Innovation,Growth,andAssetPrices_金融投资_经管

THEJOURNALOFFINANCE•VOL.LXX,NO.3•JUNE2015Innovation,Growth,andAssetPricesHOWARDKUNGandLUKASSCHMID∗ABSTRACTWeexaminetheassetpricingimplicationsofaproductioneconomywhoselong-termgrowthprospectsareendogenouslydeterminedbyinnovationandR&D.Inequilib-rium,R&Dendogenouslydrivesasmall,persistentcomponentinproductivitythatgenerateslong-rununcertaintyabouteconomicgrowth.Withrecursivepreferences,householdsfearthatpersistentdownturnsineconomicgrowthareaccompaniedbylowassetvaluationsandcommandhigh-riskpremiainassetmarkets.Empirically,wefindsubstantialevidenceforinnovation-drivenlow-frequencymovementsinag-gregategrowthratesandassetmarketvaluations.Inshort,equilibriumgrowthisrisky.ANECONOMY’SLONG-TERMGROWTHprospectsreflectitsinnovativepotential.Atafundamentallevel,innovationisakeysourceofsustainedgrowthinaggregateproductivity.Empiricalmeasuresofinnovation,suchasresearchanddevelopment(R&D)expenditures,tendtobevolatileandquitepersistent.Suchmovementsaffectthedynamicsofgrowth.Indeed,inU.S.post-wardata,productivitygrowthexhibitslongandpersistentswings.1Similarly,innovation-drivengrowthwavesassociatedwiththearrivalofnewtechnologies,suchastelecommunication,computers,andtheinternet,tonameafew,arewelldocumented.2Stockpricesreflectsuchchangesingrowthprospects.Moreover,ifagentsfearthatapersistentslowdownineconomicgrowthwilllowerassetprices,thesemovementswillgiverisetohigh-riskpremiainassetmarkets.Inthispaper,wedevelopageneralequilibriummodelofinnovationandR&Dtolinkassetpricesandaggregateriskpremiatoendogenousmovements∗HowardKungisattheLondonBusinessSchool.LukasSchmidisatDukeUniversityandtheUniversityofCalifornia,LosAngeles.WethanktheActingEditorDaveBackusandtwoanonymousrefereesaswellasHengjieAi,RaviBansal,GianLucaClementi,JohnColeman,DiegoComin,MaxCroce,IanDew-Becker,BernardDumas,LorenzoGarlappi,Jo˜aoGomes,FrancoisGourio,LeonidKogan,LarsLochstoer,PascalMaenhout,StavrosPanageas,DimitrisPapanikolaou,Adri-anoRampini,NormanSchurhoff,TomTallarini,AmirYaron,andLuZhangforhelpfuldiscussions.WehavealsobenefitedfromthecommentsofseminarparticipantsatBocconi,BostonUniversity,CarnegieMellonUniversity,CornellUniversity,DukeUniversity,FederalReserveBankofSanFrancisco,ImperialCollege,INSEAD,LondonBusinessSchool,LondonSchoolofEconomics,MIT,ParisSchoolofEconomics,UCLA,andUniversitatPompeuFabra,aswellasfromconferencepar-ticipantsatWesternFinanceAssociation,SocietyforEconomicDynamics,CEPRMacroFinanceConferenceLondon,EuropeanFinanceAssociation,TelAvivFinanceConference,EconometricSociety,AmericanFinanceAssociation,andNBERSpringAssetPricingMeeting.1See,forexample,Gordon(2010)andJermannandQuadrini(2004).2Forexample,seeHelpman(1998)andJovanovicandRousseau(2005).DOI:10.1111/jofi.1224110011002TheJournalofFinanceRinlong-termgrowthprospects.Oursetuphastwodistinguishingfeatures.First,weembedastochasticmodelofendogenousgrowthbasedonindustrialinnovation3intoanotherwisestandardproductioneconomy.Inthismodel,pro-ductivitygrowthisendogenousandsustainedbythecreationofnewpatentedtechnologiesthroughR&D.Patentsrepresentanendogenousstockofintangi-blecapital.Second,weassumethathouseholdshaverecursivepreferences,sothattheycareaboutuncertaintyregardinglong-termgrowthprospects.Whencalibratedtomatchempiricalevidenceoninnovationandlong-runeconomicgrowth,ourmodelcanquantitativelyreplicatekeyfeaturesofassetreturnsinthedata.Inparticular,ourmodelrationalizesasizeableequitypre-miumandalowandstablerisk-freeinterestrate.Moreover,ourmodelpredictsasizeablespreadbetweenthereturnsonphysicalcapitalandintangiblecap-ital,whichisrelatedtothevaluepremiuminthedata.Inshort,wefindthatequilibriumgrowthisrisky.WefirstshowthatinthemodelinnovationandR&Dendogenouslydriveasmallbutpersistentcomponentinthegrowthrateofproductivity.Inourgen-eralequilibriumsetting,theselow-frequencymovementsinproductivitytrig-gerlongandpersistentswingsinaggregategrowthrates,suchasconsumptionandoutput,whichwelabelgrowthcycles.Intuitively,shocksaffecttheincen-tivestoinnovate,whichinturnimpactlong-termgrowthprospects.Notably,transitoryshocksinthissettinghavelong-lastingpermanenteffectsthroughtheinnovationchannelandgenerateendogenouspersistenceingrowthrates.Thus,abadtemporaryshocknotonlylowersthelevelofconsumptionandcashflowstoday,butalsodepresseslong-termgrowthrates.Whenagentshaverecursivepreferences,theyaresensitivetobothshort-runandlong-rununcer-taintyaboutconsumptiongrowth.Growthcycleshelprationalizesizeableriskpremiainassetmarkets,asagentsfearthatsuchprolongedslumpsineconomicgrowthcoincidewithlowassetvaluations.Similarly,agentssaveforextendedlowgrowthepisodes,drivingdowntherealinterestrate.Furthermore,inthemodel,physicalcapitalisendogenouslymoreexposedtopredictablevariationingrowththanintangiblecapital,whichgeneratesasizeablevaluespread.Aninnovation-drivenpersistentcomponentinproductivitygrowthprovidesanequilibriumfoundationoflong-runrisksinthespiritofBansalandYaron(2004).Moreprecisely,inourmodel,long-runproductivityrisks,inthesenseofCroce(2014),arisenaturallyinequilibrium.Furthermore,persistentmove-mentsinexpectedproductivityaff