You are on page 1of 31

Determinants of Commercial Bank Interest Margins and Profitability: Some International Evidence Author(s): Ash Demirg-Kunt and Harry

Huizinga Source: The World Bank Economic Review, Vol. 13, No. 2 (May, 1999), pp. 379-408 Published by: Oxford University Press Stable URL: . Accessed: 02/07/2013 00:32
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .

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

Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to The World Bank Economic Review.

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions











Determinantsof CommercialBank InterestMargins and Profitability:Some InternationalEvidence

Asli Demirgui-Kunt and Harry Huizinga
Using bank-leveldata for 80 countriesin the years 1988-95, this articleshows that differences in interestmarginsand bank profitability reflecta varietyof determinants: bank characteristics, macroeconomic conditions,explicitand implicitbank taxation, and underlying depositinsuranceregulation,overallfinancialstructure, legaland institutionalindicators.A largerratioof bankassetsto grossdomesticproductand a lower andprofits,controlling marketconcentration ratioleadto lower margins for differences in bank activity, leverage,and the macroeconomic environment. Foreignbanks have highermarginsand profitsthan domesticbanksin developingcountries,while the opcountries. tax burden posite holds in industrial Also, thereis evidencethat the corporate is fully passed onto bank customers,while higherreserverequirements are not, especially in developingcountries.

As financial intermediaries,banks play a crucial role in the operation of most economies. Recent research, as surveyed by Levine (1997), shows that the efficacy of financial intermediationcan affect economic growth. Crucially,financial intermediation affects the net return to savings and the gross return to investment. The spread between these two returns mirrors bank interest margins, in addition to transaction costs and taxes borne directly by savers and investors. Thus bank interest spreads could be interpretedas an indicator of the efficiency of the banking system. In this articlewe investigatehow bank interestspreadsare affected by taxation, the structure of the financial system, and financial regulations, such as deposit insurance. A comprehensive review of the determinantsof interest spreads is offered by Hanson and Rocha (1986), who summarize the role that implicit and explicit taxes play in raising spreads and discuss some of the determinantsof bank costs and profits, such as inflation, scale economies, and market structure.Using aggregate interest data for 29 countries in the years 1975-83, the authors find a positive correlation between interest margins and inflation. Recently, several studies have examined the impact of international differences in bank regulation using cross-countrydata. Analyzing interest rates in 13 Organisation for Economic Co-operation and Development (OECD) countries in
AsliDemirgui-Kunt is with the Development Research Groupat the WorldBank,andHarryHuizinga is with the Department andthe Centrefor Economic of Economics, TilburgUniversity, PolicyResearch in London.The authorsgratefully acknowledgecommentsby threeanonymousreferees.They also thank AnqingShi for excellentresearch assistance and PaulinaSintim-Aboagye for help with the manuscript. Bankfor Reconstruction and Development C 1999 The International / THEWORLDBANK


This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



the years 1985-90, Bartholdy,Boyle, and Stover (1997) find that the existence of explicit deposit insurance lowers the deposit interest rate by 25 basis points. Barth, Nolle, and Rice (1997) use 1993 data from 19 industrial countries to further examine the impact of banking powers on bank return on equity, controlling for several bank and market characteristics.They find that variations in banking powers, bank concentration, and the existence of explicit deposit insurance do not significantly affect the return on bank equity. In this article we extend the existing literature in several ways. First, we use bank-level data for 80 industrialand developing countries in 198 8-95 to provide summary statistics on the size and decomposition of bank interest margins and profitability. Second, we use regressionanalysis to examine the underlyingdeterminants of interest spreads and bank profitability. This empirical work enables us to infer the extent of taxation and regulation on bank customers and on banks themselves. Apart from covering many banks in many countries, this study is unique in its coverage of the determinantsof interestmarginsand profitability.These determinants include a comprehensiveset of bank characteristics(such as size, leverage, type of business,foreignor domestic ownership),macroeconomicindicators,taxation and regulatoryvariables,financial structurevariables, and legal and institutional indexes. Among these, the ownershipvariable,the tax variables,some of the financial structure variables, and the legal and institutional indicators have not been included in any previous study in this area. To check whether some of these determinantsaffect bankingdifferentlyin developingand industrialcountries,we interactthese variableswith the country'sgross domesticproduct (GDP)per capita. I. BANK INTERESTSPREADSAND PROFITABILITY We can measure the efficiency of bank intermediationusing both ex ante and ex post spreads. The ex ante spread is the difference between the contractual rates charged on loans and rates paid on deposits. The ex post spread is the difference between banks' actual interest revenues and their actual interest expenses. The ex post spread differs from the ex ante spread by the amount of loan defaults. The ex post spread is a more useful measure because it controls for the fact that banks with high-yield, risky credits are likely to face more defaults. An additional problem with using the ex ante spread is that data are generally available at the aggregateindustrylevel and are put together from a variety of sources. Thus they are not completely consistent. For these reasons we focus on ex post interest spreads in this article.There is, however, a problem with ex post spreads, in that the interest income and loan loss reserving associated with a particular loan tend to materialize in different time periods. As a measure of what we call bank "efficiency,"we consider the accounting value of a bank's net interest income divided by total assets (TA), or the net interest margin (NIM). Bank "profitability"is a bank's before-tax profits (BTP) divided by total assets. Profitability could also be measured by the return on

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgi4-Kunt and Huizinga 381

equityas opposedto thereturn on assets.It is well knownthat,ceteris paribus, a bankwith a higherequityratiowill have a higherreturnon assetsand a lower returnon equitythan a bankwith a lower equityratio. The problemin some developingcountriesis that banksoperatewith extremelylow equitycapital, which inflatestheirreturnon eqoften supported by implicitstate guarantees, uity. Using unadjusted returnson equitymay then be moredistortionary than returns on equity, using returnson assets.Ideally,we shoulduse risk-adjusted but since theseare not available, on assetsaftercontrolling we analyzereturns the for the banks'equityratio.We do this by entering equityratioas an independentvariablein the profitregression. Thus,by straightforward accounting,

BTA ATP TX + ~ ~~~~ TA TA TA

whereATP is after-taxprofits.Fromthe bank'sincomestatement, before-tax profitsdividedby total assetsfurther satisfiesthe followingaccounting identity:









whereNII is noninterest income,OV is overhead,and LLP is loan loss provisioning.NIIITAreflects the fact that manybanksalso engagein nonlending acand brokerage tivities,suchas investment banking accounts for services, OVITA the bank'sentireoverhead with all of its activities, andLLP/FA meaassociated suresactualprovisioning for baddebts. Althoughthe net interest can be interpreted as a roughindexof bank margin efficiency,this does not meanthat its reductionalwayssignalsimproved efficiency.A reduction in the net interestmargincan, for example,reflecta reduction in bank taxationor, alternatively, a higherloan defaultrate. In the first instancethe reduction in the net interestmarginmayreflectthe improved functioningof the banking system, whilein the secondcasethe oppositemaybe true. Also,variation in an accounting ratio,suchas the netinterest margin, mayreflect or in netinterest income(thenumerator) differences differences in,say,nonlending assets(a component of the denominator). In the dataset the accounting so as to be comparable data'areorganized interin accounting nationally. However,differences the valuaconventions regarding tion of assets, loan loss provisioning, hiddenreserves,and so on may remain. Vittas(1991) reviewsthepitfallsin interpreting bankoperating ratios.Accountwith a longlagso thattheyarenot ing dataalso tendto reflect economicrealities able to flag pendingbanking crises,suchas thosethathaverecently occurred in Southeast Asia. This articlefocuseson accountingmeasuresof incomeand profitability as investors equalize(risk-adjusted) financial returns on bankstocksin the absence of prohibitive barriers. Gortonand Rosen(1995) andSchranz (1993)also focus

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



of profitability when examining on accountingmeasures entrenchmanagerial mentand banktakeovers. a usefuldecomposition The aboveaccounting identity(equation of 2) suggests the realizedinterestspread-the net interestmargin-into its constituent parts: noninterestincome, overhead,taxes, loan loss provisions,and after-taxbank withsomemodifications. profits.HansonandRocha(1986)takethis approach, As a first step to analyzingthe data, in section III we providean accounting countriesand for selected breakdownof the net interestmarginfor individual it maybe misleading to compare ratioswithout aggregates. Although accounting in the macroeconomic controllingfor differences in which banks environment in their business,productmix, and leverage,these operateand the differences breakdowns are still a usefulinitialindicatorof differences acrosscountries. Next, controllingfor bank characteristics and the macroeconomic environment, we providean economicanalysisof the determinants of the interestand profitability variables-the net interestmargin and before-tax profitsdividedby total assets.This empirical work offersinsightsinto how bank customers and banksthemselves The net interest areaffectedby thesevariables. marginregressions tell us how the spreaddeterminants affectthe combined welfareof depositors and lenders.The relationship betweenthe interestspreadand a bank'scorporatetaxes, for instance,revealsthe extentto whicha bankis ableto shift its tax bill forwardto its depositors and lenders. taxes and other variablescan affect interestrates as well as the Generally, volume of loans and deposits.In the short term the majoreffectsmay come throughpricingchanges,in which case the net interestmarginand before-tax profits as a share of total assetsimmediately revealeasily interpreted welfare consequences for banksand theircustomers. Withmarketimperfections in the form of creditrationingor imperfect in creditmarkets, competition changesin quantities generally havefirst-order welfareimplications independent of changes in prices.Wedo not, however, evaluate changes in quantities in thisarticle. Lastly, the before-taxprofit regressions show how spreaddeterminants affect bank shareholders. The regression analysisstartsfromthe followingequation:



+ a, Bij, + Xi,

S +yT?C

+ Ei+t

where Ijitis the dependent variable(eitherthe NIM or BTP/TA)for bank i in countryj at time t, Bij, arecharacteristics of banki in countryj at time t, X,tare / characteristics of country at time t, T, and C, are time and countrydummy variables,and c,,t is a white-noise errorterm.We estimateseveralspecifications of equation2 including different bankandcountryvariables. II.

In this studywe use incomestatement and balancesheetdata of commercial banksfromthe BankScope database provided by IBCA. IBCA's coverage is compre-

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgi4-Kunt and Huizinga 383

hensivein mostcountries, accounting for 90 percent of all bankassets.We begin with all commercial banksworldwide, with the exceptionof France, Germany, andtheUnited States, forwhichwe include onlyseveral hundred commercial banks listedas "large." To ensurereasonable for individual coverage we incountries, cludeonly countries with at leastthreebanksfor a givenyear.We end up with a dataset that includes80 countries duringthe years1988-95, with about7,900 individual commercial bankobservations. This data set includes all OECDcountries,as well as manydeveloping andtransition countries economies (table1). Severalcountries,such as Luxembourg, the Netherlands, and Egypt,have a netinterest margin closeto 1 percent (column 2 of table1). Thisis thelow endof the distribution. Egypt's low net interest can be explained margin by a predominanceof low-interest directed creditsby the largestate banking sector.Generally, developing countries, andespecially LatinAmerican suchas Arcountries, gentina,Brazil,CostaRica,Ecuador, andJamaica, haverelatively largespreads. This is also truefor certainEastern European countries, such as Lithuania and Romania. Columns3-6 in table1 breakdownthe net interest incomeintoits fourcomponents:overhead minusnoninterest income,taxes,loan loss provisioning, and net profits.Taxesas a shareof net interest income(column the explicit 4) reflect taxes that bankspay (mostlycorporate incometaxes). Banksalso face implicit taxationbecauseof reserve andliquidityrequirements andotherrestrictions on lendingthat come throughdirected or subsidized creditpolicies.Theseindirect formsof taxationdirectly lowerthe net interest incomerather thanthe tax variable. Nonetheless,the tax variableindicatesthat thereis considerable internationalvariation in theexplicittaxationof commercial banks.Several in countries Eastern Europeimposehighexplicittaxes on banking. (Forexample,taxesas a of net interest incomeareonly 17.5 in Lithuania percentage and 13.7 in Hungary compared with 26.2 in Romania,83.3 in Russia,and 23.2 in the CzechRepublic.)The lowestshareof taxesin net interest incomeis 0 for Qatar,wherethereis no significant taxationof banking.For some countries,such as Norway,Sweden, and Costa Rica, low tax sharesreflectthe tax deductibility of bad debts, whichare plentiful. Loanloss provisioning as a shareof net interest incomeis a directmeasure of differences in creditqualityacrosscountries(column5). It also reflectsdifferencesin provisioning regulations. This variable is high for some Eastern Europeancountries.It is also highfor some industrial suchas France and countries, the Nordiccountries. The fourthcomponent of net interest incomeis net profits (column6). As a residual, net profitsas a shareof net interest incomereflectthe extentto whichthe net interest translates margin into net-of-tax profitability. Theremaining columnsin table1 tabulate thevariousaccounting ratios(relativeto totalassets)in theaccounting identity(equation 2). Noninterest incomeas a shareof total assetsrevealsthe importance of fee-based for banksin services different countries(column7). Banksin Eastern Europe-for example,thosein Estonia,Hungary,and Russia-seem to rely heavilyon fee-basedoperations.

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

0 U^ (B of o0

'. - w

N b 00 e rf



o - eb

.(f w


O ,,*






. .


oo v- N VI
. . . . .

_* v1 N t^>w Q \ o Nb
. . . . . . . . .

v. . . .


. . . . . . .






00 k0 r-

V-- 0 V" V t', N

V Go *\

en N co r

0 \1
v -4

oq N t

ON o \0 \0 *\

oh " co r


V tl- 0 %nch _ >o >-



*tw oo o> t?


" r-

2 .

I t

et \0 o

*t 00 " 00
VN -4 V"-

00 \0 C)
V |^>

ws \aI R cro

u X e e >t> _ eN

-4 r-

<~~~~~t o6 \Z I46

4 -4 ci r4 1, 6

v: v-4 4

c t-i



I.. * *__H_ w>_

~~~~ .t



N N t" .

fE 1 en

. 0





N e







~~ ~


t (U X

~ ~


w-oow> o-> >v)o te-oo 4 t > > b ci



0 I



t E~~~~~


t .E~~~~~~38


This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



; 4

c; 6




(a' ". 6



C4 r

C4 c0 C; C

<6 CS

e4 c; cS

c) cI c;

(6 Cs c;

V: o4 4 <6



C4 C4 C



4 C4 C. i

Ci 4



4 \z


C4 c- C- c-


X~~~~C c; V OQ 14

X to C) 6 14 C4 X4 c" 14

0~) %.P -X

1: 14 1.

b CN-i O




O~ V

O, O" fn N

eq V




en t




6 ~~60~

0 a N ON 00 - \Den ON oo 00 V r r0~~~~~~~~~~~~~ f 0 _ao U
'-4 '4

ON oN ff oN N > N -. N ON"e r 00 v \ N 00 00 N

b tON ff N NV) v > O 0

N N 0 IN n \0 N% > o

t O0 - cn \0 m e ON No N tn N 0 0


No 00 _ t C, en >o I Ma0 " o m W O



o ON m O

o \o 00t tn

cr 00 NON-4 ' ' -


o.t N N

0o -0 - ~N~'-0 -

N N No oo N N

u '-4

o N



0 oo 0o N -4 NN '>-~


C No ' '-e




NO e



b~~~~~~~~~~~~~~W to>X NO v- e 00 N - e t N
0oeo e E

t -I NO N w


t w->oNoso


:6 - w






= r ra ^
=0 =



ffi 4)X -w.tE

z385z z z z o


This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions


4 C; 6 C;



6 Cs 1. C; C-i

bo 2)


_) o


co coo 04~~~~~~~~~~~~~ os



~~~~~~~ ~~~~.0
eXo c >o c \0 co m N 0 0


ti lZ f_ i 1. \_ .4

s _~ 0, Or C_4rx eiV:E


~ ~ ~ ~ ~ ~ ~

14 U,~~~~~~0 NIoo'q~~~u~r. DfWt4 ; E

E]i; 4



^ CD 1w %n CD cr0 N titOw v

- 0w co 0to 0


e w%


t 8t~~~~~~~~~~~~~~~~~~I

a. | ~

^ 2 8 sre t- N ON\

t.. 4,

I o .



This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgii-Kunt and Huizinga 387

This is also the case in some Latin Americancountries, such as Argentina,Brazil, Colombia, and Peru, and in a few African countries, such as Nigeria and Zambia. Overhead as a share of total assets reveals variations in operatingcosts across banking systems (column 8). This variablereflects variations in employment and in wage levels. Despite high wages, overhead as a share of total assets appearsto be lowest at around 1 percentfor high-incomecountries, such as Japan and Luxembourg. It is notably high at 3.6 percent for the United States, perhaps reflecting the proliferationof banks and bank branchesbecause of bankingrestrictions. Jamaica, Lithuania, and Romania stand out with high tax-to-asset ratios of around 2 percent (column 9). Loan loss provisioning,proxied by loan loss provisioning as a share of total assets, is equally high in EasternEurope and in some developing countries (column 10). Finally, net profits divided by total assets also tend to be relativelyhigh in developing countries (column 11). Table 2 presents statistics on interest spreads and profitability for selected aggregates. The first breakdown is by ownership; a bank is said to be foreignowned if 50 percentor more of its shares are owned by foreign residents.There is only a small differencein the net interestmarginfor domestic banks (3.7 percent) and foreign banks (2.9 percent). This small difference, however, masks the fact that foreign banks tend to achieve higher interest margins in developing countries and lower interestmarginsin industrialcountries.' This may reflectthe fact that foreign banks are less subject to credit allocation rules and have technical advantages (in developing countries), but also have distinct informationaldisadvantages relative to domestic banks (everywhere). Foreign banks pay somewhat lower taxes than domestic banks (column 4). This gap may reflect differencesin the tax rules governing domestic and foreign banks, as well as the ability of foreign banks to shift profits internationally to minimize their global tax bill. Foreign banks also have relativelylow provisioning, as indicated by loan loss provisioning as a share of total assets, which is consistent with the view that foreign banks generally do not engage in retail banking operations. The next breakdown is by bank size. For countries with at least 20 banks, large banks are defined as the 10 largest banks according to the value of their assets. Largebankstend to have lower marginsand profits and smalleroverheads. They also pay relativelylow direct taxes and have lower loan loss provisioning. Table 2 also considers bank groupings by national income levels and location.2 Breakingdown the data into four income levels, we see that the net interest margin is highest for countries in the middle-incomegroups. Banks operating in middle-income countries also have the highest values for overhead, taxes, and loan loss provisioning as shares of total assets. Net profits as a share of total
1. See Claessens,Demirguc-Kunt, and Huizinga(1997) for moredetailedinformation on the average spreadsof domesticand foreignbanksfor differentgroupingsof countriesby income.That articlealso considershow entryby foreignbanksaffectsthe interestspreadsand operatingcosts of domesticbanks. 2. Forcountrygroupingsby income,see WorldBank(1996).

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions


00 00



t OF-


aN t


a\ vX

O; u~~~C



? *e) ~~ .t ;; c t o1 et t_ ft e N vH C > >w \0 3n > 00 cnt H v x~~~e 00 X 411

Q l~~~~~~4 \_ W) 00 \0 N Qt 1M )3 _


~~ S:I~ ~ ~ ~ ~

C) 06




00 z~~~~0 tn



>N>>el nO C> a\ e 0 Nen=.a






>! : +.






en wio 0 rq m N

%. o

o tn 0i0EE '-1 t~~~~~~~~~~0 > ,$

s^ v0

o o~~~~~~~~~~~~~~~~~~~~t
St S t =

E n. C




ti @E 00 re CD VEq v- N \


v =

|~~~~~~~~~O <~~~~~~* F
t-- mav < m


01\ o

00 so c0 o \E u;;n m;m

C) C* xv <;





388 e cnene

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgiu-Kuntand Huizinga 389

assetstendsto be highestfor banksoperating in lower-income countries. Banks operatingin higher-income the lowest countries,instead,achieve net interest margins, and theyfacethe lowestvaluesof overhead, taxes,loan loss provisioning, and net profitsas sharesof total assets. The breakdown by regionrevealsthat the net interestmarginis highestfor banksoperatingin the transitioneconomiesat 6.4 percentand is also high in LatinAmerica in industrial at 6.2 percent. It is lowestfor banksoperating countriesat 2.7 percent. The transition economies further standout with highvalues
of overhead, taxes, loan loss provisioning, and net profits as shares of total assets. Banks operatingin industrialcountrieshave the lowest ratio of net profits to total assets at 0.4 percent, probably because of the high level of competition in banking services. Table 3 provides information on some of the macroeconomicand institutional variablesused in the regressionanalysis. The data are for 1995 or the most recent year available. The tax rate is computed on a bank-by-bankbasis as taxes paid divided by before-tax profits. The figure reported in the table is the average for all banks in the country in 1995. Reserves divided by deposits are the banking system's aggregate central bank reserves divided by aggregate banking system deposits. Actual reserve holdings reflect required and excess reserves. Reserves are generallyremuneratedat less-than-marketrates, and thereforeactual reserves may be a reasonable proxy for required reserves,as averaged over the different deposit categories. For several developing countries-Botswana, Costa Rica, Greece, and Jordan-the reserveratio is above 40 percent, indicating substantial financialrepression.In contrast, this ratio is low in Belgium,France,and Luxembourg at 0.01. The deposit insurancevariable is a dummy variablethat takes on a value of 1 if there is an explicit deposit insurance scheme (with defined insurance premia and insurance coverage) and a value of 0 otherwise. Even if there is an explicit deposit insurancescheme, however, the ex post insurancecoverage may prove to be higher than the de jure coverage, if the deposit insurance agency chooses to guaranteeall depositors. With a value of 0 there is no explicit deposit insurance, even if the authorities offer some type of implicit insurance. Next, table 3 presents some indicators of the structure of financial markets. The concentration variable is defined as the ratio of the assets of the three largest banks to the assets of the total banking sector. As is well known, the concentration of the U.S. banking market is low, at 16 percent, compared with values of about 50 percent for France and Germany. Note, however, that the U.S. figure may understatethe concentration ratio in individualbanking markets,which are protected from outside competition by interstatebanking and branchingrestrictions. The number of banks in the table reflects the number of banks in the data set with complete information. The ratio of bank assets to GDP is defined as the total assets of the deposit-money banks divided by GDP. This ratio reflects the banking sector's overall level of development.The ratio of stock market capitalization to GDPmeasuresthe extent of stock marketdevelopment.Developingcoun-

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

z> t m
I so \Z \a \o

' \zsll '


0000 c eq eq * o

000000 n -4 r" t-

000 v tGo

0000 0


r-, ch en en



C4 v t



N 00



N -

N \ko

r, .0\

ON0\ v



e 0






us0 \o


0 r4

t o Vheet>>-4 b> C 0 Cs


rn 0\ \vN C 0

co V

vX- N

oc v-

ho 0 v



o" v

-o as

0C o


000 co \0 \0

C) tn M en en


4~~~~~~~e M M

V \-


4 v-

4 00vt-

e V

s "


M v-


I -0






Qw c-i eo



o o

oo o

N4~~ N
N 00000

N o O OO-

~O nW4"-OO0WIN

O" \0M00N'
NeC N-



4 0

O~ N.

W z t

e .t~~~~~~~~~~~~~~~~~~~~l --


r O O.


O 00


- --

- fi e tO


This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions


0 1*


N- \- \0


XD -0 %


- en

N \0

e li q


C! q e_-w



- vN'0



N- e


N-00 e1 '*



'- r1



0 -l o C;c 00 ;C;< 0%r~N-~

N o-o;C




0 C; 0-0%0% NfrN-N o: O' <6 C; C; 0 o-C; C5 ci NO '0 O

%6 5 C5 C5



'0 00


n en G;6C



N 00 en Nr0

0 0 fe

\ '0 en 00 N Ml N

N- r "

\O0 V NNn


ON 0 N

V n p00 N en en

'.0 \.0 N e 00 X Nt" 0 N 0-

en N -

0 v

V -0



O t



t 0 0 \ e0


Hwtu F 0\ 0 N

0>ww> 0% o 4 00 en

n e0 f - 00 en "4 N N\.0 t% v " 0 N F N- t0 T-ON --I t

_ 0 en en 0 N o en en N- 00 '.0 0% en \ No 0 N0a 0 0 0 co e00 v n 0 00 0 00 N- en 0 \0 en O o N -0 - N- 00 N -0 -4 N N

z z z z z o

o. o. oL. so. o.o.o_oo -o


CL elooo . o.


-~~-o -~ ~ -~~ S



-~ o



This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions


4 - <

I - xD 4 t I tI e w .3 Sa "~~~~~~~%6




N ~ rl-. I wU

U-Q ~~~~~~~~~~~~~~~~~~~



0 ?> X c~~~~~~~>v i

:s vs

Y 2





.@ "-" " E "~~~~~~~~C:





IN \000 m as 0

fu *g



a 8 X r ?r).S: *=~~~~~~~t



k O b e e t

X X wo w t b e -X t O 0O Z r .R "9 9


x c c " *0Er?o<xW,

>9. |


" 6 E

ct e 0 en C)

,= se ,,

0 |>
( sYr

O 0 U " ,<, Jc

< 3 be



C;(6 C

C; 6

C;392 C

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgi4-Kuntand Huizinga 393

tries tend to have lower bank-to-GDP and capitalization-to-GDP ratios, with some notable exceptions. Malaysia, South Africa, and Thailand, for instance, have relatively high ratios for both variables. The final column in the table provides an index of law and order,which is one of the institutionalvariablesused in the regressionanalysis.This variableis scaled from 0 to 6, with higherscores indicating sound political institutionsand a strong court system. Lower scores reflect a tradition in which physical force or illegal means are used to settle claims. The table shows considerable variation among countries in the sample.

Tables 4 and 5 report the results of regressionsof the net interestmargin and before-tax profits as a share of total assets, respectively.Measuringprofitability using the return on equity (as opposed to using the returnon assets and controlling for equity ratios as we do here) does not lead to significantlydifferentresults and thus is not reported. All regressions include country and year fixed effects. The tables report severalspecifications, the basic one including a set of bank and macroeconomic indicators as regressors. These are important control variables to which we subsequentlyadd the taxation variables, the deposit insurance index, the financial structurevariables, and the legal and institutional indicators. We drop some variables from these two regressions because we want to ensure that banks from a reasonable number of countries are included. The estimation technique is weighted least squares, with the weight being the inverseof the number of banks for the country in a given year. This weighting correctsfor the fact that the number of banks varies considerably across countries. Bank Characteristics and Macroeconomic Indicators The first bank characteristicis the book value of equity divided by total assets lagged one period (EITAt-1). We lag total assets by one period to correct for the fact that profits, if not paid out in dividends, have a contemporaneousimpact on bank equity. Buser, Chen, and Kane (1981) examine the theoreticalrelationship between bank profitabilityand bank capitalization. They find that banks generally have an interioroptimal capitalization ratio in the presenceof deposit insurance. Banks with a high franchisevalue, reflectingcostly bank entry, have incentives to remain well-capitalized and to engage in prudent lending behavior (see Caprio and Summers 1993 and Stiglitz and Uy 1996). Berger(1995b) provides empirical evidence that U.S. banks show a positive relationship between bank profitability and capitalization, although the evidence is not conclusive. The author notes that well-capitalized firms face lower expected bankruptcycosts for themselves and their customers, thereby reducing their cost of funding. The basic specification (column 1 in tables 4 and 5) confirms that there is a positive relationshipbetweenEITAt-land net interestincome and bank profitability. In the regressionsthis variableis also interactedwith GDPper capita (measured

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



Table4. Determinants of Net Interest Margins

Independent variable Bankcharacteristics
Equity/laggedtotal assets



Regression results (3) (4)


Equity/laggedtotal assets interactedwith GDP per capita Loans/total assets Loans/total assets interacted with GDP per capita Non-interest eaming assets/total assets Non-interest earning assets/ total assets interactedwith GDP per capita Customer and short-term funding/total assets Customer and short-term funding/total assets interactedwith GDP per capita Overhead/totalassets Overhead/totalassets interacted with GDP per capita

0.046*** 0.047** (0.007) (0.007)

-0.001 (0.001) 0.017*** (0.004) -0.000 (0.000) -0.016** (0.007) 0.000 (0.001) 0.008** (0.004) 0.001*** (0.000)

0.044* (0.007)
-0.001 (0.001) 0.012*** (0.004) 0.001* * (0.000) -0.021*** (0.008)

0.064*** 0.063*** (0.007) (0.006)

-0.002* (0.001) 0.022** (0.004) 0.000 (0.000) -0.011 (0.007)

(0.001) 0.019*** (0.004) 0.000 (0.000) -0.020** (0.007)


40.001* (0.001) -0.007 (0.005)

-0.001 (0.001) 0.003 (0.005)

0.000 (0.001) 0.004 (0.006)

-0.002* (0.001) -0.000 (0.005)

-0.001 (0.001) -0.004 (0.005)

0.000 (0.000) 0.173*** (0.022) 0.002***

-0.000 (0.000) 0.025*** (0.019) 0.004*

0.000 (0.000) 0.213*** (0.019) 0.004*

-0.000 (0.000) 0.141*** (0.018) 0.009*

0.000 (0.001) 0.310*** (0.019) 0.005*

Foreign ownership dummy

0.003 (0.001) -0.001** (0.000)

0.004 (0.001) -.0.001* ** (0.000)

0.004*** (0.001)


(0.001) Foreign ownership dummy interactedwith GDPper capita 40.001** (0.000) Macroeconomic indicators
GDP percapita


0.000* * * 0.0001* (0.000) (0.000)

(0.001) 0.004 (0.008) 0.021 ** * (0.006) 0.044*** (0.007) 0.001 (0.002)

(0.001) 0.005 (0.008) 0.026* * * (0.006) 0.060*** (0.007) -0.004 (0.002)* *

(0.001) 0.006 (0.008) 0.025*** (0.006) 0.058*** (0.007) -0.003* (0.002)

Growth rate Inflation rate Real interest rate Real interest rate interacted with GDP per capita

0.000 -0.011w (0.001) (0.002) -0.011 -0.020)** (0.007 (0.008) 0.020*** 0.003 (0.006) (0.005) 0.051* * 0.025*** (0.007) (0.006)


-0.000 (0.002)

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgui-Kuntand Huizinga 395

Table 4. (continued)
Independentvariable Taxation Reserves Reservesinteractedwith GDP (1) (2) Regressionresults (3) (4) -0.076*** (0.015) -0.024* (0.016) (5) -0.104*** (0.016)

-0.076*** (0.015)

Tax rate Tax rate interactedwith GDP per capita

(0.003) 0.016*** (0.002) -0.001 (0.000)

(0.003) 0.015*** (0.002) 0.001*** (0.000)

0.009*** 0.004
(0.003) 0.017*** (0.002) (0.004) 0.017*** (0.002)

-.0.001**I -0.001 ' (0.000) (0.000)

Deposit insurance Deposit insurancedummy Financialstructure Bank assets/GDP Bank assets/GDP interactedwith GDPper capita Stock market capitalization/GDP Stock market capitalization/GDP interactedwith GDP per capita Stock market capitalization/bankassets Stock market capitalization/bankassets interactedwith GDP per capita Number of banks Market concentration Total assets (U.S. dollars) Legal and institutional indicators Contract enforcementdummy Contract enforcementdummy interactedwith GDP per capita Law and order index Law and order index interactedwith GDP per capita Corruption

-0.009*** (0.003) 0.024* (0.010) 0.001* (0.001)


(0.005) -0.002** (0.001)


(0.003) 0.001 (0.001) -0.001 (0.015) 0.004 (0.005) 0.003*** (0.000) -0.042*** (0.007) 0.003*** (0.001) -0.003 ** (0.001)

(0.000) 0.009*** (0.001) (fable continued on following page.)

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



Table 4. (continued)
Independent variable Corruption interacted with
GDP percapita



Regression results (3) (4)


0.50 5,841 0.51 5,276 0.50 5,212 0.58 5,054 (0.000) 0.63 4,497

Adjusted R2 Number of observations


at the 10 percentlevel. Significant at the 5 percentlevel. Significant at the 1 percentlevel. Significant Note: The regressionsare estimatedusingweightedleast squarespooling bank-leveldata across80 countries for the 1988-95 time period. The number of banks in each period is used to weight the The also includecountryand time dummyvariablesthat are not reported. observations.The regressions incomedividedby totalassets.Standard definedas netinterest margin variableis the net interest dependent errorsare given in parentheses. Source:Authors'calculations.
** * *

Table S. Determinants of Bank Profitability

Independent variable Bank characteristics Equity/laggedtotal assets (1) (2) Regression results (4) (3) (5)

Equity/laggedtotal assets interactedwith GDP per capita Loans/total assets Loans/total assets interacted with GDP per capita Non-interest earning assets/ total assets Non-interest earning assets/ total assets interacted with GDP per capita Customer and short-term funding/total assets Customer and short-term funding/total assets interacted with GDP per capita Overhead/total assets Overhead/total assets interacted with GDP per capita


(0.009) 0.002*** (0.001) -0.024*** (0.005)

(0.009) 0.003*** (0.001) -0.023 ** (0.005)


(0.009) 0.002 (0.001) -0.013*** (0.005) 0.001 (0.000) -0.005 (0.010)

0.015*** (0.006)

0.002*** 0.003*** (0.001) (0.001) -0.015*** -0.018*** (0.004) (0.005)


(0.000) -0.010 (0.010)

(0.000) -0.011 (0.010)

(0.000) -0.014 (0.010)

0.001 (0.000) -0.033*** (0.007)

-0.007*** (0.001) -0.029*** (0.006)


(0.001) -0.017** (0.007)

-0.007*** (0.001) -0.014*** (0.008)

-0.008*** (0.001)

0.002** (0.001)



0.002*** (0.000) -0.023 (0.025) -0.030***

-0.000 (0.001) -0.006 (0.026) -0.049***

-0.000 (0.001) -0.004 (0.026) -0.049***

0.001 (0.001) -0.024 (0.026)


0.002*** (0.000) -0.114*** (0.019) 0.007***






This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgiu-Kuntand Huizinga 397

Table 5. (continued)
Independentvariable Foreignownership dummy (1) 0.005***


Regressionresults (3) (4) 0.006**


(5) 0.006***


(0.001) -0.001*** (0.000)

Foreign ownership dummy interactedwith GDP per capita -0.001





Macroeconomic indicators
GDP per capita

0.008*** (0.001)

0.008*** (0.001)




Growth rate Inflation rate Real interest rate Real interest rate interacted with GDP per capita

0.002 (0.010) 0.011 (0.008) 0.023* * I (0.009) -0.000 (0.002)

-0.006 (0.011) 0.015* (0.008) 0.029*** (0.010) -0.001 (0.002)

-0.007 (0.011) 0.014* (0.008) 0.029 ** (0.010) -0.001 (0.002)

-0.019 (0.011) 0.009 (0.008)


(0.009) -0.000 (0.002)

0.004 (0.007) 0.011* (0.005) 0.026* * * (0.006) -0.003** (0.002)

Taxation Reserves Reservesinteracted with GDP per capita Tax rate Tax rate interacted with GDP per capita

-0.126*** (0.021) 0.029* (0.004) 0.022*** (0.003) -0.000


-0.129*** (0.021) 0.031 (0.004) 0.022*** (0.003) -0.000?


-0.106*** -0.091*** (0.023) (0.016) 0.032*** (0.004) 0.021*** (0.003) 0.003


0.005S** (0.004) 0.017*** (0.002) 0.000?


Deposit insurance Deposit insurance dummy Financialstructure

Bank assets/GDP Bank assets/GDP interacted with
GDP per capita

-0.005 (0.004)
-0.028* (0.014)


Stock market capitalization/GDP Stock market capitalization/ GDP interacted with GDP per capita Stock market capitalization/bankassets

0.010 (0.007)



{Table continued on following page.)

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



TableS. (continued)
Indicator Stock market capitalization/bank assets interacted with GDP per (1) (2) Regressionresults (3) (4) (5)

Number of banks Market concentration Total assets (U.S. dollars) Legal and institutional indicators Contract enforcement dummy Contract enforcement dummy interactedwith GDP per capita Law and order index Law and order index interactedwith GDP per capita Corruption Corruption interacted with GDP per capita Adjusted R2 Number of observations

(0.001) 0.000 (0.000) 0.010* (0.007) 0.000 (0.000)

(0.007) 0.001VE (0.001) -0.000 (0.001) {0.000?




0.21 5,841

0.27 5,276

0.27 5,212

0.31 5,054

0.35 4,497

at the 5 percentlevel. Significant Significant at the 1 percent level. Note:Theregression leastsquares is estimated usingweighted poolingbank-level dataacross80 countries for the 1988-95 time period.The number of banksin eachperiodis usedto weightthe observations. The thatare not reported. regressions also includecountryand timedummyvariables The dependent variable is before-taxprofitsdividedby total assets.Standard errorsaregiven in parentheses. Source:Authors'calculations.

Significant at the 10 percent level.

in constantthousands dollarsfor the year 1987). The positivecoefficient on the interaction variables in the before-tax profitsregression mayreflecta higherbank franchise valuein wealthier countries. Thecoefficients on ETA,.1andthe interaction variable togetherindicate how the ratioof equityto assetsaffectsbankvariables in countries with differentincome levels. For a country with a per capita GDP of $10,000, for instance, the point estimate of the effect of ETA, 1 on before-tax profits divided by total assets is 0.067 (or 0.047 + 10 x 0.002). There is a negative and significantcoefficient on non-interest-earningassets as a share of total assets in the net interest margin equation, but there is no significant relationship for the before-tax profits equation. Note that the sign on this variable interacted with per capita GDPis negative in both the net interest margin

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgui-Kuntand Huizinga 399

the presence of non-interest Apparently, profitsspecifications. andthe before-tax morein wealthier net interestincomeand profitability earningassetsdepresses the signon loansdividedbytotal Bycontrast, thanin poorercountries. countries assetsis positivein the net interestmarginequationand negativein the beforeinteracted withGDP of thisvariable the coefficient However, tax profitsequation. income levels bank that at higher indicating is positive, equation in the profits tend to be moreprofitable. lendingactivities fundingconsistsof demand On the liabilityside, customerand short-term this typeof customer deposits,savingsdeposits,and time deposits.On average, fundingmay carrya low interestcost, but it is costly in termsof the required affectthe net network.This liabilitycategorydoes not significantly branching thatit lowersbankprofitability. althoughthereis evidence variable, interest in bank businessand in overheadmay also capturedifferences Differences Theratioof in therangeandqualityof services. mix, as well as variation product of 0.173 in the net interest coefficient overhead to total assetshas an estimated cost is thatabouta sixthof a bank'soverhead whichsuggests regression, margin with per of overhead passedon to its depositorsand lenders.The interaction
capita GDPalso enters with a positive coefficient, indicating that a largershare of overhead is passed onto financial customers in wealthier countries. This may reflect more competitive conditions in banking markets in industrialthan in developing countries. In the before-tax profits regression the interaction of overhead with per capita GDP enters negatively, indicating that higher overheads eat into bank profits. The foreign ownership variable equals 1 if at least 50 percent of the bank's stock is in foreign hands and equals 0 otherwise. In both tables 4 and 5 this variablehas a positive coefficient, while its interactionwith per capita GDPhas a negative coefficient. These results suggest that foreign banks realize relatively high net interest margins and profitability in relativelypoor countries. It may be that foreign banks are frequently exempt from unfavorable domestic banking regulationsand apply superiorbankingtechniques.Note, however, that the point estimate of the effect of foreign ownership for a wealthy country with a per capita GDP of $20,000 is negative in the net interest margin equation at -0.016 (that is, 0.004 - 20 x 0.001) and in the profitabilityequation at -0.015 (that is, 0.005 - 20 x 0.001). Foreign banks' technological and efficiency advantages in countries may be insignificant because, while there, they face informational disadvantages. This could explain why foreign banks in industrial countries are relatively unprofitable on average. Turningto the macroeconomicindicators,we see, first, that per capita GDPhas no significant impact on the realized net interest margin, although it enters the profitabilityequation with a positive coefficient. Percapita GDPis a generalindex of economic development, and it thus reflectsdifferencesin bankingtechnology, the mix of banking opportunities, and any aspects of banking regulations omitted from the regression. Growth, defined as the growth rate of real per capita GDP, is insignificant in both regressions.The percentagechange in the GDP defla-

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



tor, or inflation,is estimated to increase the net interest and bankprofitmargin ability.However,the significance of the coefficients in the profitability regressions is low, possiblybecausebanksobtainhigherearnings fromfloator because thereare delaysin crediting customeraccountsin an inflationary environment. Withinflation,bankcosts also tendto rise.A larger number of transactions may leadto higher laborcostsand,as shownbyHansonandRocha(1986:40), result in a higherratio of bank branches per capita.On net, however,the regression resultssuggestthat the impactof inflationon profitability, althoughnot very
significant, is positive throughout.

We constructed the real interestrate using the short-term debt government yield and, if that measure was not available, othershort-term marketrates.The realinterest rateentersthe net interest and before-tax margin profitsregressions positively,althoughthis variableinteracted with per capitaGDPhas a significantly negativecoefficientin the ne t interestmarginequation.Thus there is someevidence thatincreases in therealinterest ratedo not raisespreads as much
in industrialcountries, perhaps because their deposit rates are not tied down by

depositrateceilings. TaxationVariables Banksare subjectto directtaxationthroughthe corporateincometax and other taxes, and they are subjectto indirecttaxationthroughreserverequirements.Reserve requirements are an implicittax on banksif, as is usual,official reservesare remunerated at less-than-market rates.The corporateincometax and the reserve tax differin important respects. First,the corporate incometax, in principle at least,can be targeted at pureprofit.To the extentthatit is a profit
tax, the corporate income tax is relatively nondistorting. In practice, however, it

may not be a pureprofittax if completeexpensing of costs is not allowed. The reservetax, by its verynature,is proportional to the volumeof deposit takingandis therefore a distorting tax. Froma welfareperspective the corporate incometax thus appearsto be superior to the reservetax. A secondimportant difference is that the severity of the reserve tax dependson the opportunity cost of holdingreserves. Thismaydependon financial market conditions as muchas on any tax code. Relatedto this secondcondition,reserve requirements arealso
an instrumentof monetary policy.

As faras we know, therehas beenno otherempirical research on the effectof the corporate incometax on the banking sector.Several studieshaveconsidered the impactof reserverequirements on bank profitability. Some, in particular, have examinedhow FederalReservemembership affectedthe profitability of U.S. commercial banksin the 1970s (see Rose and Rose 1979 and Gilbertand Rasche1980). Federal Reserve membership subjected banksto generally higher reserverequirements. Most of the studiesin this area supportthe notion that nonmember banksweremoreprofitable thanmember banks(withsimilar characteristics)becausenonmember banks held relativelylittle cash. Competition amongmember and nonmember banksin the samemarketappears to havepre-

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgui-Kuntand Huizinga 401

ventedmember banksfrompassingtheirhigherreserve costs onto theircustomers. In relatedwork Kolari,Mahajan, and Saunders (1988) studythe impactof announced changesin reserve on bankstockpricesusingan event requirements studymethodology. andLemmen Huizinga (1996)andEijffinger, Huizinga, (1998) examinehow nonresident taxes affectinterestrates,while Fabozzi withholding in reserve andThurston arepriced (1986) examinehow differences requirements into money-market instruments. in our Becausedetailedinformation of all countries on the reserve regulation We construct sampleis not available, we use a proxyto capturebankreserves. this variablein the regressions as the productof the bankingsystem'sratio of aggregate reserves to deposits(as in table3) and the individual bank'sratio of short-term andshort-term fundingto total assets.Customer funding, consisting of demand deposits, savings andtimedeposits, hereproxyforreservable deposits, deposits.The reserves variable is thus an approximation of actualbankreserves thatreflectsdifferences in systemwide reserve requirement rules. In tables4 and 5 the reserves variableentersthe regressions negatively. The coefficientsin the net interestmarginequationsshow two effects:less-thanmarketremuneration and the impacton banks'lendingand depositrates.The firsteffectis expectedto be negativebecause underremunerated reserves lowera The secondeffectcould be either bank'snet interestincomeand profitability. zero,in whichcasethe bankbearsthe fullcost of higherreserves, or positive,in which case the cost of reservesis passedonto bankcustomersin the form of higherinterestmargins.In table 5 we see that the reserves variablenegatively affectsbankprofitability. thatthesecond,or pass-through, Thissuggests effectis eithernonexistentor too smallto offset the first,or direct,effect.Abstracting fromany pass-through, the coefficient on the reserves in eitherregresvariable sioncan also be interpreted cost of holdingreserves. as a bank'sopportunity The reserves variable interacted withpercapitaGDPentersbothregressions positively, whichmayreflectthe factthatthe opportunity cost of holdingreserves is higher in wealthier countries. We capturethe explicittaxesthat bankspaywith the variable tax rate,which is measured by a bank'stax billdividedby its pretaxprofits.Thisvariable has a and profitability. significantly positiveimpacton interestmargins The tax rate interacted with per capita GDP is negativeand significant in both regressions. Theseresultssuggestthat boththe net interest marginandprofitability increase with tax rates,but less so in richercountries. Thusthe corporate incometax is to somedegree. passedthroughto bankcustomers To calculate theextentof thispass-through, we usetheestimated coefficients on the tax ratevariable andits interaction with percapitaGDP. Letthe pass-through be defined in before-tax as the increase profitsfollowing a one-unit increase in the = ,B, tax bill,or aBTPI3X. Next, notethat(aBTP/1c)/TA corporate whereXis the tax rate, TA is total assets,and X is estimated at 0.022 - (0.0004)x (percapita = (aBTPIat) + BTP,as TX = r x BTP.It now followsthat GDP). Further, aTX/ar = f/[frr+ (BTP/TA)]. aBTP/aTX We canevaluate thisexpression usingmeanval-

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



ues of t, BTPITA, and per capitaGDP separately for countriesin four different incomegroups(low,lower-middle, andhigh),wherepercapitaGDP upper-middle, is theinternational for1995.Thecalculations average thatthepass-through suggest coefficient,MBTPI/TX, equals1.01, 0.72, 1.00, and 1.21 forcountries in the four in 1995BTP/TAandT have incomegroups,respectively. Forlow-income countries meanvaluesof 0.016 and0.225 for all banks,whilethe average GDPpercapitais $426. Thecalculations thefactthatin high-income reflect themeanvalue countries of BTPITAis lower,whilethevalueof X changes little. income tax completely thatthecorporate Essentially, theseresults suggest passes throughto bankcutomers. Thusthereis no supportfor the notionthat the corporateincometax is a nondistorting tax on bankprofits.Generally, it is a sourcebasedtax on domestically A completepass-through employed capitalresources. of this tax is consistent with the assumption thatinternational investors demand a net-of-taxreturnon capitalinvestedin a particular countryindependent of the country'ssource-based taxes.
Deposit Insurance

Severalstudieshaveexamined the impactof depositinsurance usinginternational data. Demirguc-Kunt and Detragiache (1997) find that the existenceof explicitdepositinsurance is positively withtheprobability associated of banking crises.Barth, Nolle, andRice(1997),however,findthatdepositinsurance hasno significant impacton banks'return on equityfor a sampleof 142 banksin 1993. Bartholdy, Boyle,and Stover(1997) estimatethat depositinsurance lowersthe depositrateby 25 basispoints,usingaggregate depositinterestratedata for 13 OECD countriesduring1985-90. These authorsdiscusswhy depositinsurance has a theoretically ambiguous effecton interestmargins. On the one hand,the depositratefor insured depositsshoulddecrease giventhe insurance protection. On the otherhand,mispriced depositinsurance bankswithan incentive provides to engage in more risky lendingstrategiesto increasethe contingentpay-out fromthe depositinsurance agency. Brewerand Mondschean (1994) offer empirical supportfor the notion that depositinsurance createsincentives for banksto acquire riskyassetsby examining the junkbond holdingsof U.S. banks,while Demirgii-Kuntand Huizinga (1993) arguethat depositinsurance was an important determinant of bankstock pricesduringthe international debtcrisisof the 1980s. Thismoralhazardproblem and the associatedriskscan leadbankcreditors to demanda higherinterest rate. Also, for a given level of risk, deposit insurance may lead banksto lend moneymorecheaplythantheyotherwise would,depressing net interest margins andprofitability. Evenbanksthatdo not engagein riskylendingstrategies themselves may experiencea downwardeffect on interestmarginsbecauseof bank competition. The depositinsurance variable equals1 if the countryhas an explicitdeposit insurance regime.Forsomecountries this variable changeswith time,reflecting changesin the regimeduringthe sampleperiod.The resultssuggestthat deposit

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgiu-Kuntand Huizinga 403

insurancelowers net interestmargins. Deposit insurancemay also influencemargins and profits through its effect on financial structure-it encouragesnew entry and enables small banks to operate. However, when we include financial structure variables in the regression, the results do not change. The impact on bank profits is negative, but it is not significant, possibly because of the offsetting impact of mispricedsubsidies in actual deposit insuranceschemes. These results suggest that explicit deposit insurance regimesdo not produce higher bank profitability and margins, perhaps because of design and implementationproblems.

Variables Financial Structure

In the regressionsreportedin column 4 of tables 4 and 5 we includetwo sets of financial structure variables. The first set -comprises the market concentration ratio, the number of banks, and total bank assets as indicators of market structure and scale effects. Various authors, such as Gilbert (1984), Berger(1995a), and Goldberg and Rai (1996), have pointed out that such variables may proxy for market power as well as for differencesin bank efficiency.We do not attempt here to distinguish between the correspondingmarket power and efficient structure hypotheses. The second set consists of financial structure variables that measure the importance of bank and stock market finance relative to GDP and to each other. Reasons why these variablesmatter may also hinge on marketpower arguments. A high ratio of bank credit to GDP, for instance, may reflect a high demand for banking services fueling competition among banks. Or these variables may reflect the complementarityof or substitutability between bank and stock market finance. The Miller-Modigliani theorem states that debt and equity finance are purely substitutes in the absence of taxes and bankruptcycosts. In practice,however, debt and equity finance may also be complementary, as modeled in Boyd and Smith (1996). Demirguq-Kuntand Maksimovic (1996) provide empirical evidence that an ability to attract equity capital may also enhance firms' borrowing capacity, especially in developing countries' financial markets.In this setting easier equity finance may increase rather than decrease the demand for debt finance, reflecting that these sources of finance are complements. Turning to the first set of market concentration and scale variables, we see that the bank concentration ratio has a significant and positive impact on bank profitability, while bank size, as proxied by total assets, has a significant and positive impact on interest margins. The number of banks has no significant impact on either interestmargins or profits. The second set of financial structurevariables affects bank marginsmore significantly than bank profits. This may indicatethat these variableshave a greater impact on banks' loan and deposit customers than on other clients. The ratio of bank assets to GDP has a significantly negative impact on margins and profits, perhaps reflecting more intense interbank competition in well-developed financial systems. This effect is smaller in richercountries that alreadyhave relatively developed bankingsectors. The ratio of stock marketcapitalizationto GDPenters

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



the net interestmarginequationpositively,which suggeststhat a largerstock marketper se enablesbanksto obtain higherinterestmargins,supporting the complementarity betweendebtandequityfinancing hypothesis discussed above. As stockmarketsdevelop,betteravailability of information the potenincreases tial pool of borrowers, makingit easierfor banksto identifyand monitorthem. Thisraisesthe volumeof business for banks,makinghighermargins possible.In the regression the ratio of stock marketcapitalization to bankingassetsenters the interestmarginequationnegatively. Thusit may be that a largerstockmarket relativeto the bankingsectorlowers bank margins,reflecting substitution possibilities betweendebt and equity.For both stockmarketdevelopment indicators the interaction with per capita GDPentersthe interestmarginequation with the oppositesign, suggesting that the impactof any stock marketdevelopmenton interestmargins is mutedin wealthier countries. Legaland Institutional Indicators The finalregressions repottedin tables4 and 5 includea varietyof legaland institutional variables(column5). The contractenforcement dummy,ranging from 1 to 4, measures the degreeto which contractual agreements are honored and not subjectto languageand mentalitydifferences. A highervalue means greater contract enforcement. Inboththenetinterest margin andbefore-tax profits regressions, the contract enforcement variable hasa negative andsignificant sign. Poorcontractenforcement maypromptbanksto require higherinterest margins and investorsto requirehigherprofitability to compensatefor the additional risk. In both regressions the contractenforcement variableinteracted with per capitaGDPenterspositively, suggesting a mutedeffectof thisvariable in wealthier

The law and orderindex, rangingfrom 0 to 6, captureshow well the legal systemworksin adjudicating disputes.Fromtable4 we see thata higher valueof this indexis significantly associated with lowerinterest margins. Thereasonmay againbe that an effectivelegalsystemreducesthe required riskpremiaon bank lending.The interactionbetweenthe law and orderindex and per capitaGDP entersthe equationnegatively, however. Finally,the corruptionindex, rangingfrom 0 to 6, measuresthe degreeof government A higherscore indicatesthat government corruption. officialsare less likelyto take bribes.Table4 indicatesthat a cleanergovernment is associated with lower realizedinterestspreads,and this relationshipis weaker in wealthiercountries.Again, banks may requirea lower risk premium on their investments in countries thatarerelatively freeof corruption. Overall, the regressions indicatethat the underlying legaland institutional variables areimportant
in explaining cross-country variation in interest spreads and bank profitability.

Fortwo of the threevariables the interaction withpercapitaGDPhasa coefficient with the oppositesign,suggesting that the effectsof institutional differences are mutedin wealthiercountries.

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

and Huizinga 405 Demirgua-Kunt


Bankingsystemsaroundthe world differwidelyin theirsize and operation. Acrosscountries commercial macroeconomic bankshaveto dealwith different finanenvironments, explicitandimplicittax policies,depositinsurance regimes, cial marketconditions, and legaland institutional realities. Usinga comprehensive cross-country how bank dataset with bank-level data,this articleanalyzed affecthow banksfunction characteristics and the overallbankingenvironment as reflected in interest margins and bankprofitability. Wecanconfirm research: somefindings of earlier a positive forinstance, relationand a negative ship betweencapitalization and profitability between relationship reserves and profitability. Butotherimportant determinants of bankmargins and profitability, suchasownership, taxation, andthelegal corporate financial structure,
and institutionalsetting,have not been treatedextensivelyin the literature. Differences in the mix of bank activity also have an impact on spreads and profitability. Our results show that banks with relatively high non-interest earning assets are less profitable. Banks that rely largely on deposits for their funding are also less profitable, because deposits apparentlyentail high branching and other expenses. Similarly,variation in overheadand other operatingcosts is reflected in variation in bank interest margins, because banks pass on their operating costs to their depositors and lenders. The international ownership of banks also has a significant impact on bank spreads and profitability.Foreign banks, specifically,realize higher interestmargins and higher profitability than domestic banks in developing countries. This finding may reflect the fact that in developing countries a foreign bank's technological edge is relativelystrong, apparentlystrong enough to overcome any informational disadvantage in lending or raising funds locally. Foreign banks, however, are shown to be less profitable in industrialcountries, where they may not have a technological edge. Macroeconomic factors also explain variation in interest margins. We found that inflation is associated with higher realized interest marginsand higher profitability. Inflation entails higher costs-more transactions and generally more extensive branch networks-and also higher income from bank float. The positive relationship between inflation and bank profitability implies that bank income increasesmore with inflation than do bank costs. Further,high real interest rates are associated with higher interest margins and profitability, especially in developing countries. This may reflect the fact that in developing countries demand deposits frequentlypay zero or below-marketinterest rates. Regardingfinancialstructure,banksin countrieswith morecompetitivebanking sectors-where banking assets constitute a larger portion of GDP-have smaller margins and are less profitable. The bank concentration ratio positively affects bank profitability, and larger banks tend to have higher margins. A larger ratio of stock marketcapitalizationto GDPincreasesbank margins,suggestingpossible

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



betweendebt andequityfinancing. complementarity A largerratio,however,is negativelyrelatedto margins,suggestingthat relativelywell-developed stock marketscan substitute for bankfinance. Similarly, severalinstitutional factors,suchas indexesof creditrights,law and in financialstructure, have more proorder, and corruption,and differences nouncedeffectson interest in developing thanin and bankprofitability margins industrial of bankcountries. the relatively closednature Theseresults mayreflect ing markets in developing countries. evidence that Coupledwithearlier empirical a weak institutional environment makesbankingcrisesmorelikely(DemirgiiuKuntand Detragiache 1997), theseresultssuggestthat returns to improving underlyinginstitutions imalso havea morepronounced areindeedhigh.Reserves pact on margins in developing thanin industrial andprofitability countries. This resultmay simnply reflectthe relatively cost of holdingreserves highopportunity in poorerand moreinflationary countries. The corporate incometax appears to be passedon fullyto bankcustomers in both developing Thisfindingis consistent and industrial countries. with the notion that bankstock investorsrequirenet-of-company-tax returnsindependent of the levelof companytaxation.It also impliesthatthe corporate incometax on banks is likely to distortthe underlying savingand investment decisions,with possiblynegativeimplications for economicgrowth.Theseconsiderations must weigh heavilyin considering the meritsof the corporate incometax on banksas part of the overalltax system. However, we also found that official reservesdepressbank profits.Prima facie, this suggeststhatreserve requirements area betterinstrument with which to tax bankprofitsthanthe corporateincometax. Note, however,thatthe implicit reservetax in manycountriesis muchmore variablethan the corporate incometax. The levelof banking investment and activityis therefore unlikelyto be adjustedto each changein the implicitreservetax. Variability in the reserve tax can thusgo a longway towardexplaining the responsiveness of bankprofits to this tax. These issuesare pursuedfurtherin Demirguc-Kunt and Huizinga (1997). Policymakers inpromoting haveaninterest banking sectors thatarebothstable and efficient.Stability clearlyrequires sufficient banking profitability, whileeconomic efficiencyrequires bankingspreadsthat are not too large.A prerequisite to formulating effectivebanking policiesis thusto understand the determinants of bankprofitability andinterest margins. Several othertopicsremain for further differconstudy.Countries worldwide siderablyin the extentof foreignownership of theirbankingsystems.An interestingissueis how entryby foreign banksaffectsthe operation of domestic banking firms.In principle,foreignentrycan affect pricingby domesticfirmsand force them to reducetheiroperatingcosts and to remaincompetitive. Both of these effectsdetermine whetherthe entryof foreignfirmsis welfare-improving overall.We addressthe impactof foreignentry in Claessens,Demirgui-Kunt, and Huizinga(1997).

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions

Demirgii-Kunt and Huizinga 407

As a relatedissue,it wouldbe interesting to consider whatdetermines foreign bankentry.Foreign bankentry,and foreigndirectinvestment generally, maybe drivenby the different taxationof domestic andforeign (worldwide) firmsrather thansimplyby countries' in providing comparative advantage financial services. We havefoundsomeevidence thatgovernment suchas the design regulations, of depositinsurance schemes, have an impacton bankmargins. It would be interesting to analyze thisissuefurther by takinginto accountdifferences in design features. We intendto return to theseissuesin futurework.

The word "processed" describes informally reproduced works that may not be commonly available through librarysystems. Barth,James R., Daniel E. Nolle, and Tara N. Rice. 1997. "CommercialBankingStructure, Regulation, and Performance:An International Comparison." Comptroller of the CurrencyEconomicsWorkingPaper97-6. Comptrollerof the Currency,Washington, D.C. Processed. Bartholdy,Jan, Glenn W. Boyle, and Roger D. Stover. 1997. "Deposit Insurance,Bank Regulation, and Interest Rates: Some InternationalEvidence." University of Otago, New Zealand. Processed. Berger,Allen N. 1995a. "TheProfit-Structure Relationship in Banking:Tests of MarketPower and Efficient StructureHypotheses." Journal of Money, Credit, and Banking 27(2):404-31. 1995b. "TheRelationshipbetween Capital and Earningsin Banking."Journal of Money, Credit, and Banking27(2):432-56. Boyd, John, and BruceSmith. 1996. "The Co-Evolutionof the Real and FinancialSectors in the Growth Process." The World Bank Economic Review 10(2):371-96. Brewer, Elijah III, and Thomas H. Mondschean. 1994. "An EmpiricalTest of the Incentive Effects of Deposit Insurance."Journal of Money, Credit, and Banking 26(1): 146-64. Buser,Stephen A., Andrew H. Chen, and EdwardJ. Kane. 1981. "FederalDeposit Insurance, RegulatoryPolicy, and Optimal BankCapital."Journal of Finance35(1):51-60. Caprio, Gerard,Jr., and LawrenceH. Summers.1993. "Financeand Its Reform:Beyond Laissez-Faire."Policy ResearchWorking Paper 1171. World Bank, Policy Research Department, Washington, D.C. Claessens, Stijn, Asli Demirgui-Kunt, and Harry Huizinga. 1997. "How Does Foreign Entry Affect the Domestic Banking Market?" World Bank, Policy ResearchDepartment, Washington, D.C. Processed. Demirgui-Kunt, Asli, and EnricaDetragiache. 1997. "The Determinantsof BankingCrises: Evidence from Developed and Developing Countries." World Bank, Policy Research Department, Washington,D.C. Processed. Demirgiiu-Kunt,Asli, and HarryHuizinga. 1993. "Official Creditsto Developing Countries: Implicit Transfers to the Banks." Journal of Money, Credit, and Banking 25(3):430-44. . 1997. "The Taxation of Domestic and Foreign Banking." World Bank, Policy Research Department,Washington, D.C. Processed.

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions



Demirgiiu-Kunt,Asli, and Vojislav Maksimovic. 1996. "StockMarket Developmentand Financing Choices of Firms." The World Bank Economic Review 10(2): 341-69. Eijffinger,Sylvester, Harry Huizinga, and Jan Lemmen. 1998. "Short-Termand LongTerm Government Debt and Nonresident Interest Withholding Taxes." Journal of Public Economics 68(3):309-34. Fabozzi, FrankJ., and Thom B. Thurston. 1986. "StateTaxes and ReserveRequirements as Major Determinantsof Yield Spreadsamong Money Market Instruments." Journal of Financialand Quantitative Analysis 21(4):427-36. Gilbert, R. Alton. 1984. "BankMarket Structureand Competition."Journal of Money, Credit, and Banking 16(4):617-60. Gilbert, R. Alton, and Robert H. Rasche. 1980. "FederalReserve Bank Membership: Effects on Bank Profits."Journal of Money, Credit,and Banking 12(3):448-61. Goldberg, Lawrence G., and Anoop Rai. 1996. "The Structure-Performance Relationship for European Banking."Journal of Banking and Finance 20(4):745-71. Gorton, Gary, and Richard Rosen. 1995. "CorporateControl, Portfolio Choice, and the Decline of Banking."Journal of Banking 50(5):1377-420. Hanson, James A., and Roberto de Rezende Rocha. 1986. "High InterestRates, Spreads, and the Cost of Intermediation:Two Studies." Industryand Finance Series 18. World Bank, Industry Department,Washington, D.C. Huizinga, Harry. 1996. "TheIncidenceof InterestWithholdingTaxes: Evidencefrom the LDCLoan Market." Journal of Public Economics 59(3):435-51. InternationalMonetary Fund. Various years. InternationalFinancialStatistics.Washington, D.C. Kolari,James, Arvind Mahajan,and EdwardM. Saunders.1988. "The Effectsof Changes in Reserve Requirements on Bank Stock Prices." Journal of Banking and Finance 12(2):183-98. Kyei, Alexander. 1995. "Deposit Protection Arrangements:A ComparativeStudy." IMF Working Paper 951134. InternationalMonetary Fund, Washington, D.C. Levine, Ross. 1997. "FinancialDevelopment and Economic Growth." Journal of Economic Literature35(2):688-726. Rose, John T., and PeterS. Rose. 1979. "The Burdenof FederalReserveSystemMembership: A Review of the Evidence."Journal of Banking and Finance 3(2):331-45. Schranz,Mary S. 1993. "TakeoversImproveFirmPerformance: Evidencefrom the Banking Industry."Journal of Political Economy 101(2):299-326. Stiglitz,Joseph E., and Marilou Uy. 1996. "FinancialMarkets,Public Policy, and the East Asian Miracle." The WorldBank Research Observer 11(2):249-76. Talley, Sam, and Ignacio Mas. 1990. "Deposit Insurancein Developing Countries."World Bank, Latin America and the Caribbean Regional Office, Washington, D.C. Processed. Vittas, Dimitri. 1991. "MeasuringCommercialBankEfficiency,Use, and Misuse of Bank OperatingRatios." Policy ResearchWorking Paper806. World Bank, Policy Research Department, Washington, D.C. World Bank. 1996. WorldDevelopment Report 1996: From Plan to Market.New York: Oxford University Press.

This content downloaded from on Tue, 2 Jul 2013 00:32:50 AM All use subject to JSTOR Terms and Conditions