �0.03[0.02]*** [0.02]*** [0.02]
Average speed of adjustment �0.21 �0.21 �0.27 �0.25 �0.25[0.03]*** [0.04]*** [0.05]*** [0.05]*** [0.05]***
Number of observations 933 933 848 826 826
Source: World Bank staff.Note: M2 � broad money. ** significant at the 5% level; *** significant at the 1% level.
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 114
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
115
pooled mean group estimates when the controlvariables include the ratio of M2 to GDP (M2/GDP),the ratio of domestic credit to GDP (credit/GDP),and the fraction of assets in the banking sectorowned by foreign banks (foreign banks), all in log-arithms. Because of the high collinearity betweenM2/GDP and credit/GDP, we did not include bothregressors simultaneously.
From this table we conclude:
• As expected, money-market rates are highlysignificant and with coefficients close to 1,suggesting a large long-run pass-though.
• Economies with deeper financial systems, asmeasured by M2/GDP and credit/GDP, havelower lending rates.
• Economies with deeper financial systems, asmeasured by M2/GDP and credit/GDP, havehigher sensitivity of lending rates to money-market rates (see the positive and significantcoefficients in rows 5 and 6).
• The presence of foreign banks does not seemto affect the levels of lending rates.
• Foreign bank presence reduces the sensitivityof lending rates to money-market rates (seethe significantly negative coefficients in the in-teraction term of row 7).
• The dynamics of the pass-through are stable:the average speed of adjustment is significant,and between �2 and 0.
Summarizing, the estimates shown in table3C.1 suggest that deeper financial markets in-crease the pass-through of interest rates, but ahigher foreign bank presence reduces the transmis-sion of policy interest rates. This last result is con-sistent with the view that foreign banks are lesssensitive to domestic monetary conditions becauseof their access to a large pool of funds beyond thecontrol of the monetary authority.
Notes1. Data on foreign bank claims on developing-country
residents are from the BIS (consolidated banking statistics).They measure claims denominated in foreign currency aswell as the local currency of the country in which the bor-rower is domiciled. The number of countries whose banksreport foreign claims to the BIS has increased from 10 in1964—Belgium, France, Germany, Italy, Luxembourg, theNetherlands, Sweden, Switzerland, the United Kingdom,and Japan—to 30 today, including all members of the Or-ganisation for Economic Co-operation and Development
plus Brazil, Chile, Hong Kong (China), India, Panama, andSingapore.
2. By definition, FDI is “investment made to acquirelasting interest in enterprises operating outside of the economyof the investor,” where lasting interest is defined as 10 per-cent or more of the ordinary shares or voting power of anincorporated firm or its equivalent for an unincorporatedfirm. FDI in the banking sector is proxied by FDI in finan-cial sector data, which are collected from central banks ofselected economies. The definition of the banking sector,however, may differ among countries. The FDI data arecompiled for Argentina, Brazil, Colombia, Peru, and Mexicoin Latin America; Bulgaria, Hungary, Kazakhstan, Poland,Romania, Russia, the Slovak Republic, and Turkey in Europeand Central Asia; Pakistan in South Asia; and China, In-donesia, Malaysia, Pakistan, the Philippines, Thailand, andVietnam in East Asia. Cross-border M&A transactions inthe banking sector reflect purchased domestic banks in 150developing countries by nonresidents as recorded at the timeof closure of the deals. M&A values may not be paid out ina single year and may also include the financing that is gen-erated in the host country. The foreign bank database usedin Claessens and others (2008) includes bank-specific infor-mation for all banks operating in 100 developing countriesduring 1995–2006. These data also include foreign banks,defined as banks domiciled in a developing country but 50 per-cent or more owned by foreign nationals in a given year.
3. This figure includes all transactions that led to atleast 10 percent minority share holdings as well as expan-sion of existing foreign banks.
4. In the case of U.S. bank branches, section 25C ofthe Federal Reserve Act establishes that “a member bankshall not be required to repay any deposit made at a foreignbranch of the bank if the branch cannot repay the depositdue to an act of war, insurrection, or civil strife or (2) anaction by a foreign government or instrumentality (whetherde jure or de facto) in the country in which the branch islocated, unless the member bank has expressly agreed inwriting to repay the deposit under those circumstances”(Cerutti, Dell’Ariccia, and Martinez Peria 2005).
5. Banks have traditionally been heavily regulated for anumber of reasons including potential systemic risk and pol-icy makers’ desire to control and influence the supply andallocation of credit. A large literature exists on the degreeand nature of such banking regulation in both developedand developing country; see Dinç (2003); Demirgüç-Kunt,Laeven, and Levine (2004); and Bertrand, Schoar, andThesmar (2007). For more detail on barriers against for-eign competition, see Berger (2007) and Berger and others(2008).
6. Limited foreign entry was permitted in 1992 andwas expanded in 1994 with new bank regulations and theadoption of NAFTA. Following the Tequila crisis in late1994, the government further relaxed foreign bank acquisi-tions and kept an ownership cap in only the three majordomestic banks. In 1999 this cap was abolished, andin 2001 FDI in the Mexican banking sector surged withthe acquisition of Banamex by Citigroup, a deal valued at$12.5 billion.
7. China has removed geographic and client restrictionsand allowed foreign banks to establish locally incorporated
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 115
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
116
subsidiaries to provide full renminbi services to all clients,but it maintains a cap on foreign ownership of a domesticbank at 25 percent, with a limit of 20 percent on a single for-eign shareholder.
8. Note that this argument refers to the medium-termimpact of foreign bank entry. The short-term implication offinancial sector liberalization, which often includes openingto foreign capital inflows, is a more complicated subject.
9. The literature has reached different conclusionsregarding the efficiency of domestic versus foreign banks indeveloping countries. For example, Martinez Peria andMody (2004) find that foreign banks charge lower spreadsand have lower costs than domestic banks, while Claessens,Demirgüç-Kunt, and Huizinga (2001) report that for low-income countries, foreign banks had significantly higher netinterest margins, overhead expenses, and profitability thandomestic banks (these comparisons tended to be not signifi-cant, or reversed, for middle-income countries).
10. An overnight index swap is a fixed-rate/floating-rate swap, where the floating-rate leg is linked to a dailyovernight reference rate during the term of the swap.
11. During a recession, when even borrowers represent-ing otherwise acceptable credit risks might not be able to ser-vice their debt, banks tend to exert more effort in identifyingabove-average borrowers. In the current credit crunch, how-ever, the pool of acceptable credit risks has dwindled somuch that the marginal benefit of more intensive screening isnot worth the extra expenditure of time and cost (Ruckes2004). As a consequence of the decrease in information col-lection, banks are likely to reduce their credit offers. But asthe economic outlook improves, and the average repaymentprobability of borrowers rises along with it, lenders willbe willing to spend more on borrower screening becauseexpected returns on that activity will also increase.
12. Blank and Buch (2007) report that cross-borderlending not only responds to macroeconomic shocks butalso contributes to their propagation, echoing the findingsof Forbes and Chinn (2004), who show that bilateral banklending was an important determinant of cross-countryfinancial links and the transmission of market shocks in thelate 1990s. In analyzing the determinants of the amount ofbilateral cross-border assets and liabilities in OECD coun-tries, Blank and Buch (2007) find that geographical distancehas a negative effect on banks’ cross-border assets, so thatbanks limit their exposure in unfamiliar markets where dis-tance exacerbates difficulties in information collection(Agarwal and Hauswald 2006).
13. Developing countries contracted a total of $68 bil-lion of syndicated loans in the fourth quarter of 2007, com-pared with $81 billion in the fourth quarter of 2006 and animpressive $126 billion in the third quarter of 2007. The fig-ure declined to $56 billion in the first quarter of 2008, com-pared with $94 billion a year ago. There were 324 and 164deals in the fourth quarter of 2007 and first quarter of 2008,respectively, compared with 418 in the third quarter of 2007.
14. The sample of countries is those with an average an-nual growth rate above 33 percent in the period 2003–06.These countries are Albania, Angola, Armenia, Azerbaijan,Belarus, Democratic Republic of Congo, Georgia, Ghana,Guinea, Guinea-Bissau, Islamic Republic of Iran, Kazakhstan,Kyrgyz Republic, Latvia, Liberia, Lithuania, Malawi,
Mongolia, Montenegro, Romania, Russia, Serbia, Tajikistan,Tanzania, Ukraine, República Boliviana de Venezuela, andZambia.
15. In analyzing the relationship between foreign bankpresence and private credit growth, we estimate the follow-ing model with time and regional fixed effects using paneldata for 51 countries over the period 1995–2005:
�PCGDPi,t � � � � foreign_banki,t � � controlsi,t � �i,t ,
where the dependent variable is the first difference of privatecredit/GDP, foreign bank is the ratio of foreign bank assets tototal banking assets and the control variables include laggedGDP growth, logarithm of GDP per capita, the ratio of stockmarket capitalization to GDP, inflation, ICRG compositerating, KOF index of globalization economic openness, cred-itor rights, number of foreign banks as a proportion of totalbanks, ratio of overseas borrowing by banking sector toGDP, and a banking crisis dummy. Regression results showthat the relationship between foreign bank presence and pri-vate credit growth is positive and statistically significant.
16. Indeed, many developing countries initially placedlittle emphasis on prudential regulation, because they hadinherited colonial-era financial systems dominated by estab-lished and reputable foreign banks subject to strict pruden-tial control from home country authorities (Brownbridgeand Kirkpatrick 2000).
17. These have been set out in Minimum Standards forthe Supervision of International Banking Groups and theirCross-Border Establishment (1992); The Supervision ofCross-Border Banking (1996); and subsequent reports bythe Working Group on Cross-Border Banking.
18. For burden-sharing issues arising in the context ofthe European banking system, see Srejber (2006).
19. The countries in the sample are Argentina, Bolivia,Brazil, Bulgaria, Chile, Colombia, Czech Republic, Estonia,Hungary, Latvia, Malaysia, Mauritius, Mexico, Moldova,Peru, Poland, Russia, Slovak Republic, Thailand, Ukraine,Uruguay, and República Boliviana de Venezuela. The panelis unbalanced.
20. The banking data come in annually. Quarterlyobservations were log-linearly interpolated. For the construc-tion of the banking data, see Claessens and others (2008).
References
Agarwal, S., and R. Hauswald. 2006. “Distance and Infor-mation Asymmetries in Lending Decisions.” Unpub-lished paper, American University, Washington, DC.
Arnold, Jens, Beata Javorcik, and Aaditya Mattoo. 2007.“Does Services Liberalization Benefit Manufactur-ing Firms? Evidence from the Czech Republic.”Policy Research Working Paper 4109, World Bank,Washington, DC.
Arnone, Marco, Salim M. Darbar, and Alessandro Gambini.2007. “Banking Supervision: Quality and Gover-nance.” Working Paper 07/82. International MonetaryFund, Washington, DC.
Barth, J., G. Caprio, and R. Levine. 2001. “Banking Systemsaround the Globe: Do Regulations and OwnershipAffect Performance and Stability?” In Prudential
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 116
T H E C H A N G I N G R O L E O F I N T E R N A T I O N A L B A N K I N G I N D E V E L O P M E N T F I N A N C E
117
Supervision: What Works and What Doesn’t, ed. R.Mishkin. Chicago: University of Chicago Press.
Barth, J., J. Marchetti, D. Nolle, and W. Sawangn-goenyuang. 2008. “WTO Commitments vs. ReportedPractices on Foreign Bank Entry and Regulation: ACross-Country Analysis.” Available at http://ssrn.com/abstract=1033268.
Baudino, Patrizia, Giacomo Caviglia, Ettore Dorrucci, andGeorges Pineau. 2004. “Financial FDI to EU AccessionCountries.” European Central Bank, Frankfurt. Avail-able at http://www.bis.org/publ/cgfs22ecb.pdf.
Beck, Thorsten, Asli Demirgüç-Kunt, and Vojislav Maksi-movic. 2004. “Bank Competition and Access to Finance:International Evidence.” Journal of Money, Credit andBanking 33: 627–48.
Berger, Allen N. 2007. “Obstacles to a Global Banking Sys-tem: ‘Old Europe’ versus ‘New Europe’.” Journal ofBanking and Finance 31(7): 1955–73.
Berger, Allen N., Leora F. Klapper, Maria Soledad MartinezPeria, and Rida Zaidi. 2008. “Bank Ownership Typeand Banking Relationships.” Journal of FinancialIntermediation 17(1): 37–62.
Bertrand, Marianne, Antoinette Schoar, and David Thesmar.2007. “Banking Deregulation and Industry Structure:Evidence from the French Banking Reforms of 1985.”Journal of Finance 62 (2): 597–628.
BIS (Bank for International Settlements). 2004. “ForeignDirect Investment in the Financial Sector of Emerg-ing Market Economies.” Committee on the GlobalFinancial System, Bank for International Settlements,Geneva.
Blank, Sven, and Claudia M. Buch. 2007. “InternationalBank Portfolios: Short- and Long-Run Responses tothe Business Cycle.” Available at http://ssrn.com/abstract=985070.
Brownbridge, M., and C. Kirkpatrick. 2000. “FinancialRegulation in Developing Countries.” Journal of De-velopment Studies 37 (1): 1–24.
Bruno, Valentina, and Robert Hauswald. 2007. “The RealEffects of Foreign Banks.” Unpublished paper, AmericanUniversity, Washington, DC.
Cerutti, Eugenio, Giovanni Dell’Ariccia, and Maria SoledadMartinez Peria. 2005. “How Banks Go Abroad:Branches or Subsidiaries?” World Bank Policy Re-search Working Paper No. 3753. Available at http://ssrn.com/abstract=844766
Claessens, Stijn, Asli Demirgüç-Kunt, and Harry Huizinga.2001. “How Does Foreign Bank Entry Affect DomesticBanking Markets?” Journal of Banking and Finance25: 891–911.
Claessens, Stijn, and Neeltje van Horen. 2008. “LocationDecisions of Foreign Banks and Institutional Com-petitive Advantage.” Available at http://ssrn.com/abstract=904332.
Claessens, Stijn, Neeltje van Horen, Tugba Gurcanlar, andJoaquin Mercado Sapiain. 2008. “Foreign BankPresence in Developing Countries 1995–2006: Dataand Trends.” Available at http://ssrn.com/abstract=1107295.
Clarke, George R. G., Robert Cull, and Maria SoledadMartinez Peria. 2006. “Foreign Bank Participation
and Access to Credit across Firms in DevelopingCountries.” Journal of Comparative Economics 34:774–95.
Coppel J., and M. Davies. 2003. “Foreign Participation inEast Asia’s Banking Sector.” Contribution to the CGFSWorking Group on FDI in the Financial Sector ofEmerging Market Economies. Available at http://www.bis.org/publ/cgfs22cbpapers.htm.
Corsetti, Giancarlo, and Paolo Pesenti. 2005. “InternationalDimensions of Optimal Monetary Policy.” Journal ofMonetary Economics 52: 281–305.
Cull, Robert, and Maria Soledad Martinez Peria. 2007.“Foreign Bank Participation and Crises in DevelopingCountries.” Policy Research Working Paper 4128.World Bank, Washington, DC. Available at http://ssrn.com/abstract=961096.
De la Torre, Augusto, Eduardo Levy-Yeyati, and SergioL. Schmukler. 2002. “Argentina’s Financial Crisis:Floating Money, Sinking Banking.” World Bank,Washington, DC.
Demirgüç-Kunt, Asli, Luc Laeven, and Ross Levine. 2004.“Regulations, Market Structure, Institutions, and theCost of Financial Intermediation.” Journal of Money,Credit and Banking 36 (3): 593–622.
Detragiache, Enrica, Poonam Gupta, and Thierry Tressel.2006. “Foreign Banks in Poor Countries: Theory andEvidence.” IMF Working Paper 6/18. InternationalMonetary Fund, Washington, DC.
Dinç, Sedar. 2003. “Government Ownership of Banks andPolitical Lending in Developing Countries.” Unpub-lished paper, Ross School of Business, University ofMichigan.
ECB (European Central Bank). 2008. “The Euro Area BankLending Survey.” Frankfurt.
ECLAC (Economic Commission for Latin America and theCaribbean). 2002. Foreign Investment in Latin Amer-ica and the Caribbean. May 2001 Report LC/G.2178-P.Santiago, Chile.
Financial Stability Forum. 2008. “Report of the FinancialStability Forum on Enhancing Market and Institu-tional Resilience.” Bank for International Settle-ments, Basel, Switzerland. Available at http://www.fsforum.org/publications/FSF_Report_to_G7_11_April.pdf.
Forbes, Kristin J., and Menzie D. Chinn. 2004. “A Decom-position of Global Linkages in Financial Markets overTime.” Review of Economics and Statistics 86 (3):705–22.
Galindo, A., A. Micco, and C. Serra. 2003. “Better theDevil That You Know: Evidence on Entry Costs Facedby Foreign Banks.” Working Paper 477, Inter-AmericanDevelopment Bank, Washington, DC.
Garber, Peter. 2000. “What You See vs. What You Get: De-rivatives in International Capital Flows.” In Manag-ing Financial and Corporate Distress: Lessons fromAsia, ed. Charles Adams, Robert E. Litan, andMichael Pomerleano. Washington, DC: BrookingsInstitution.
Gelos, R. G., and Jorge Roldos. 2004. “Consolidation andMarket Structure in Emerging Market Banking Sys-tems.” Emerging Markets Review 5: 39–59.
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 117
G L O B A L D E V E L O P M E N T F I N A N C E 2 0 0 8
118
Giannetti, Mariassunts, and Steven Ongena. 2005. “Finan-cial Integration and Entrepreneurial Activity: Evidencefrom Foreign Bank Entry in Emerging Markets.”Working Paper 91/2005, European Corporate Gover-nance Institute, Brussels.
Goldberg, Linda. 2004. “Financial-Sector FDI and HostCountries: New and Old Lessons.” Working Paper10441, National Bureau of Economic Research,Cambridge, Mass.
Hagmayr, Bettina, Peter R. Haiss, and Kjell Sümegi. 2007.“Financial Sector Development and EconomicGrowth: Evidence for Southeastern Europe.” Availableat http://ssrn.com/abstract=968253.
International Swaps and Derivatives Association. 2007.Market Survey. http://www.isda.org/.
Levine, Ross. 2001. “International Financial Liberalizationand Economic Growth.” Review of InternationalEconomics 9 (4): 688–701.
Martinez Peria, Maria Soledad, and Ashoka Mody. 2004.“How Foreign Participation and Market ConcentrationImpact Bank Spreads: Evidence from Latin America.”Journal of Money, Credit and Banking 36: 511–37.
Mian, A., 2004. “Distance Constraints: The Limits ofForeign Lending in Poor Economies.” The Journalof Finance 61: 1465–1505.
Peek, J., and E. S. Rosengren. 2000. “Implications of theGlobalization of the Banking Sector: The Latin Ameri-can Experience.” Federal Reserve Bank of Boston NewEngland Economic Review (September/October):45–62.
Pesaran, M. H., Y. Shin, and R. J. Smith. 2000. “StructuralAnalysis of Vector Error Correction Models with
Exogenous I(1) Variables.” Journal of Econometrics97: 293–343.
Rajan, Raghuram G., and Zingales, Luigi. 1998. “FinancialDependence and Growth.” American Economic Review88 (3): 559–86.
Rose, Andrew K. 2006. “A Stable International MonetarySystem Emerges: Inflation Targeting Is Bretton Woods,Reversed.” Working Paper 12711, National Bureau ofEconomic Research, Cambridge, Mass.
Ruckes, M. 2004. “Bank Competition and Credit Standards.”Review of Financial Studies 17 (4): 1073–102.
Srejber, Eva. 2006. “Footing the Bill for Europe’s Next Bank-ing Crisis.” The Financial Regulator 11(1): 29–35.
Sturm, Jan-Egbert, and Barry Williams. 2005. “WhatDetermines Differences in Foreign Bank Efficiency?Australian Evidence.” CESifo Working Paper No. 1587.http://ideas.repec.org/p/ces/ceswps/_1587.html
UNCTAD (U.N. Conference on Trade and Development).2006. “Measuring Restrictions on FDI in Services inDeveloping Countries and Transition Economies.”New York. http://www.unctad.org/Templates/Webflyer.asp?intItemID=1397&docID=7158.
UNIDO (United Nations Industrial Development Office).2005. Industrial Statistics. Geneva.
U.S. Federal Reserve Board. 2008. “Senior Loan OfficerOpinion Survey on Bank Lending Practices.” Washing-ton, DC (January).
Wooldridge, Philip D., Dietrich Domanski, and AnnaCobau. 2003. “Changing Links between Mature andEmerging Financial Markets.” BIS Quarterly ReviewSeptember: 45–57. Available at http://www.bis.org/publ/qtrpdf/r_qt0309e.pdf.
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 118
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 119
gdf_ch03_081-120.qxd 5/26/08 3:18 PM Page 120