Financial innovation, new assets, and the

~i,mt~lffUELSEVIERJournalofBanking&Finance20(1996)207-225JournalofBANKING&FINANCEFinancialinnovation,newassets,andthebehaviorofmoneydemandDennisGlennonaJuliaLaneb,,aDepartmentofEconomicandPolicyAnalysis,OfficeoftheComptrolleroftheCurrency,Washington,DC20219,USAbDepartmentofEconomics,CollegeofArtsandSciences,TheAmericanUniversity,4400MassachusettsAvenueN.W.,Washington,DC20016,USAReceived15June1993;accepted15August1994AbstractTwofundamentalchangesinUSbankingregulationshaveaffectedthebehaviorofmoneydemand(MI).Thefirstauthorizedcheckabledepositaccountspayingexplicitinterestrates.Thesecondallowedtheseratestobemarketdetermined.Thetheoreticalliteraturedoesnotdirectlyaddresstheimpactoftheseevents,suggestingthattheyareprimarilyanempiricalissue.However,theempiricalliteraturehasyettoagreeontheimpactoffinancialinnovationonmoneydemand;forexample,severalstudiesreportanincreaseintheelasticityofmoneydemand,severalothersreportadecline.ThispaperusesaLancaster-typechoicemodeltoanalyzeformallytheexpectedimpactofthesetwochangesonthedemandformoney.Themodelderivesspecificconditionsunderwhich(i)thedemandformoneyincreasesasnewassetsareintroducedand(ii)theimpactofeithertheintroductionofnewassetsortheeliminationofinterestraterestrictionsontheelasticityofmoneydemand.JELclassification:E41;E52;E53Keywords:Moneydemand;Financialinnovation;Elasticity*Correspondingauthor.Tel.:[phone];fax[phone]-4266/96/$15.00©1996ElsevierScienceB.V.AllfightsreservedSSD!0378-4266(94)00130-8208D.Glennon,J.Lane/JournalofBanking&Finance20(1996)207-2251.IntroductionFinancialinnovationsinthe1970sandthe1980scoincidedwiththebreakdownintherelationshipbetweenmoney,incomeandinterestrates(GoldfeldandSichel,1990).Thetheoreticalliteraturedoesnotdirectlyaddressthebreakdown,suggest-ingthatitisprimarilyanempiricalissue.Infact,conventionaltheoryisnotwellsuitedforanalyzingtwoofthemoreimportantissuesassociatedwiththeimpactofderegulationonthedemandformoney:theemergenceofnew(monied)assetsandthephaseoutoftheinterestrestrictionsondeposits.Inparticular,inventorymodelsarelimitedtoananalysisofinnovationsthatcanbedirectlyassociatedwithchangesintransaction/brokercosts(Milboume,1986;MilboumeandMoore,1986).Liquiditypreferencemodels,ontheotherhand,areinappropriatefortheanalysisofinnovationsthatinvolvetheemergenceofnewmoniedassets,sincetheseassetsgenerallypossesssimilarriskcharacteristics(Changetal.,1983;Greene,1992).Inthispaperwedevelopatheoreticalapproachtomodellingassetdemandwhichisdesignedtoaddresstheimpactoffinancialinnovationonthedemandformoney.Themodel,anextensionofLancaster's(Lancaster,1971,1991)'modemconsumertheory'model,ismoreflexiblethanconventionalmodelssinceitpermitstheendogenousemergenceofnewassets.Itsuggeststhatfinancialinnovationand/orderegulation,suchastheemergenceofnewassetsorthephaseoutofRegulationQ,shouldresultinspecificstructuralshiftsintherelationshipbetweeninterestratesandmoneydemand.ThisresultisconsistentwithempiricalstudiesthatfindasignificantchangeinmoneydemandandcorrespondswellwiththemajorinnovationschronicledinHester(1981,1987)andEisenbeis(1985).Thepaperisstructuredasfollows.ThemodelisdevelopedinSection2.AnanalysisoftheexpectedimpactoffinancialinnovationonthemarketdemandformoneyisgiveninSection3.Section4discussestheconsistencyoftheresultsoveranalternativedefinitionofmoney.2.MonetarycharacteristicsandmonetaryassetsAmodelwhichseekstoincorporatetheimpactoftheintroductionofanewassetonthedemandforexistingassetsshouldbegeneralenoughtoallowfortheendogenousemergenceofnewassets.However,conventionalutilitymodelsrequireabandoningtheoldpreferencepattensandbeginninganewwithanexpandedassetspaceforeachnewassetthatemerges.TheLancasterapproachavoidsthisproblem(Lancaster,1991,1971).Thisapproachisparticularlywellsuitedforanalyzingmoneydemand,espe-ciallyduringaperiodofrapidfinancialinnovation.Moneycanbedescribedasanassetthatfulfillstheadditionalroleofamediumofexchange,andthisliquidity(spendability)rolesetsitapartfromotherassetsheldinaportfolio.Weaccord-D.Glennon,J.Lane/JournalofBanking&Finance20(1996)207-225209inglyassumethatwealthholders'demandformoneyisderivedfrommaximizingtheirutilitywithrespecttothecharacteristicscommonlyassociatedwithassets:liquidityandreturn(Milbourne,1986;DesaiandLow,1987;Fisher,1978).zThisassumptionisusedtomodelindividualportfoliochoiceandisusedasafounda-tionuponwhichtoderiveamarketdemandcurve.2.1.individualportfoliodecisionsThewealthholder'sproblem(Lancaster,1971)istomaximizeutilitybasedonacompositionofcharacteristics(Z)satisfyingtheprimaryportfoliofunctionsofmoney:MaxU(Z)s.t.Z~K,whereKisthefeasiblesetofcharacteristicsobtainablefromholdingvariouscombinationsofassets.Thecombinationofassetswhichisnecessarytoobtaintheutilitymaximizingcharacteristicsmixisderivedfromthemappingofthefeasiblesetinassetspaceintocharacteristicsspace.Thefollowingassumptionsaremadetofacilitatethediscussion:1.Thereareonlytwocharacteristics:liquidity(r~)andprofitability(~'2),thatwealthholdersseekinanassettobeheldintheirportfolio(DesaiandLow,1987;Fisher,1978).2.Therearethr