A4 Martinez Mody (2004)

How Foreign Participation and Market Concentration Impact Bank Spreads: Evidence from
Latin America
Author(s): Maria Soledad Martinez Peria and Ashoka Mody
Source: Journal of Money, Credit and Banking, Vol. 36, No. 3, Part 2: Bank Concentration and
Competition: An Evolution in the Making A Conference Sponsored by the Federal Reserve
Bank of Cleveland May 21-23, 2003 (Jun., 2004), pp. 511-537
Published by: Ohio State University Press
Stable URL: http://www.jstor.org/stable/3838950 .
Accessed: 11/07/2014 04:30
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
Ohio State University Press is collaborating with JSTOR to digitize, preserve and extend access to Journal of
Money, Credit and Banking.
http://www.jstor.org
This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

All use subject to JSTOR Terms and Conditions
MARIA SOLEDAD MARTINEZ PERIA
ASHOKA MODY
How Foreign Participationand Market

ConcentrationImpact Bank Spreads:Evidence
from Latin America
Increasingforeign participationand high concentrationlevels characterize

the recent evolution of banking sectors' market structuresin developing
countries.We analyze the impact of these factors on Latin Americanbank
spreadsduringthe late 1990s. Our results suggest that foreign banks were
able to charge lower spreadsrelative to domestic banks. This was more so
forde novo foreignbanksthanforthosethatenteredthroughacquisitions.The
overall level of foreign bank participationseemed to influence spreads
indirectly,primarilythroughits effect on administrativecosts. Bankconcen-
trationwas positively and directlyrelatedto both higher spreadsand costs.
JEL codes: D4, G21, Lll, N26

Keywords:marketstructure,concentration,
foreign bank participation,bank spreads.
The marketstructureof the bankingindustryin many devel-

oping countrieshas recentlyundergonesignificantchanges.In particular,the ongoing
and, often, extensive entry of foreign banks has been the sournce of a far-reaching
transformation.Between 1994 and 1999, the share of assets held by foreign banks
(i.e., those banks that are at least 50% foreign) increased from 7.8% to 52.3%
among countries in EasternEurope (IMF 2000). For countries in Latin America,
the correspondingfigures reflect an increase in foreign bank participationfrom
The opinions expressed in this paper are entirely those of the authorsand do not reflect the views
of the IMF,the WorldBank, or theirexecutive directors.We thankAdolfo Barajas,GiovanniDell'Aricia,
Asli Demirgiiu-Kunt,Tom Glaessner,StephenHaber,Sergio Schmukler,Jessica Serrano,andparticipants
in the WorldBank conference on Bank Concentrationand the JMCB/ClevelandFederalReserve Bank
conference on Banking Consolidationand Competitionfor comments and suggestions. We are greatly
indebted to Juan Miguel Crivelli and Adrian de la Garza for excellent researchassistance.
MARIASOLEDADMARTINEZPERIAis a senior economist in the Finance Research Group
of the World Bank. E-mail: mmartinezperia@worldbank.org ASHOKAMODY is a Division
Chief at the Research Department of the IMF. E-mail: mmody@imf.org
Received May 29, 2003; and accepted in revised form May 29, 2003.
Journal of Money, Credit, and Banking, Vol. 36, No. 3 (June 2004, Part 2)
Published in 2004 by The Ohio State University Press.

512 : MONEY, CREDIT,AND BANKING
13.1% to 44.8% over this period. The rise in foreign bank participationhas often
occurredin the context of alreadyhigh and in some countriesrising levels of bank
concentration.Among a sample of 33 developing countries,the level of bank assets
held by the three largest banks averaged64% during 1995-99.1
The increase in foreign bank presence and the high levels of bank concentration
in developing countries have been the result of a number of factors, some of
them interrelated.Growingforeign bank participationhas been a facet of the larger
process of financial liberalization and internationalintegration experienced by
developingcountriesin recentyears.In manycases, however,foreignentryarosefrom
the exigencies of crises, encouragedby local bankingauthoritiesto reducethe costs
of recapitalizingthe domestic financial system. The high levels of concentration
have also been, in part,the consequenceof bank closures, mergers,and acquisitions
following crises. Furthermore,foreign bank entry, in some cases, contributedto
bank concentration.In many countries,foreign banks enteredthe system mainly by
acquiringexisting domestic banks. Also, it is possible that in some countries the
observed domestic bank consolidation and concentrationoccurred in response to
foreign competition.
Foreignbankentryandbankconcentrationmay influencemany aspectsof banking
sectorsin developingcountries.2A questionof interestfor policymakersandacadem-
ics alike is the impact of these factors on bank spreads-the differencebetween the
rate chargedto borrowersand the rate paid by depositors. Spreadsare commonly
interpretedas a measure of the cost of financialintermediation(see Saundersand
Schumacher,2000, Brock and Rojas Suarez, 2000). High spreads can hinder the
growthof savingsandinvestmentand imply thatthe cost of using the financialsystem
maybecomeprohibitivefor certainborrowers.Furthermore, the impactof high spreads
is likelyto be moreseverefordevelopingcountrieswhere,given thatcapitalmarketsare
generally small and under-developed,a larger percentageof firms and individuals
tend to dependon banks to meet their financialneeds.
A number of recent papers have investigated the impact of foreign entry on
bank spreadsand other variables (see for example Claessens, Demirgiiu-Kunt,and
Huizinga, 2000, Barajas, Steiner, and Salazar, 2000, Denizer, 2000). Yet, only
some of these studies have also taken into account the parallel trend toward more
consolidationin the sector.None has examinedhow differenttypes of foreign bank
entry affect spreads.
At the same time, manypapershave focused on the impactof concentrationon the
degree of competitionin the bankingsectorand,hence, on overallbankprofitability.3
However, typically, these studies have not examined the direct impact on bank
1. Figures obtainedfrom Demirgiiq-Kunt,Laeven, and Levine (2004).

2. For a review of the potential consequences of foreign bank participation,see Levine (1996). For
a discussion of the impact of bank concentrationon profitability,see Berger (1995). For a test of
whether bank consolidation and concentrationhas worsened competition in developing countries, see
Gelos and Rold6s (2002).
3. See Gelos and Rold6s (2002) for developing countries and Berger (1995) for a review of the
U.S. evidence.

MARIA SOLEDAD MARTINEZPERIA AND ASHOKA MODY : 513
spreads. Demirgtiu-Kunt,Laeven, and Levine (2004) is an exception. This paper

analyzes the effect of concentrationand bankregulationon spreads,but in this study
little attention is paid to the simultaneousimpact that foreign bank participation
could have on spreads.
Ourpaperinvestigatesthe impact of foreign bank participationand concentration
on bank spreads in a sample of developing Latin American countries during the
late 1990s. Using bank-leveldatafor Argentina,Chile, Colombia,Mexico, andPeru,
we examine a numberof hypotheses. First, we investigate whether foreign banks
are able to operate with lower spreads, directly benefiting borrowers.We refer to
this effect as the "own effect" of foreign bank presence.
Second, we examine whetherthe type of foreign bank entry influences how big
the "own effect" might be. In other words, among the foreign banks we distinguish
between those that entered or increased their presence in the system by acquiring
domestic banks and those that established de novo operations. Dell'Aricia and
Marquez(2004) suggest that even if all banks are equally cost efficient,they might
changespreadsbasedon theirspecializationin differentmarketsegments.An alterna-
tive explanationfor variationsin spreadsacross foreign banks might be related to
differencesin the pricing strategiespursuedto gain marketshare, even if they lend
to the same segments.Thoughwe are not able to formallytest whethervariationsin
market segments or in pricing strategies account for differencesin spreads across
banks, to our knowledge, our study is the first to examine whether all forms of
foreign entry have the same impact on spreads.
Third,we analyze whetherthereis a "spillovereffect" as a result of foreign bank
participation.Thatis, once we controlfor the origin (domesticor foreign) of individ-
ual banks, we test whether the overall level of foreign bank participationin the
banking system raises or lowers spreadsacross the board, and in particularamong
domestic banks.A priori,the spillovereffect of foreign bankparticipationis indeter-
minate. Spreadswould be lowered if foreign banks compete directly with domestic
banks forcing them to lower their spreads.4Alternatively,faced with foreign bank
competition, domestic banks may redirecttheir lending to segments that are more
opaque and where they have an informationaladvantageand greatermarketpower,
allowing them to charge higher spreads (Dell'Arricia and Marquez2004).
Finally, we study the impact of bank concentrationon bank spreadsby including
several measuresof system-widebankconcentrationin our estimations.At the same
time, we control for banks' marketshare and for cases of bank consolidation.
We believe this paper contributesto the existing literaturenot only by testing
some hypothesesthathave been overlookedbefore, but also by focusing on a region
that has been at the forefront of the recent changes in bank market structure
in developing countriesand that has been traditionallycharacterizedby having high
spreads.Thus, we feel that Latin America makes for an interestingcase study for
4. Domestic banks may lower spreads either because they are driven to become more efficient
following bank entry (by imitating some of the practicesintroducedby foreign banks) or because they
are forced to give up some of the marginsthey were able to chargebefore. In other words, lower spreads
could be the result of lower costs or lower revenues.

514 : MONEY, CREDIT, AND BANKING
a numberof reasons.First,despite having embarkedin a process of financialmarket

liberalizationduringthe late 1980s and early 1990s, which includedthe elimination
of interest rates and direct credit controls, spreadsin the region remainedhigh by
the mid-1990s.5Second, perhapsafter EasternEurope,Latin America has been the
region to witness the sharpestincrease in foreign bank participation(IMF 2000).
Third,concentrationrose or remainshigh (dependingon the country)in partbecause
many foreignbanksincreasedtheirparticipationby acquiringdomesticbanks.Also,
in many of these countries, there has been a trend toward consolidation among
domestic banks.
Ourempiricalanalysis yields a numberof interestingresults.We find that foreign
banks are able to charge lower spreads and have lower costs than domestic banks.
Moreover, those foreign banks that acquired domestic institutions have higher
spreadsthanthose thatestablishedde novo operations,suggestingeithersome market
segmentationor differencesin pricing strategiesto gain marketshare.However, we
do not find consistent evidence of a direct spillover effect on spreads. Instead,
the degreeof system-wideforeignbankparticipation(as measuredby the shareof total
loans) appearsto influencespreadsthroughits effects on costs. Greaterparticipation
of foreign banks lowers costs all around. On the other hand, the degree of
concentrationin the banking system has a positive and economically significant
impact on both spreadsand costs.
The remainderof the paper is organized as follows. Section 1 describes the
structureof the banking sector and the behavior of bank spreads in Argentina,
Chile, Colombia, Mexico, and Peru over the late 1990s. Section 2 discusses the
empiricalmethodology and data used to study the determinantsof bank spreadsin
Latin America. Section 3 presents the empiricalresults and Section 4 concludes.
1. FOREIGNBANK PARTICIPATION,
CONCENTRATION,
AND SPREADS IN LATIN AMERICA
As in many developing economies, countriesin LatinAmericaexperienceda sig-

nificant increase in foreign bank participationduring the late 1990s (see Table 1).
In Argentina,foreign bank participationincreasedfrom 18.9%in 1995 to 49.4% of
outstandingloans in 2000. In Chile, Mexico, and Peru,the shareof bank loans held
by foreignbanksrosefrombelow 15%in 1995to exceed 40%by theendof the decade.6
Colombiais theonly countryin oursamplewhereforeignbanksconsistentlyaccounted
for one-fourthof the loans duringthe periodunderconsideration.
Despite the dramaticincrease in foreign bank participation,the total numberof
banks in the region droppedin four of the five countries and concentrationlevels
5. Brock and Rojas Suarez (2000) study spreadsin Latin Americaduring 1990-96 and conclude that
they have not gone down significantly(perhapswith the exception of Mexico) and in many cases are
still three times higherthan those observedfor industrialcountries(thoughless so for Chile). In general,
the study finds that high operatingor administrativecosts are particularlysignificant in explaining the
behavior of bank spreadsin the region.
6. In the case of Mexico, the foreign bank share exceeded 40% by 2001.

TABLE 1
BANKMARKET
STRUCTURE
ANDSPREADS
IN LATINAMERICA
1995-2000
Country Variables 1995 1996 1997 1998 1999 2000
Argentina Total numberof 141 122 115 106 96 90

banks
Number of foreign 32 32 35 38 38 40
banks
Foreign bank share 18.9 24.2 30.4 40.9 47.4 49.4
(percent)
Top-3 Bank Share 30.0 29.9 29.5 30.8 32.1 33.9
(percent)
Top-5 Bank Share 40.9 41.7 40.9 43.8 46.7 49.4
(percent)
Herfindahlindex 483.3 489.6 482.6 545.3 605.5 656.7
Average annualized 15.2 11.4 10.6 11.9 12.7 12.3
spreads-all banks
Average annualized 16.7 12.8 12 13.5 15.2 14.2
spreads-domestic
banks
spreads-foreign
banks
Chile Total numberof 31 31 29 29 29 28
banks
banks
(percent)
Top-3 Bank Share 36.6 35.7 42.5 42.1 41.5 41.1
(percent)
Top-5 Bank Share 51.9 52.6 62.5 62.1 61.9 61.5
(percent)
spreads-all banks
Average annualized 4.8 4.5 4.5 4.3 4.5 6
spreads-domestic
banks
spreads-foreign
banks
Colombia Total numberof banks 33 33 28 27
Number of foreign 13 14 12 10
banks
Foreign bank share 27.6 27.6 26.0 24.9
(percent)
Top-3 Bank Share 29.5 31.5 32.3 29.9
(percent)
Top-5 Bank Share 44.1 47.4 50.2 47.3
(percent)
Herfindahlindex 584.7 644.4 714.4 691.6
Average annualized 17 15.9 13.3 11.3
spreads-all banks
Average annualized 18.7 17.4 14.7 13
spreads-domestic
banks
Average annualized 14.2 13.6 11.6 9.1
spreads-foreign
banks
(Continued)

TABLE 1
CONTINUED
Country Variables 1995 1996 1997 1998 1999 2000
Mexico Total numberof banks 39 39 40

Number of foreign 18 18 20
banks
Foreign bank share 13.2 14.0 25.8
(percent)
Top-3 Bank Share 50.0 49.2 47.5
(percent)
Top-5 Bank Share 63.8 62.7 60.8
(percent)
Herfindahlindex 1,108.0 1,055.5 1,078.2
Average annualized 4.3 8.8 7
spreads-all banks
Average annualized 4.2 8.0 7.3
spreads-domestic
banks
Average annualized 4.4 9.8 6.7
spreads-foreign
banks
Peru Total numberof 29 27 27 27 24 20
banks
banks
(percent)
Top-3 Bank Share 60.6 61.2 58.2 54.2 53.3 55.7
(percent)
Top-5 Bank Share 74.4 74.8 70.7 67.0 68.2 72.5
(percent)
spreads-all banks
Average annualized 17 16.6 14.5 12.2 10.9 10.5
spreads-domestic
banks
Average annualized 13.3 19.4 17 13.6 9.9 8.4
spreads-foreign
banks
increased or remained high (see Table 1). In Argentinaand Peru, the number of
banksdeclinedby morethan30%between 1995 and2000. The totalnumberof banks
in Argentinafell from about 141 in 1995 to 90 in 2000. While Peruhad 29 banks in
1995, this numberdroppedto 20 by 2000. For Colombia and Chile, the numberof
banks fell by 18% and 11%,respectively,duringthis period. The one exception is
Mexico, wherethe numberof banksincreasedfrom 39 to 40 between 1997 and 2000.
In all five countries,the share of loans held by the top (i.e., largest) three (five)
banks exceeded 30% (40%) for most of this period and the Herfindahlindex was
above 650. Concentrationlevels rose significantlyfor Argentinaand Chile between
1995 and 2000. In Argentina,the share of loans held by the top five largest banks
increased from 40.9% in 1995 to 49.4% in 2000. Similarly, this share increased
from 51.9% to 61.5% in the case of Chile between 1995 and 2000.

The drop in the numberof banks and the high or rising concentrationlevels can
be ascribedto severalreasons.First,therewere manybankclosuresduringthisperiod.
Such closures typically followed periods of financial distress in the countries,like
the Tequila crisis in Argentinain 1995, when 32 banks closed, and the 1998-99
period of financialturmoil in Colombia, when four institutionswere liquidated.
Second, muchof the increasein foreignbankparticipationresultedfrompurchases
of domestic banks. Thus, foreign entry did not typically add to the number of
banks. In Argentina,these were 16 cases where foreign banks acquireddomestic
financialinstitutionsduring the period 1995-2000. The Spanish banks BBVA and
Santander,the British bank HSBC, and the CanadianScotia Bank were among the
most significant entrantsin Argentina.During the same period, foreign banks ac-
quired five domestic banks in Chile, two in Colombia, and three in Mexico. As in
Argentina,Santander,BBVA, and Scotia were importantplayers in these countries.
Though there were also some truly de novo entries, i.e., cases of foreign banksthat
startedtheir own operationswithout any affiliationwith domestic banks,these were
not the norm.7Six foreign banks set up de novo operationsin Argentina,while two
banks settled in Peru over this period. This explains why the total number of
foreign banks in these countries did not increase at the same pace as the increase
in foreign bank participationin the system.
At the same time, many domestic banks also consolidated with other domestic
banksdue to financialdistressor as a strategyto competewith foreignbanks,bringing
down the total number of institutions. Thirty-seven such transactionstook place
in Argentina,fourin Chile, threein Colombia,andthreein Peruduringthe late 1990s.
What has been the impact of foreign bank participationand concentrationon
bank spreads? While a detailed econometric analysis is required to answer this
question, it is interesting to note some trends in these variables. In Argentina,
Colombia, and Peru, spreadshave droppedduringmost of the late 1990s (see Table
1). A cursory look at the data for these countries suggests that spreads tended to
decline in periods when foreign participationincreased, but concentrationlevels
remained constant. In Chile and Mexico, the increase in foreign participation
appearsto have had little effect on spreadsperhapsbecause both countrieshad high
levels of concentrationat the start of the period and throughout.Note, however,
that by the mid-1990s, spreads in Chile and Mexico were already quite low by
regional standards.
Foreign and domestic bank spreadsappearto move very much in tandemacross
countriesin the region. This behaviorcould signal the influence of macroeconomic
factors and/or similar cost structuresthat affect all banks in the system, as well as
the possibility that in general or at least in certain markets,foreign and domestic
banks compete with each other for customers. However, in Argentina,Colombia,
and Peru, throughoutmost of the sample, and in all countries by the end of the
period, foreign banks seemed to be able to operate with lower spreads.
7. Following the lifting of restrictions on foreign entry, 15 foreign banks initiated operations in
Mexico during 1995 and 1996. These entrantswere small relative to the existing domestic banks, and
the main increase in foreign bank participationoccurred through the acquisition of domestic banks
after 1998.

2. EMPIRICALMETHODOLOGYAND DATA
In this section, we turnto an econometricanalysis of the impact of concentration

and foreign bank presence on bank spreads. In particular,we study the effect of
market structurechanges on bank spreads, while controlling for a host of bank
characteristicsandmacroeconomicvariables,by estimatingregressionsof the follow-
ing form:
Spreadij,t= oo + talLiquidityij,t+ a2AdministrativeCostsi,t + a3NPLsij,t
+ a4Equityij, + oa5BankMarketShareij,t+ a6ForeignBankijt
+ a7ForeignM&Ai, t + a8ForeignM&A x Agei,t
+ octOtherM&Aij,t+ aloOtherM&A x Ageij,t
+ a(l ForeignDe Novoi,t + al2ForeignDe Novo x Ageij,t
+ al3Foreign Bank ParticipationJ,
+ al4Banking Sector Concentrationj,t
+ al5Real OutputGrowthj,+ a16Inflationj,
+ a17Short-Term Real Interest Ratej,t
+ o18Argentinai, + al9Chileij,t+ a2oColombiai,t

+ (2ilMexicoij,t + ?i,t (1)
where i is the bank id, j identifies the country, and t refers to the time period
considered.
Equation(1) is motivatedby the dealershipmodel of bank spreadsdeveloped by
Ho and Saunders(1981), extended by Allen (1988), Angbazo (1997), and others,
and the firm-theoreticalframeworkdeveloped by Zarruck(1989) andWong (1997).8
Both models predict that operating costs, regulatory costs, credit risks, and the
marketstructureof the banking sector can affect spreads.9
In Equation(1), the variableSpreadis the differencebetween the implicit average
interestchargedon loans and the implicit averageinterestpaid on deposits. In other
words, the spread is calculated by taking the total interest received by banks on
loans duringone quarterdividedby the averageloans for thatperiodand subtracting
from it the total interestpaid on deposits throughoutthe quarterdivided by average
deposits. Liquidity is measuredas the ratio of liquid to total assets. Liquid assets
8. According to the dealership approach,banks are risk-aversedealers trying to balance loan and
deposit markets,where loan requests and deposit flows are not necessarily synchronized.In this setup,
bank spreads are interpretedas fees charged by banks for the provision of liquidity under transactions
uncertainty.The firm-theoreticalmodel of banks assumes that these operatein a static frameworkwhere
the demand and supply for loans and deposits clears both markets.
9. A common limitation of the empirical applicationsof these frameworksis that market structure
differencesacross countrieshave been modeled by including countrydummies (see Saundersand Schu-
macher 2000), that is, they have been implicitly assumed to be constant over time.

refer to cash and deposit balancesin otherbanks (includingreserve requirementsat

the central bank). High liquidity ratios, either self-imposed for prudentialreasons
or as a result of regulation(e.g., reserve or liquidity requirements),inflict a cost on
banks since they have to give up holding higher-yielding assets. To the extent
that banks are able to transferthis opportunitycost to borrowers,spreadswill rise
with liquidity ratios.
AdministrativeCosts refer to the ratio of administrativeexpenses (including
payroll and overhead)to average assets. If banks incur high administrativecosts in
the processof providingtheirservices as intermediaries,they arelikely to increasethe
spread they charge their customers. NPLs is the ratio of non-performingloans to
total loans. This variableis intendedto capturecredit risk. Faced with higher credit
risk, banks are likely to charge higher rates on their loans, as equity holders
demand risk-adjustedreturns. Equity refers to the share of bank equity to total
assets. Holding large equity ratios either on a voluntary basis or as a result of
regulationcan be costly for banks. We would expect bank spreadsto rise with this
variable.BankMarketShareis the ratioof each bank'sloans to total system loans. To
the extent that market shares get translatedinto marketpower, banks with higher
sharesof the marketmay be able to chargehigherrates on loans. On the otherhand,
largerbanks may be able to reap economies of scale and may pass on some of these
benefits to their customers in the form of lower spreads.
ForeignBank is a dummy thattakes the value 1 if a bankis foreign at each point
in time. By introducingthis variable, we can test whether the average spread for
foreign banks is significantlydifferentfrom the averagespreadfor domestic institu-
tions. That is, this variable allows us to test for the own effect of foreign bank
presence.ForeignM&A is a dummyvariablethatidentifiesthose transactionswhere
foreignbanksincreasedtheirsize or began operationswithinour sampleby acquiring
domestic banks. Foreign De Novo, on the other hand, is a zero/one variable that
capturesthose foreign banks that set up de novo operationsin a given country.The
purpose of including the latter two variables is to determinehow the spreads for
these banks comparewith those that have been foreign since the startof the sample
andhow differentmodes of foreignbankentryand/orstrategiesto increaseparticipa-
tion in local marketsaffect bank spreads.
We also controlfor othertypes of mergersandacquisitions,namelythose involving
domestic banks or foreign banks, by including the variable Other M&As, which
takes the value 1 for those domestic or foreign banks that acquiredan institution
of the same type. Both M&A variables (i.e., Foreign and Other) plus the dummy
identifyingforeign de novo entryare interactedwith Age, the time since entry(mea-
sured in years), to allow for the possibility that there is an adjustmentperiod until
banks can attain their desired level of spreads after they enter a new market or
purchase/mergewith a bank.
Foreign Bank Participationis the share of loans in the hand of foreign banks.
This variablecapturesthe dynamicimpact of changes in the relative importanceof
foreignbankson the overall level of spreads.In otherwords, this variableis included
to test whetherthere is a spillover effect arising from the presence of foreign banks

in the system. Banking Sector Concentrationmeasuresthe extent to which loans are

concentratedon the hands of few banks within a system. In most of the estimations,
we include threedifferentmeasuresof concentration,namely,the Herfindahlindex-
defined as the sum of squaredloan marketshares-plus the share of loans held by
the top three and top five largest banks, respectively. We expect concentration
measuresto have a positive impacton bank spreads,once we controlfor differences
in cost ratiosacrossbanks.Furthermore,contraryto the literatureon bankconcentra-
tion and profitability,where a positive associationbetweenthese variablescan signal
differentthings, we interpreta positive sign on bank concentrationas an indication
of greatermarketpower and less competition in the banking sector.10
Given that the level of bank spreads can be affected by the macroeconomic
environment in which banks operate, we control for the Inflation rate, the Real
OutputGrowth,and a measureof the money marketShort-TermReal InterestRate.
Following Smith (2001), we include the inflationrate for two reasons. First, given
that bank spreadsare the differencebetween two nominal rates, if inflation shocks
are not passed through to both rates equally fast, then spreads should reflect
this. Second, Cottarelliand Kourelis (1994) have found that inflationcan affect the
flexibility of loan rates and thereforeof bank spreads. The real growth of output
variable could help pick up business cycle effects as those discussed by Beranke
and Gertler (1989) and Kiyotaki and Moore (1997). These studies suggest that
changesin outputcan affectlendingrates,andconsequentlyspreads,becauseborrow-
ers' creditworthinessis countercyclical.As outputgrowthslows down, creditworthi-
ness deterioratesand, otherthings equal, this is likely to be reflectedin higherbank
loan rates and, consequently,spreads. Finally, we include a measure of the short-
term money marketreal interestrate to control for the marginalcost of funds faced
by banks.
We obtainedbank-levelbalancesheet andincome statementdatafromthe Superin-
tendency of Banks in each of the countries in our sample. For Argentina, Chile,
and Peru, the data cover the period 1995-2000. For Colombia, we obtained data
for 1997-2000. For Mexico, where a change in accountingstandardsdoes not allow
us to use data before 1997, the sample studied is 1998-2001. The data frequency
is quarterlyin all cases. The correspondingbank authoritiesalso provideddetailed
accountson the foreign banks operatingin each countryat each point in time along
10. An extensive literatureexists studyingthe impactof concentrationon bankprofitability(see Berger

1995 for a review). While the literatureunanimouslypredictsa positive associationbetweenconcentration
and profitability,differenttheories exist explaining what is behind this result. The structure-conduct-
performancetheory arguesthat bank concentrationsignals marketpower and that a positive association
between profitsand concentrationis unambiguouslybad for the economy. A relatedtheoryis the relative
marketpower hypothesis,which claims thatonly firmswith largemarketshareanddifferentiatedproducts
can obtain marketpower and are able to earn profits above normal.On the other hand, the efficiency-
structurehypothesis contends that larger concentrationlevels and market shares could reflect greater
efficiency by the largest banks, which in turn are able to lower costs and obtain higher profits. While
a problemof observationalequivalenceexists in interpretingthe relationbetween bankconcentrationand
profits, this issue should not arise in analyzing bank spreads.Relatively more efficientbanks should be
able to charge lower spreads, as a result of having lower costs. Consequently,a positive association
between bank spreadsand concentrationshould signal greatermarketpower and less competitionin the
banking sector.

with informationon their mode of entry (e.g., via acquisitionsor by de novo entry).
They also supplied us with the list of mergers and acquisitions among domestic
banks and between existing foreign banks.
Data on inflation,outputgrowth, and the real short-terminterestrate came from
the IMF InternationalFinancial Statistics database. Table 2 contains a detailed
descriptionof the variablesused in this papertogetherwith means andstandarddevia-
tions for each of them.11
3. EMPIRICALRESULTS
Table3 presentsthe estimationresultsfor Equation(1), analyzingthe determinants

of loan-deposit spreadsfor privatebanksin LatinAmerica.12In particular,resultsare
reported for all private banks and, separately, for domestic and foreign banks,
respectively. Throughout,the t-statistics shown were calculated allowing standard
errorsto be correlatedfor observationscorrespondingto the same bank within a
country (i.e., using clustered standarderrorsas describedby Rogers, 1993).
The estimates reportedin Table 3 were obtainedby pooling all countriesin our
sample (Argentina, Chile, Colombia, Mexico, and Peru). However, they are not
intendedto explainvariationsacrosscountries.Instead,becausethey includecountry
fixed-effects,they explainchangesin spreadsover time withina country.The purpose
of pooling observationsin this context is to increasethe power of our estimations.13
At the same time, pooling assumes that the relation between bank spreads and
its determinantscan be characterizedby the same coefficients for all countries.
Thus, as part of our sensitivity analysis, we report and discuss below results in
which we do not include all countries in the estimation.
The determinantsof spreadsmay be categorized into three groups. First, bank-
specificvariablesthatincludeoperationalcharacteristics(such as Liquidity,Non-Per-
forming Loans, and AdministrativeCosts), the bank's market share, whether it is
foreign or domestic, whetherthe formationof the bank was the result of a mergeror
acquisition(M&A), or whetherthe bankwas a new (de novo) entrant,andthe interac-
tion of bank'sage with the foreign and M&A dummies.'4Second, system-widemea-
sures of market structure,including the degree of foreign bank participationand
11. Note that while Table 1 reportsannualizedspreads,Table 2 presentsquarterlyspreads,since the

regressions are conductedwith quarterlyobservations.
12. Becausethe spreadschargedby publicbanksmaybe subjectto constraintsdueto directsubsidiesand
other political considerations,we do not include these banks in our sample. Also, since implicit bank
spreadscalculatedfrom quarterlyincome and balance sheet data can be quite volatile, we exclude those
observations in the top and bottom 5 percentile of the distributionof the change in bank spreads.
The purpose of doing so is to avoid the possibility that outliers drive our results. However, eliminating
these observationsdoes not change the results describedbelow.
13. For example, we are interestedin analyzingif and how the mode of foreign bankentry,by merger
and acquisition or by de novo entry, affects bank spreads.However, there are few such transactionsin
each countryto study this question on a country-by-countrybasis.
14. To addressthe concernof possible reverse causalityfrom spreadsto bank's operationalcharacter-
istics (such as liquidassets, non-performingloans, andmarketshare),we also estimatedsimilarregressions
using one-quarterlags of these variables as regressorswith virtually identical findings. To save space,
these results are available upon request.

TABLE 2
DEFINITION OF VARIABLES USED AND DATA DESCRIPTIVE STATISTICS
Standard
Variable Definitions Source of original data Mean deviationk
Spread Interestincome received Bank superintendencies 0.025 0.016

on loans (over total
loans) minus interest
expenses paid on deposits
(over total deposits)
Liquid Assets Cash and deposits with Bank superintendencies 0.108 0.077
(over total other banks (including
assets) the central bank)
Non-Performing Loans considered to be Bank superintendencies 0.109 0.151
Loans (over non-performingby the
total loans) banking authorities
(in most cases 90
days overdue)
Administrative Includes payroll and other Bank superintendencies 0.016 0.0128
Costs (over operatingexpenses
total assets)
Foreign Bank Dummy equal to 1 when Bank superintendencies 0.412 0.492
bank is at least 50%
foreign owned
Bank Market Share of loans held by Bank superintendencies 0.024 0.043
Share each bank to total loans
Equity (over Bank capital plus reserves Bank superintendencies 0.145 0.116
total assets)
Foreign M&A Dummy equal to 1 for Bank superintendencies 0.062 0.289
cases when a foreign
bank acquireda domestic
bank
Foreign Interactionof Foreign Bank superintendencies 0.048 0.263
M&A x Age M&A with time (in
fraction of years) since
acquisitionof a domestic
bank by a foreign bank
Foreign De Novo Dummy equal to 1 for Bank superintendencies 0.002 0.044
foreign banks that entered
the country by setting up
de novo operations
Foreign De Interactionof Foreign Bank superintendencies 0.002 0.044
Novo x Age De Novo with time
(in fraction of years)
since entry
Other M&A Dummy equal to 1 for Bank superintendencies 0.113 0.439
domestic banks that
acquiredother domestic
banks, or for foreign
banks that acquired
other foreign banks
Other Interactionof Other Bank superintendencies 0.091 0.432
M&A x Age M&A with time
(in fraction of years
since entry)
Foreign Bank Share of loans held by Bank superintendencies 0.273 0.119
Share foreign banks (those
that are at least 50%
foreign owned)
Top-3 Bank Share Share of loans held Bank superintendencies 0.385 0.105
by the top 3 banks
in the system
(Continued)

MARIA SOLEDAD MARTINEZ PERIA AND ASHOKA MODY : 523
TABLE 2
CONTINUED
Standard
Variable Definitions Source of original data Mean deviation
Top-5 Bank Share Share of loans held Bank superintendencies 0.526 0.114
by the top 5 banks
in the system
Herfindahlindex Sum of squaredbank Bank superintendencies 794.405 325.403
marketshares
Inflation Rate of growth of the IMF International 0.011 0.015
consumer price index Financial Statistics
Real Output Rate of growth of IMF International -0.011 0.054
Growth industrial/manufacturing Financial Statistics
production
Real Interest Money marketrate- IMF International 13.189 8.222
Rate inflation Financial Statistics
NOTE: The spreadsreportedhere are quarterlyspreadsas opposed to those shown in Table 1, which are annualizedspreads.
concentration.15 And, finally, variables that control for the macroeconomic environ-
ment-inflation, real growth of production, and the real market interest rate.
For the sample including all banks, among the bank-specific variables, bank
liquidity and administrative costs have a positive and significant impact on bank
spreads in all three specifications, corresponding to the different measures of concen-
tration. Banks that either decide or are required by regulation to hold a high
proportion of their assets in the form of liquid assets seem to charge higher spreads.
This can be interpreted as banks' response to the fact that in holding higher liquidity
ratios, banks forego a return on such assets. However, the impact of higher liquidity on
bank spreads seems to be quantitatively small: a one standard deviation increase
in liquidity raises spreads by 0.14 standard deviation. On the other hand, administra-
tive costs have a larger impact on bank spreads: a one standard deviation change
in administrative costs results in an almost 0.6 standard deviation change in spreads.
As discussed below, administrative costs are influenced by macro country character-
istics (inflation, growth, and domestic interest rates) and subsume their effects in
these regressions, as a consequence of which the macro variables do not appear to
have a direct influence on spreads.
Foreign banks, on average, charge lower spreads (0.5% lower per quarter) than
their domestic counterparts. For foreign banks that enter through an M&A process, the
full effect on spreads is the sum of the Foreign Bank dummy (which has a negative sign)
and the Foreign M&A dummy (which has a positive sign). This sum is negative and
statistically different from zero, as noted in the F-test reported at the bottom of
Table 3. The estimated coefficients indicate that spreads for foreign banks that
15. Because bank origin might be correlatedwith the degree of foreign bank participation(i.e., the
larger the number of foreign banks, the more likely it is that foreign bank participationwill be high)
and bank marketshare might be positively associated with the level of system-wide concentration,we
reestimatedthe equations after excluding these bank-level variables to confirm the robustness of our
findings. Our main findings remain unchanged.These results are available upon request.

TABLE 3
PANEL ESTIMATIONS FOR BANK SPREADS INCLUDING ALL COUNTRIES
All banks Domestic banks
Variables (3.1) (3.2) (3.3) (3.4) (3.5) (3.6)
Liquid Assets 0.028 0.029 0.028 0.053 0.054 0.052

(3.75)*** (3.86)*** (3.68)*** (4.77)*** (4.85)*** (4.72)***
Non-Performing -0.001 -0.00113 -0.00095 0.000331 0.000179 0.000435
loans (0.25) (0.29) (0.24) (0.08) (0.04) (0.11)
Administrative 0.653 0.656 0.652 0.569 0.573 0.568
Costs (7.88)*** (7.94)*** (7.87)*** (7.73)*** (7.79)*** (7.71)***
Foreign Bank -0.005 -0.005 -0.005
(3.92)*** (3.93)*** (3.93)***
Bank Market Share -0.014 -0.014 -0.014 -0.023 -0.024 -0.024
(1.38) (1.37) (1.41) (2.00)** (2.01)** (2.02)**
Equity 0.01 0.01 0.01 0.019 0.019 0.019
(1.59) (1.61) (1.60) (1.35) (1.37) (1.35)
Foreign M&A 0.002379 0.002377 0.002439
(2.54)** (2.55)** (2.61)***
Foreign M&A 0.000209 0.000228 0.000159
x Age (0.22) (0.24) (0.17)
Foreign De Novo -0.022 -0.022 -0.022
(2.08)** (2.08)** (2.07)**
Foreign De 0.007 0.008 0.007
Novo x Age (1.06) (1.08) (1.06)
Other M&A 0.000195 0.000152 0.000207 0.000599 0.000579 0.000616
(0.25) (0.19) (0.26) (0.82) (0.79) (0.84)
OtherM&A 0.000016 0.000048 0.000001 -0.00025 -0.00026 -0.00027
x Age (0.03) (0.08) 0.00 (0.37) (0.38) (0.40)

TABLE 3
CONTINUED
Variables (3.1) (3.2) (3.3) (3.4) (3.5) (3.6)
Foreign Bank Share 0.002 0.002 0.001 0.005 0.003 0.003

(0.71) (0.46) (0.22) (1.12) (0.56) (0.64)
Top-3 Bank Share 0.038 0.05
(3.03)*** (3.26)***
(1.83)* (2.76)***
HerfindahlIndex 0.000011 0.000015
(2.82)*** (3.17)***
Inflation -0.002 0.005 -0.005 -0.023 -0.017 -0.027
(0.07) (0.14) (0.13) (0.57) (0.41) (0.67)
Real Growth of 0.002 0.002 0.003 0.004 0.004 0.005
Production (0.78) (0.67) (0.96) (0.98) (0.87) (1.14)
Real Market 0.000014 -0.000004 0.00003 0.000068 0.000056 0.000089
InterestRate (0.23) (0.06) (0.48) (1.20) (0.96) (1.52)
Argentina 0.007 0.002 0.007 0.014 0.01 0.013
(1.78)* (0.54) (1.59) (3.10)*** (2.45)** (3.01)***
Chile 0.000009 -0.00442 -0.00157 0.003479 -0.00104 0.001457
0.00 (1.91)* (0.63) (1.10) (0.40) (0.54)
Colombia 0.02 0.015 0.018 0.028 0.023 0.026
(4.68)*** (3.95)*** (4.55)*** (5.79)*** (5.60)*** (5.78)***
Mexico -0.013 -0.014 -0.013 -0.009 -0.01 -0.009
(4.47)*** (4.92)*** (4.75)*** (3.29)*** (3.76)*** (3.58)***
Constant -0.00651 0.002967 0.000417 -0.0198 -0.01382 -0.01089
(0.85) (0.39) (0.07) (2.21)** (1.59) (1.60)
Observations 2618 2618 2618 1539 1539 1539 1
R-squared 0.52 0.52 0.52 0.53 0.53 0.53
F-test, Foreign 5.39 5.41 5.19
+ Foreign
M&A = 0
p-Value 0.02 0.02 0.02
NOTES:Estimationsinclude Argentina,Chile, Colombia, Mexico, and Peru. Robust t-statistics (calculatedallowing for clustered standarderrorsby bank) are
***significant at 1%.

entered the system through acquisitions of domestic banks are 0.26% per quarter
lower than those for domestic banks. Since the Foreign De Novo dummy is also
negative, its sum with the Foreign Bank dummy is a large negative, implying that
while both types of foreign banks charge lower spreads than domestic banks, the
de novo foreign banks charge much lower spreads(around2.7% per quarterlower
than those for domestic banks). The interactionsbetween the mode of entry by
foreign banks and the time since entry (Age) are never significant.16
Two factorscould explain why the spreadschargedby foreign banksthat entered
the marketby acquiringdomestic banksmight differfrom those of de novo entrants.
First, de novo banks, interestedin gaining market share, may be more willing to
charge lower rates to reach their desired size. Second, the two types of foreign
banks may be targetingdifferentmarketsegments. Dell'Aricia and Marquez(2004)
suggest that differences in the informationavailable to banks influence who they
lend to and the spreads they are able to charge. By virtue of being newcomers to
the sector, de novo banks are likely to possess the least informationaboutdomestic
borrowersand, hence, would have an incentive to focus on the more transparent
segmentsof the market(i.e., where informationaboutborrowersis most accessible).
At the same time, since transparentmarketsegments are likely to be more competi-
tive, de novo banks would be requiredto charge lower spreads relative to those
possible in other marketsegments. In contrast,foreign banks acquiringor merging
with domestic bankswould inheritproprietarycustomerinformation,allowing them
to serve somewhat less transparentfirms, in less contestable markets,where they
might have some marketpower and the ability to chargehigher spreads.Since both
types of foreign banks charge lower spreadsthan their domestic competitors,it is
possible thatdomesticbanksare forced to increasetheirlendingto the least transpar-
ent borrowersfrom whom they are able to obtain the highest spreads.17
Beyond their incentive and ability to charge lower spreads, do foreign banks
have a spillover effect on the overall level of spreads?We test this possibility by
investigatingif the foreign bank participationvariable (i.e., the shareof loans held
by banks that are at least 50% foreign owned) influences spreads. In our basic
estimations in Table 3, this variable is statisticallyinsignificant.This result could
imply either that no spillover effect exists (lower spreadschargedby foreign banks
do notcreatesufficientpressureon otherbanksto lowertheirspreads,perhaps,because
of market segmentation)or that the spillover effect operates mainly in an indirect
manner;for example, throughthe impact of foreign competitionon administrative
costs, as we examine below. It is possible, of course, that because foreign bank
participationis rising over time, the variablepicks up mainly a time trend and does
not speak to the issue of "spillovers."We also explore this possibility below.
16. Excluding these interactionterms does not change our results.

17. Testingthese speculationswould requirespecific dataon the portfolioof the differentbanks,which
are not availableat the presenttime in the detail that is necessary.As a second best alternative,we tried
controlling for the share of loans to assets and the ratio of non-interestexpenses to assets to take into
account that some de novo banks might be investing in bonds and/or securities rather than lending.
However, these variables never proved to be significant and in some cases reduced our sample size.
Thus, these results are not reportedhere, but are available upon request.

Finally,for the sampleincludingall banks,higherbankconcentrationraisesspreads

significantly. Regardless of the measure of concentrationincluded, spreads rise
as a responseto increasesin bankconcentration.A one standarddeviationincreasein
concentrationresults in a 0.13-0.25 standarddeviation change in bank spreads.18
In the rest of Table 3, we present estimations for the determinantsof spreads
among domestic banks only (columns 4-6) and foreign banks only (columns 7-9).
We continueto find thatliquidityand administrativecosts have a positive impact on
bank spreads,with administrativecosts exercising the strongerinfluence,especially
among foreign banks. Withinthe sample of domestic banks, we also find that those
with higher market shares are able to charge lower spreads. This may point to
the presence of economies of scale among large domestic banks. Withinthe group
of foreign banks, the evidence for lower spreads charged by the new entrantsis
reaffirmed.Once again,changes in foreignbankparticipationdo not seem to directly
affect the overall level of spreadsfor domestic or foreign banks.Finally,as before, a
rise in bank concentrationleads to higher spreads,with the effect being particularly
high and significantfor domestic banks.
The spread estimations reportedin Table 3 make three assumptions.First, by
pooling observationsacross countries,we are forcing the coefficientsin the spread
equations(except for the constant)to be the same for all countries.Second, we are
also assuming that there are no structuralshifts (over time) in the relationbetween
bankspreadsandtheirdeterminants.Finally,we areignoringpossible commonshocks
or time trends.
Because of the short time series at our disposal, we are unable to run separate
regressionsfor each countryand formally test the pooling assumption.19However,
we conduct alternativeestimations to analyze the sensitivity of our results to this
assumption.In particular,we obtain results excluding Mexico and Colombia, the
countrieswith the shortesttime series andthe lowest levels of foreignbankparticipa-
tion (see Table4).20Reassuringly,the results are virtuallythe same for this smaller
sample. Also, to mitigate the concern that our findings are driven by Argentina,the
countrywith the longest time series, Table4 also reportsestimationsexcluding this
country.Again, results remainedlargely unchanged.
To test for structuralshifts in the relationbetween spreadsand their determinants
over time, we try two possibilities (see Table 5). First, we interact administrative
costs (the most consistentlysignificantvariableacross all spreadspecifications)with
a dummythatequals one for the period 1999 andbeyond. Second, to assess whether
18. In particular,a one standarddeviation change in the share of loans held by the top three banks
(top five banks) results in a 0.25 (0.13) standarddeviation change in bank spreads.At the same time,
a one standarddeviation change in the Herfindahlindex leads to a 0.20 standarddeviation rise in
bank spreads.
19. Argentinais the exception, given its large numberof banks.Results for Argentinayield the same
results and conclusions as those for the panel. These results are available upon request.
20. To save space, in Tables4 and 5, we only reportestimationsincludingthe loan shareof the three
largest banks as a measure of concentration.However, regressions using the top five bank share and
the Herfindahlindex produce virtuallythe same results. These are available upon request.

TABLE 4
PANEL ESTIMATIONS FOR BANK SPREADS EXCLUDING SOME COUNTRIES
Excluding Mexico and Colombia Excluding Argentina
Variables All banks Domestic banks Foreign banks All banks Domestic banks Foreign banks
Liquid Assets 0.031 0.0603 0.0056 0.021 0.042 0.018

(2.73)*** (4.89)*** (0.38) (2.67)*** (2.66)** (2.03)**
Non-Performing -0.0002 0.0012 0.0013 -0.011 0.005 -0.037
Loans (0.04) (0.28) (0.10) (0.59) (0.31) (1.75)*
Administrative 0.611 0.5251 0.713 0.815 0.922 0.854
Costs (5.71)*** (7.28)*** (3.67)*** (5.81)*** (4.10)*** (5.53)***
Foreign Bank -0.0047 -0.004
(3.26)*** (2.68)***
Bank Market -0.027 -0.034 0.0209 -0.005 -0.014 0.065
Share (2.65)*** (2.81)*** (0.85) (0.45) (1.10) (2.22)**
Equity 0.009 0.0245 0.0012 0.007 0.004 0.007
(1.35) (1.58) (0.21) (0.91) (0.21) (0.97)
Foreign M&A 0.0023 0.0004 0.002 -0.002
(2.60)*** (0.32) (1.99)** (1.15)
Foreign M&A x 0.0001 0.0012 -0.00042 0.001
Age (0.09) (1.02) (0.55) (0.67)
Foreign De Novo -0.0223 -0.0184
(2.15)** (2.28)**
Foreign De 0.0073 0.0069
Novo x Age (1.05) (1.24)
Other M&A 0.0006 0.001 -0.003 -0.001
(0.76) (1.46) (2.65)*** (0.85)
Other M&A x -0.0001 -0.0004 0.001 0.001
Age (0.19) (0.65) (0.58) (0.59)
Foreign Bank 0.0046 0.0087 -0.0034 -0.006 -0.007 -0.005
Share
(1.33) (1.71)* (0.88) (1.52) (1.15) (1.23)
Top-3 BankShare 0.0565 0.0648 0.0497 0.027 0.031 0.015
(4.16)*** (3.83)*** (2.39)** (2.25)** (2.26)** (0.82)
Inflation 0.1036 0.0572 0.1425 -0.02 -0.03 -0.003
(3.52)*** (1.45) (3.72)*** (0.52) (0.71) (0.04)
Real Growth of 0.000469 0.000805 -0.00125 0.002 0.003 -0.00009
Production (0.18) (0.21) (0.43) (0.65) (0.82) (0.02)
Real Market 0.0001 0.0002 0.0001 -0.00011 -0.0001 -0.0001
InterestRate (1.93)* (2.23)** (1.12) (1.45) (1.57) (0.79)
Argentina 0.0152 0.0206 0.0113
(3.12)*** (3.93)*** (1.73)*
Chile 0.0039 0.0075 0.0038 -0.002 0.002 -0.003
(1.30) (2.15)** (0.81) (0.49) (0.60) (0.53)
Colombia 0.02 0.027 0.012
(4.42)*** (5.37)*** (1.74)*
Mexico -0.014 -0.009 -0.018
(4.87)*** (3.65)*** (3.28)***
Constant -0.0211 -0.0332 -0.0173 0.003 -0.006 0.006
(2.54)** (3.23)*** (1.42) (0.42) (0.71) (0.48)
Observations 2188 1303 885 1342 695 647
R-squared 0.57 0.55 0.58 0.56 0.64 0.53
NOTES:The estimationsexcluding Colombia and Mexico include: Argentina,Chile, and Peru. Those excluding Argentinainclude: Chile,
Colombia,Mexico, and Peru. Robust t-statistics(calculatedallowing for clusteredstandarderrorsby bank) are in parentheses.*significant
at 10%;**significantat 5%; ***significantat 1%.
the impact of administrativecosts on bank spreads changed with the increase in

foreign bank presence, we interact administrativecosts with the foreign bank
share. All interactionterms are always insignificant and our main results do not
change.

TABLE 5
PANEL ESTIMATIONS FOR BANK SPREADS ALLOWING FOR STRUCTURAL SHIFTS
Interactingadministrativecosts Interactingadministrativecosts
with a dummy post 1999 with foreign bank share
Variables All banks Domestic banks Foreign banks All banks Domestic banks Foreign banks
Liquid Assets 0.0281 0.0522 0.0187 0.028 0.0514 0.0198

(3.74)*** (4.70)*** (2.32)** (3.79)*** (4.62)*** (2.55)**
Non- -0.0009 0.0005 -0.0002 -0.001 0.0004 0.0007
Performing (0.23) (0.12) (0.02) (0.25) (0.11) (0.06)
Loans
Administrative 0.6418 0.5485 0.7748 0.6502 0.4609 0.8745
Costs (6.24)*** (7.35)*** (3.98)*** (4.60)*** (3.98)*** (3.56)***
Administrative 0.0319 0.1051 -0.0783
Costs x
Dummy (0.39) (1.66)* (0.46)
1999-2000
Administrative 0.0099 0.4513 -0.5312
Costs x Foreign (0.03) (1.37) (0.87)
Bank Share
Foreign Bank -0.0052 -0.0051
(3.91)*** (3.95)***
Bank Market -0.0141 -0.0235 0.0469 -0.014 -0.0234 0.0451
Share (1.39) (1.99)** (1.79)* (1.38) (1.98)** (1.69)*
Equity 0.0095 0.019 0.0043 0.0095 0.0194 0.0037
(1.59) (1.36) (0.76) (1.58) (1.38) (0.65)
Foreign M&A 0.0025 -0.0003 0.0024 -0.0003
(2.52)** (0.22) (2.50)** (0.20)
Foreign 0.0001 0.0018 0.0002 0.0018
M&A x Age (0.11) (1.50) (0.22) (1.52)
Foreign De -0.0215 -0.0195 -0.0217 -0.0193
Novo (2.07)** (2.06)** (2.08)** (2.17)**
Foreign De 0.0073 0.0083 0.0075 0.0078
Novo x Age (1.05) (1.26) (1.07) (1.23)
OtherM&A 0.0002 0.0007 0.0002 0.0006
(0.28) (0.90) (0.25) (0.86)
OtherM&A x -0.000007 -0.00029 0.000015 -0.00029
Age (0.01) (0.42) (0.03) (0.42)
Foreign Bank Share 0.0009 -0.000027 -0.001 -0.0128 -0.009 -0.0159
(0.18) (0.00) (0.15) (4.45)*** (3.35)*** (2.97)***
Top-3 Bank 0.0391 0.0516 0.018 0.0382 0.0513 0.0175
Share (3.06)*** (3.37)*** (0.87) (3.01)*** (3.36)*** (0.90)
Inflation -0.0012 -0.0213 0.0053 -0.0024 -0.0267 0.0046
(0.03) (0.52) (0.09) (0.07) (0.66) (0.08)
Real Growthof 0.0017 0.0019 0.0002 0.0023 0.0028 -0.0006
Production (0.52) (0.46) -0.04 (0.77) (0.72) (0.16)
Real Market 0.000013 0.000064 -0.00003 0.000014 0.000068 -0.000029
InterestRate (0.22) (1.13) -0.24 (0.23) (1.19) (0.22)
Argentina 0.0078 0.0151 0.0009 0.0021 -0.003 0.0032
(1.70)* (3.26)*** (0.11) (0.39) (0.44) (0.42)
Chile 0.0002 0.0038 -0.0012 0.0074 0.0143 0.0013
(0.06) (1.21) (0.24) (1.75)* (3.15)*** (0.19)
Colombia 0.0207 0.0292 0.0118 0 0.0035 -0.0011
(4.60)*** (5.91)*** (1.61) 0.00 (1.12) (0.23)
Mexico -0.013 -0.0095 -0.0151 0.0204 0.0287 0.012
(4.75)*** (3.48)*** (2.96)*** (4.65)*** (5.92)*** (1.70)*
Constant -0.0067 -0.0195 0.0011 -0.0065 -0.0183 0.0002
(0.87) (2.20)** (0.09) (0.85) (2.10)** (0.02)
Observations 2618 1539 1079 2618 1539 1079
R-squared 0.52 0.53 0.48 0.52 0.53 0.49
NOTES: These estimationsinclude all countries:Argentina,Chile, Colombia,Mexico, and Peru. Robust t-statistics(calculatedallowing for
clustered standarderrorsby bank) are in parentheses.*significantat 10%;**significantat 5%; ***significantat 1%.

To control for possible time trends, we repeat our spreadestimations including

quarterlytime dummies(see Table6). Most of our resultsremainunchanged,except
that among domestic banks, the foreign bank share has a positive impact on
spreads. One possible explanationfor this result is that competition from foreign
banks causes domestic banks to redirecttheir lending to more opaqueborrowersto
whom they can charge higher spreads. However, this finding does not imply that
the net effect of foreign bank participationon domestic banks is to increase
their spreads, since as discussed above, the presence of foreign banks can affect
spreadsindirectlythroughits impact on administrativecosts. We turnto this issue
next.
Table 7 presentsthe determinantsof administrativecosts (expressedas a ratio of
total assets) for all banks and separately for domestic and foreign banks.21The
macro variables are now seen to be significant among domestic banks, unlike in
the spreadsequations.Inflationis negatively signed, suggesting that bank costs do
not respond immediately to general inflation. Higher interest rates, which are a
proxy for the marginalcost of capital, raise administrativecosts.
In general, foreign banks appearto operatewith lower costs relative to domestic
banks.However,in the estimationsincludingall banks,differenttypes of foreignbank
entry (via M&As or throughde novo entry)do not seem to have differentialeffects
on costs. On the other hand, in the specification for foreign banks, we find that
those thatenteredthroughmergerand acquisitionswith domestic bankshave higher
costs than other foreign banks.
Regardless of their origin, the overall level of foreign bank presence seems to
exert a downwardpressureon the administrativecosts of all banks. Thus, despite
evidence consistent with the hypothesis of market segmentation in our spreads
results, foreign bank presence apparentlygenerates sufficientcompetitive pressure
to induce an all-roundlowering of costs.
However, if foreign bank entry is also associated with increased concentration,
there may be an offsettingeffect. Our results indicatethat costs, indeed, go up with
concentration.Note that unlike for spreads, where the influence of concentration
was especially large for domestic banks, more concentrationseems to raise costs
all aroundin similar measure.This result is consistent with the notion that in more
concentratedsystems there is less pressurefor banks to lower their administrative
costs in orderto offer more competitive spreads.Since bankconcentrationwas also
seen to raisespreads,it has a particularlypowerfuleffecton the costs of intermediation.
As with the spreads estimations, we repeated the regressions dropping Argen-
tina, Mexico, and Colombia and adding time dummies. The results remain the
same. To save space, we do not reportthese resultshere, but they are availableupon
request.
21. Again, recognizing the possibility of reverse causation from costs to marketshare, we reestim-
ated the regressions with lagged values of the market share, with results that are the same as those
described here. However, these estimations are available upon request.

TABLE 6
PANEL ESTIMATIONS FOR BANK SPREADS INCLUDING QUARTERLY TIME DUMMIES (NOT SHOWN)
Variables (6.1) (6.2) (6.3) (6.4) (6.5) (6.6)
Liquid Assets 0.0272 0.0275 0.027 0.0525 0.0531 0.0523

(3.59)*** (3.64)*** (3.56)*** (4.65)*** (4.69)*** (4.63)***
Non-Performing -0.0005 -0.0005 -0.0004 0.0012 0.0012 0.0013
Loans (0.12) (0.13) (0.09) (0.30) (0.29) (0.33)
Administrative 0.6424 0.6446 0.6424 0.5498 0.5506 0.5488
Costs (7.28)*** (7.29)*** (7.27)*** (7.19)*** (7.20)*** (7.18)***
Foreign Bank -0.0051 -0.0051 -0.0051
(3.90)*** (3.91)*** (3.90)***
Bank MarketShare -0.0141 -0.0141 -0.0143 -0.0239 -0.0241 -0.0242
(1.40) (1.39) (1.42) (1.99)** (2.00)** (2.01)**
Equity 0.0095 0.0096 0.0095 0.0185 0.0186 0.0185
(1.57) (1.59) (1.58) (1.29) (1.30) (1.29)
Foreign M&A 0.0023 0.0023 0.0024
(2.50)** (2.50)** (2.57)**
Foreign M&A x 0.0002 0.0003 0.0002
Age (0.26) (0.34) (0.21)
Foreign De Novo -0.0211 -0.0214 -0.0213
(2.04)** (2.05)** (2.06)**
Foreign De Novo x 0.0071 0.0075 0.0072
Age (1.00) (1.05) (1.01)
Other M&A 0.0004 0.0003 0.0004 0.0007 0.0007 0.0008
(0.45) (0.41) (0.47) (1.02) (0.98) (1.05)
Other M&A x Age -0.00002 0.000018 -0.000051 -0.000251 -0.000227 -0.000293
(0.04) (0.03) (0.09) (0.36) (0.33) (0.42)
Foreign Bank Share 0.0099 0.0109 0.0077 0.0201 0.0204 0.0166
(1.51) (1.70)* (1.23) (2.64)*** (2.72)*** (2.30)**

TABLE 6
CONTINUED
Variables (6.1) (6.2) (6.3) (6.4) (6.5) (6.6)

(3.18)*** (3.46)***
(2.09)** (3.03)***
Herfindahlindex 0.000011 0.000015
(2.99)*** (3.49)***
Inflation -0.0048 0.0022 -0.0078 -0.0489 -0.0425 -0.054
(0.12) (0.05) (0.19) (1.04) (0.90) (1.14)
Real Growth of -0.0056 -0.0057 -0.0046 -0.0057 -0.0058 -0.0046
Production (1.71)* (1.70)* (1.37) (1.34) (1.33) (1.06)
Real Market 0.000003 -0.00002 0.000019 -0.000003 -0.000019 0.000023
InterestRate (0.04) (0.28) (0.27) (0.04) (0.26) (0.30)
Argentina 0.0066 0.0015 0.006 0.0112 0.0067 0.0111
(1.70)* (0.41) (1.55) (2.60)** (1.81)* (2.65)***
Chile -0.0004 -0.0048 -0.002 0.002 -0.0027 0.0004
(0.15) (2.17)** (0.80) (0.67) (1.08) (0.14)
Colombia 0.0199 0.0146 0.0177 0.0288 0.0236 0.0265
(4.53)*** (3.70)*** (4.44)*** (6.06)*** (5.69)*** (6.27)***
Mexico -0.0108 -0.012 -0.0111 -0.0047 -0.0057 -0.005
(3.10)*** (3.40)*** (3.18)*** (1.52) (1.88)* (1.64)
Constant -0.0091 -0.0012 -0.0028 -0.0265 -0.0207 -0.0191
(0.88) (0.12) (0.30) (2.51)** (2.03)** (2.09)**
Observations 2618 2618 2618 1539 1539 1539 1
R-squared 0.53 0.53 0.53 0.54 0.54 0.54
NOTES:These estimations include all countries: Argentina,Chile, Colombia, Mexico, and Peru. Quarterlytime dummies are included in all regressions, but
allowing for clustered standarderrorsby bank) are in parentheses.*significantat 10%;**significantat 5%; ***significantat 1%.

TABLE 7
PANEL ESTIMATIONS FOR ADMINISTRATIVE COSTS INCLUDING ALL COUNTRIES
Variables (7.1) (7.2) (7.3) (7.4) (7.5) (7.6)
Foreign Bank -0.0029 -0.0029 -0.0029

(2.57)** (2.57)** (2.57)**
Bank Market Share -0.0546 -0.0548 -0.0549 -0.0321 -0.0322 -0.0323
(4.57)*** (4.58)*** (4.58)*** (3.12)*** (3.13)*** (3.14)***
Foreign M&A 0.0017 0.0017 0.0018
(1.52) (1.50) (1.56)
Foreign M&A x Age 0.000397 0.000353 0.000316
(0.16) (0.14) (0.13)
Foreign De Novo 0.0116 0.0115 0.0115
(0.70) (0.70) (0.70)
Foreign De Novo x Age 0.0035 0.0035 0.0035
(0.39) (0.39) (0.39)
Other M&A 0.0015 0.0015 0.0015 0.0011 0.001 0.0011
(2.83)*** (2.79)*** (2.87)*** (2.06)** (2.03)** (2.12)**
OtherM&A x Age -0.001 -0.001 -0.001 -0.0007 -0.0007 -0.0007

(2.24)** (2.24)** (2.28)** (1.41) (1.40) (1.48)
Foreign Bank Share -0.0165 -0.0183 -0.0185 -0.0209 -0.0226 -0.0233
(4.97)*** (5.51)*** (5.67)*** (4.45)*** (4.80)*** (5.01)***
(5.75)*** (4.10)***
(4.79)*** (3.18)***
Herfindahlindex 0.000014 0.000014
(5.72)*** (4.56)***
Inflation -0.026 -0.0185 -0.0285 -0.0698 -0.0623 -0.0748
(1.57) (1.12) (1.69)* (3.47)*** (3.18)*** (3.59)***

TABLE 7
CONTINUED
Variables (7.1) (7.2) (7.3) (7.4) (7.5) (7.6)
Real Growthof Production -0.0106 -0.0109 -0.0098 -0.0173 -0.0178 -0.0164

(3.99)*** (4.05)*** (3.77)*** (4.46)*** (4.48)*** (4.29)***
Real MarketInterestRate 0.000113 0.0001 0.000135 0.000167 0.000154 0.000193
(3.16)*** (2.85)*** (3.59)*** (3.55)*** (3.37)*** (3.97)***
Argentina 0.015 0.011 0.015 0.019 0.0151 0.0204
(4.78)*** (3.79)*** (4.87)*** (6.29)*** (5.42)*** (6.83)***
Chile -0.0075 -0.0119 -0.0089 -0.0056 -0.0097 -0.0061
(3.09)*** (5.62)*** (3.95)*** (2.76)*** (6.22)*** (3.41)***
Colombia 0.003635 -0.00146 0.001709 0.004937 0.000041 0.004434
(1.37) (0.64) (0.72) (1.88)* (0.02) (1.94)*
Mexico 0.0026 0.001238 0.002388 -0.000028 -0.00139 0.000251
(0.85) (0.40) (0.82) (0.01) (0.74) (0.13)
Constant -0.0025 0.0039 0.0043 -0.0033 0.0031 0.0008
(0.49) (0.81) (1.03) (0.56) (0.55) (0.18)
Observations 2982 2982 2982 1752 1752 1752
R-squared 0.36 0.36 0.37 0.40 0.40 0.40
F-test, Foreign + 0.86 0.85 0.74
Foreign M&A = 0
p-Value 0.36 0.36 0.39
NOTES:These estimations include observationsfor Argentina,Chile, Colombia, Mexico, and Peru. Robust t-statistics (calculated allowing for clustered stand
**significantat 5%; ***significantat 1%.

4. CONCLUSIONS
This paper studies the impact of foreign participationand concentrationon bank

spreads.Ourresultsshow thatthese factorsinfluencethe spreadschargedto borrow-
ers, and hence the process of financial intermediation,in a complex manner.The
overall effect depends on three differentchannels of influence:the spreadscharged
by foreign banks relative to domestic banks, the spillover effects from the presence
of foreign bankson both spreadschargedand on operatingcosts, and the concentra-
tion in the bankingsectorthathas accompaniedforeign entry.Considereach in turn.
First, foreign banks chargelower interestmarginsand potentiallyfoster financial
intermediation.New establishments(i.e., de novo banks) operate with particularly
low spreads.However, whethersuch entry generateswelfare gains is unclear.This
will depend on whether the lower spreadschargedare the consequence of a more
aggressivepricingstrategyor because de novo bankschoose to lend only to the most
transparentsegments with high marketcontestability.
Second, greaterforeign presencedoes not imply a generaldecline in spreads,but
appearsto influence the intermediationthroughlowering costs of operation.More
widespread foreign bank presence is associated with cost reduction throughout
the banking system. Possibly, a combinationof demonstrationeffects and potential
competition, with banks threateningto encroach on each others' customer base,
generatesthe pressuresfor cost reductionthatultimatelybenefit bank clients. Thus,
the long-termbenefitsof foreignentryarelikely to come fromlower cost structuresin
the banking system.
Third,greaterconcentrationraises spreadsin an economically importantmanner.
This is so especially for domestic banks. At the same time, concentrationis also
associated with higher administrativecosts all around.The implicationis that part
of the benefits from foreign entry may be offset where concentrationlevels also
increase.As notedin the introduction,the consolidationthatdid occurin the banking
sectors of the countries concerned was not necessarily related to foreign entry,
althoughthe fact that much of the entry was in the form of takeovers,ratherthan
new establishments,did not help create more competition.For policymakers,this
creates a challenge since more competition is desirable for lowering spreads, but
could generatevulnerabilitywhere the "franchise"value of domestic banks is seri-
ously eroded.
Finally, while we believe this paper adds to our understandingof the impact of
foreign participationand concentrationon the costs of financial intermediationin
developing countries, more work in this area is clearly needed. Given the limited
numberof countriesand short sample period we study,thereis a need to extend the
analysis in both these directions.Also, furtherresearchlinking bank-leveldata with
the banks' customerprofiles would help to explain the apparentdifferencesin the
spreadschargedby foreign and domesticbanks,somethingthatthis paperspeculates
on but cannot answer definitively.

LITERATURE CIT'ED
Allen, Linda (1988). "The Determinantsof Bank Interest Margins: A Note." Journal of
Financial and QuantitativeAnalysis 23, 231-235.
Angbazo, Lazarus(1997). "CommercialBank Net InterestMargins,Default Risk, Interest-
Rate Risk, and Off-BalanceSheet Banking."Journal of Bankingand Finance 21, 55-87.
Barajas,Adolfo, RobertoSteiner,and Natalia Salazar(2000). "The Impactof Liberalization
andForeignInvestmentin Colombia'sFinancialSector."Journalof DevelopmentEconom-
ics 63, 157-196.
Berger, Allen N. (1995). "The Profit-StructureRelationshipin Banking-Tests of Market
Power and Efficient StructureHypotheses."Journal of Money, Credit, and Banking 27,
404-456.
Bernanke,Ben, and MarkGertler(1989). "Agency Costs, Net Worth,and Business Fluctua-
tions." AmericanEconomic Review 79, 14-31.
Brock, Philip L., and Liliana Rojas Suarez (2000). "Understandingthe Behavior of Bank
Spreadsin Latin America."Journal of DevelopmentEconomics 63, 113-134.
Claessens, Stijn, Asli Demirgiiu-Kunt,and Harry Huizinga (2000). "The Role of Foreign
Banks in Domestic Banking Systems." In The Internationalizationof Financial Services:
Issues and Lessonsfor Developing Countries,editedby StijnClaessens andMarionJansen,
pp. 117-138. Boston: Kluwer Academic.
Cottarelli,Carlo, and Angeliki Kourelis (1994). "FinancialStructure,Bank Lending Rates,
and the TransmissionMechanismof MonetaryPolicy."InternationalMonetaryFund Staff
Papers 41, 587-623.
Dell'Arricia, Giovanni, and RobertMarquez(2004). "Informationand Bank Credit Alloca-
tion." Journal of Financial Economics 72, 185-214.
Demirguiq-Kunt, Asli, Luc Laeven, and Ross Levine (2004). "Regulations,MarketStructure,
Institutions,and the Cost of Financial Intermediation."Journal of Money, Credit, and
Banking 36, 593-622. (this issue of JMCB).
Denizer, Cevdet (2000). "ForeignEntry in Turkey's Banking Sector, 1980-1997." In The
Internationalizationof Financial Services: Issues and Lessonsfor Developing Countries,
edited by Stijn Claessens and Marion Jansen, pp. 389-406. Boston: Kluwer Academic.
Gelos, Gast6n, and Jorge Rold6s (2002). "Consolidationand MarketStructurein Emerging
MarketSystems." IMF WorkingPaper No. 02/186.
Ho, Thomas, and Anthony Saunders(1981). "The Determinantsof Bank InterestMargins:
Theory and Empirical Evidence." Journal of Financial and QuantitativeAnalysis 4,
581-600.
IMF (InternationalMonetaryFund) (2000). InternationalCapital Markets: Developments,
Prospects, and Key Policy Issues. Washington,DC.
Kiyotaki, Nobuhiro, and John Moore (1997). "CreditChains."Journal of Political Econ-
omy 105, 211-248.
Levine, Ross (1996). "ForeignBanks, Financial Development, and Economic Growth."In
InternationalFinancial Markets:Harmonizationversus Competition,edited by C.E. Bar-
field, pp. 224-254. Washington,DC: AmericanEnterpriseInstitutePress.
Rogers,WilliamH. (1993). "RegressionStandardErrorsin ClusteredSamples."Stata Techni-
cal Bulletin Reprints3, 88-94.
Saunders, Anthony, and Liliana Schumacher(2000). "The Determinantsof Bank Interest
Rate Margins:An InternationalStudy."Journal of InternationalMoney and Finance 19,
813-832.
Smith, R. Todd (2001). "Bank Spreads and Business Cycles." Mimeo, IMF.

Wong, Kit Pong (1997). "On the Determinantsof Bank InterestMargins under Credit and
InterestRate Risks." Journal of Banking and Finance 21, 251-271.
Zarruck,Emilio R. (1989). "BankMarginwith UncertainDeposit Level and Risk Aversion."
Journal of Banking and Finance 14, 803-820.


A4 Martinez Mody (2004)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

A4 Martinez Mody (2004)

Uploaded by

Copyright:

Available Formats

How Foreign Participation and Market Concentration Impact Bank Spreads: Evidence from

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

How Foreign Participationand Market

Increasingforeign participationand high concentrationlevels characterize

JEL codes: D4, G21, Lll, N26

The marketstructureof the bankingindustryin many devel-

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

1. Figures obtainedfrom Demirgiiq-Kunt,Laeven, and Levine (2004).

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

spreads. Demirgtiu-Kunt,Laeven, and Levine (2004) is an exception. This paper

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

a numberof reasons.First,despite having embarkedin a process of financialmarket

As in many developing economies, countriesin LatinAmericaexperienceda sig-

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

Country Variables 1995 1996 1997 1998 1999 2000

Argentina Total numberof 141 122 115 106 96 90

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

Country Variables 1995 1996 1997 1998 1999 2000

Mexico Total numberof banks 39 39 40

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

In this section, we turnto an econometricanalysis of the impact of concentration

+ o18Argentinai, + al9Chileij,t+ a2oColombiai,t

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

refer to cash and deposit balancesin otherbanks (includingreserve requirementsat

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

in the system. Banking Sector Concentrationmeasuresthe extent to which loans are

10. An extensive literatureexists studyingthe impactof concentrationon bankprofitability(see Berger

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

Table3 presentsthe estimationresultsfor Equation(1), analyzingthe determinants

11. Note that while Table 1 reportsannualizedspreads,Table 2 presentsquarterlyspreads,since the

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

Spread Interestincome received Bank superintendencies 0.025 0.016

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

All banks Domestic banks

Variables (3.1) (3.2) (3.3) (3.4) (3.5) (3.6)

Liquid Assets 0.028 0.029 0.028 0.053 0.054 0.052

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

All banks Domestic banks

Variables (3.1) (3.2) (3.3) (3.4) (3.5) (3.6)

Foreign Bank Share 0.002 0.002 0.001 0.005 0.003 0.003

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

16. Excluding these interactionterms does not change our results.

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

Finally,for the sampleincludingall banks,higherbankconcentrationraisesspreads

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

Excluding Mexico and Colombia Excluding Argentina

Liquid Assets 0.031 0.0603 0.0056 0.021 0.042 0.018

the impact of administrativecosts on bank spreads changed with the increase in

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

Liquid Assets 0.0281 0.0522 0.0187 0.028 0.0514 0.0198

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

To control for possible time trends, we repeat our spreadestimations including

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

All banks Domestic banks

Variables (6.1) (6.2) (6.3) (6.4) (6.5) (6.6)

Liquid Assets 0.0272 0.0275 0.027 0.0525 0.0531 0.0523

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM

All banks Domestic banks

Variables (6.1) (6.2) (6.3) (6.4) (6.5) (6.6)

Top-3 Bank Share 0.0386 0.0489

This content downloaded from 82.16.238.55 on Fri, 11 Jul 2014 04:30:44 AM