Noninterest-income-and-financial-performance-at-US

TheFinancialReview39(2004)101--127NoninterestIncomeandFinancialPerformanceatU.S.CommercialBanksRobertDeYoungFederalReserveBankofChicagoTaraRice∗FederalReserveBankofChicagoAbstractNoninterestincomenowaccountsforover40%ofoperatingincomeintheU.S.com-mercialbankingindustry.Thispaperdemonstratesanumberofempiricallinksbetweenbanknoninterestincome,businessstrategies,marketconditions,technologicalchange,andfinancialperformancebetween1989and2001.Theresultsindicatethatwell-managedbanksexpandmoreslowlyintononinterestactivities,andthatmarginalincreasesinnoninterestincomeareassociatedwithpoorerrisk-returntradeoffsonaverage.Thesefindingssuggestthatnoninter-estincomeiscoexistingwith,ratherthanreplacing,interestincomefromtheintermediationactivitiesthatremainbanks’corefinancialservicesfunction.Keywords:banks,noninterestincome,deregulationJELClassifications:G21,G28∗Correspondingauthor:Economist,DepartmentofBankSupervisionandRegulation,FederalRe-serveBankofChicago,230SouthLaSalleStreet,Chicago,IL60604;Phone:[phone];E-mail:[e-mail]rethoseoftheauthorsanddonotnecessarilyreflecttheviewsoftheFederalReserveBankofChicagoortheFederalReserveSystem.TheauthorsthankPhilBartholomew,StevenDennis,GerryHanweck,IftekharHasan,CathyLemieux,TomLutton,LarryWall,andananonymousreviewerfortheirhelp,comments,andencouragement,andJenniferBlairforexpertresearchassistance.101102R.DeYoungandT.Rice/TheFinancialReview39(2004)101–1271.IntroductionAnumberoftheorieshavebeenadvancedtoexplainwhybanks,andmoregen-erallyfinancialintermediaries,exist.Inmostofthesetheories,banksexistbecausetheymitigateahostofproblemsthatotherwisepreventliquidityfromflowingdi-rectlyfromagentswithexcessliquidity(depositors)toagentsinneedofliquidity(borrowers).Theseproblemsarisebecauseofinformationalasymmetries,contract-ingcosts,andscalemismatchesbetweenliquiditysuppliersandliquiditydemanders.Intermediation-basedtheoriesoffinancialinstitutionsseebanksasthesolutiontotheseproblems,because:bankshaveacomparativeadvantageatgatheringinforma-tiononborrowercreditworthiness;banksarebetterablethanindividuallenderstomonitorborrowers;banksprovideincreasedliquiditybypoolingfundsfrommanyhouseholdsandbusinessesandbyissuingdemandabledepositsinexchangeforthesefunds;andbanksdiversifyawayidiosyncraticcreditriskbyholdingportfoliosofmultipleloans.1Muchoftheempiricalliteratureincommercialbankinghasfollowedtheserichtheoreticalleads,analyzingthefinancialflowsfundamentaltotheintermediationprocess(e.g.,interestpaidondeposits,interestreceivedfromloansandsecurities,andtheresultingnetinterestmargins)andtherisksassociatedwiththoseflows(e.g.,liquidityriskassociatedwithdeposits,creditriskassociatedwithloans,marketriskassociatedwithfixedincomesecurities,andinterest-rateriskassociatedwiththerelativematuritiesofdeposits,loans,andsecurities).However,commercialbankbusinessmodelshaveevolvedoverthepasttwodecades,andtodaybanksgenerateanincreasedportionoftheirincomefromnonintermediationand/ornoninterestactivi-ties.Forexample,between1980and2001,noninterestincomeintheU.S.commercialbankingsystemincreasedfrom0.77%to2.39%ofaggregatebankingindustryas-sets,andincreasedfrom20.31%to42.20%ofaggregatebankingindustryoperatingincome.Theincreasingpresenceofnoninterestincomeatcommercialbankshasbeenwidelydocumentedanddiscussedintheindustrypressandregulatorypublications(forexample,FeldmanandSchmidt,1999),butonlyafewacademicstudieshaveinvestigatedtheimpactofincreasednoninterestincomeonthefinancialperformanceofcommercialbanks.Whileitiswellknownthatlargebanksandbankswithspe-cializedstrategies(e.g.,creditcardbanks,mortgagebanks)relymoreheavilyonnoninterestincomethandosmallbankswithtraditionalbusinessstrategies,thereislittlesystematicunderstandingofwhynoninterestincomevariesacrossbanksandhownoninterestincomeisassociatedwithbankfinancialperformance.1SeminaltheoreticalstudiesinthisareaincludeGurleyandShaw(1960),Pyle(1971),BenstonandSmith(1976),LelandandPyle(1977),Fama(1980),DiamondandDybvig(1983),Diamond(1984),BoydandPrescott(1986),James(1987),andGortonandPennacchi(1990).SeeSaunders(2000,chapter6)andFreixasandRochet(1999,chapter2)forgeneraldiscussionsofwhybanksexistandoverviewsofthetheoreticalliterature.R.DeYoungandT.Rice/TheFinancialReview39(2004)101–127103Thispaperattemptstofillinsomeofthesegaps.InSection2wedocumentthelong-runtrendsintheamountandcompositionofnoninterestincomeatU.S.commercialbanks.InSection3wediscusstheregulatoryandtechnologicaldeter-minantsofnoninterestincomeatcommercialbanks,andconsiderwhynoninterestincomehasgrownmorequicklyatsomebanksthanatothers.InSection4wediscussthepotentialeffectsofincreasednoninterestincomeonthefinancialperformanceofcommercialbanks.Werefertotheexistingliteratureonnoninterestincomeatcommercialbanksthroughouteachofthesefirstthreesections.InSection5wespecifyaneconometricmodeldesignedtoanswertwobroadquestions:Whichbankcharacteristics,marketconditions,andtechnologicaldevelopmentsaremostcloselyassociatedwithincreasednoninterestincome?Isnoninterestincomeassociatedwithimpr