房价，借用限制和在经济周期的货币政策(1)

HousePrices,BorrowingConstraintsandMonetaryPolicyintheBusinessCycleMatteoIacoviello∗BostonCollegeDecember3,2004AbstractIdevelopandestimateamonetarybusinesscyclemodelwithnominalloansandcollateralconstraintstiedtohousingvalues.Demandshocksmovetogetherhousingandnominalprices,andareamplifiedandpropagatedovertime.Thefinancialacceleratorisnotuniform:nominaldebtdampenssupplyshocks,stabilizingtheeconomyunderinterestratecontrol.Structuralestimationsupportstwokeymodelfeatures:collateraleffectsdramaticallyimprovetheresponseofaggregatedemandtohousepricesshocks;nominaldebtimprovesthesluggishresponseofoutputtoinflationsurprises.Finally,policyevaluationconsiderstheroleofhousepricesanddebtindexationinaffectingmonetarypolicytrade-offs.(JELE31,E32,E44,E52,R21)∗DepartmentofEconomics,BostonCollege,ChestnutHill,MA02467,USA(email:[e-mail]).IamdeeplyindebtedtomyPh.D.advisorattheLondonSchoolofEconomics,NobuhiroKiyotaki,forhiscontinuoushelpandinvaluableadvice.IthankFabioCanova,RaffaellaGiacomini,ChristopherHouse,PeterIreland,RaoulMinetti,FrançoisOrtalo-Magné,MarinaPavan,ChristopherPissarides,FabioSchiantarelli,twoanonymousrefereesandseminarparticipantsattheBankofEngland,BostonCollege,theEuropeanCentralBank,theEnteLuigiEinaudi,theFederalReserveBankofNewYork,theFederalReserveBankofSt.Louis,theLondonSchoolofEconomics,theNBERMonetaryEconomicsMeetingandNortheasternUniversityfortheirhelpfulcommentsonvariousversionsofthiswork.ViktorsStebunovsprovidedsuperbresearchassistance.“Thepopulationisnotdistributedbetweendebtorsandcreditorsrandomly.Debtorshaveborrowedforgoodreasons,mostofwhichindicateahighmarginalpropensitytospendfromwealthorfromcurrentincomeorfromanyotherliquidresourcestheycancommand.Typicallytheirindebtednessisrationedbylenders[...].Businessborrowerstypicallyhaveastrongpropensitytoholdphysicalcapital[...].Theirdesiredportfolioscontainmorecapitalthantheirnetworth[...].Householddebtorsarefrequentlyyoungfamiliesacquiringhomesandfurnishingsbeforetheyearnincomestopayforthemoutright;giventhedifficultyofborrowingagainstfuturewages,theyareliquidity-constrainedandhaveahighmarginalpropensitytoconsume”-JamesTobin(1980),AssetAccumulationandEconomicActivity:ReflectionsonContemporaryMacroeconomicTheoryAlongtraditionineconomics,startingwithIrvingFisher’s(1933)debt-deflationexplanationoftheGreatDepression,considersfinancialfactorsaskeyelementsofbusinesscycles.Inthisview,deterioratingcreditmarketconditions,likegrowingdebtburdensandfallingassetprices,arenotjustpassivereflectionsofadecliningeconomy,butarethemselvesamajorfactordepressingeconomicactivity.Althoughthis“creditview”hasalonghistory,mostoftheoreticalworkonthissubjecthasbeenpartialequilibriuminnatureuntilthelate1980s,whenBenBernankeandMarkGertler(1989)formalizedtheseideasinageneralequilibriumframework.Followingtheirwork,variousauthorshavepresenteddynamicmodelsinwhichfinancingfrictionsonthefirmsidemayamplifyorpropagateoutputfluctuationsinresponsetoaggregatedisturbances:examplesincludetherealmodelsofNobuhiroKiyotakiandJohnMoore(1997)andCharlesCarlstromandTimothyFuerst(1997),andthesticky-pricemodelofBernanke,GertlerandSimonGilchrist(1999).Empirically,variousstudieshaveshownthatfirms’investmentdecisionsaresensitivetovariousmeasuresoffirms’networth(seeGlennHubbard,1998,forareview).Atthesametime,evidenceoffinancingconstraintsatthehouseholdlevelhasbeenwidelydocumentedbyStephenZeldes(1989),TullioJappelliandMarcoPagano(1989),JohnCampbellandGregoryMankiw2(1989)andChristopherCarrollandWendyDunn(1997).Whilethesestudieshavehighlightedtheimportanceoffinancialfactorsformacroeconomicfluctua-tions,todatetherehasbeennosystematicevaluationoftheextenttowhichageneralequilibriummodelwithfinancialfrictionscanexplaintheaggregatetime-seriesevidenceontheonehand,andbeusedformonetarypolicyanalysisontheother.Thisistheperspectiveadoptedhere.Fromthemodelingpointofview,mystartingpointisavariantoftheBernanke,GertlerandGilchrist(1999)new-Keynesiansetupinwhichendogenousvariationsinthebalancesheetofthefirmsgeneratea“financialaccelerator”byenhancingtheamplitudeofbusinesscycles.Tothisframework,Iaddtwomainfeatures:(1)collateralconstraintstiedtorealestatevaluesforfirms,asinKiyotakiandMoore(1997),andforasubsetofthehouseholds;(2)nominaldebt.Thereasonforhousing1collateralispracticalandsubstantial:practi-calbecause,empirically,alargeproportionofborrowingissecuredbyrealestate;substantialbecause,althoughhousingmarketsseemtoplayaroleinbusinessfluctuations,2thechannelsbywhichtheyaffecttheeconomyarefarfrombeingunderstood.Thereasonforhavingnominaldebtcomesfromthewidespreadobservationthat,inlowinflationcountries,almostalldebtcontractsareinnominalterms,eveniftheyappearhardtojustifyonwelfare-theoreticgrounds:understandingtheirimplicationsformacroeconomicoutcomesisthereforeacrucialtask.Inaddition,Iaskwhetherthemodelisabletoexplainbothkeybusinesscyclefactsandtheinteractionbetweenassetpricesandeconomicactivity.Tothisend,IestimatethekeystructuralparametersbyminimizingthedistancebetweentheimpulseresponsesimpliedbythemodelandthosegeneratedbyanunrestrictedVectorAutoregre