FinanceandEconomicsDiscussionSeries FederalReserveBoard,Washington,D.C.ISSN1936-2854(Print) ISSN2767-3898(Online) TheInformationalCentralityofBanks NathanFoley-Fisher,GaryGorton,St´ephaneVerani2024-006 Pleasecitethispaperas: Foley-Fisher,Nathan,GaryGorton,andSt´ephaneVerani(2024).“TheInformationalCen-tralityofBanks,”FinanceandEconomicsDiscussionSeries2024-006.Washington:BoardofGovernorsoftheFederalReserveSystem,https://doi.org/10.17016/FEDS.2024.006. NOTE:StaffworkingpapersintheFinanceandEconomicsDiscussionSeries(FEDS)arepreliminarymaterialscirculatedtostimulatediscussionandcriticalcomment.TheanalysisandconclusionssetfortharethoseoftheauthorsanddonotindicateconcurrencebyothermembersoftheresearchstaffortheBoardofGovernors.ReferencesinpublicationstotheFinanceandEconomicsDiscussionSeries(otherthanacknowledgement)shouldbeclearedwiththeauthor(s)toprotectthetentativecharacterofthesepapers. TheInformationalCentralityofBanks∗ NathanFoley-Fisher1,GaryGorton2,andStéphaneVerani11FederalReserveBoard 2YaleSchoolofManagementandNBER Firstversion:January2021;thisversion:December2023 Abstract Theequityanddebtpricesoflargenonbankfirmscontaininformationaboutthefuturestateofthebankingsystem.Inthissense,banksareinformationallycentral.Theamountofthisinformationvariesovertimeandoverequityanddebt.Duringafinancialcrisisbanksare,bydefinitionofacrisis,atriskoffailure.Debtpricesbecameabout50percentmoreinformativethanequitypricesaboutthefuturestateofthebankingsystemduringthefinancialcrisisof2007-2009.Thiswaspartlyduetoinvestors’fearsthatbanksmightnotbeabletorefinancethefirms’debt. JELCodes:D82,E44,G14 Keywords:priceinformativeness,assetpricing,bankingsystem,financialcrises ∗Forprovidingvaluablecomments,wewouldliketothank,withoutimplication,Elena Afanasyeva,CelsoBrunetti,EduardoDavila,BorghanNarajabad,AnnaOrlik,GeorgioOttonello,DinoPalazzo,MichaelPalumbo,CeciliaParlatore,CocoRamirez,SkanderVanDenHeuvel,participantsintheNBERSummerInstituteConferenceonCapitalMarketsandtheEconomy2023,theFederalReserveBoardMacro-FinanceWorkshop2023,andattheFederalReserveBankofPhiladelphia.Theviewsinthispaperaresolelytheauthors’andshouldnotbeinterpretedasreflectingtheviewsoftheBoardofGovernorsoftheFederalReserveSystemorofanyotherpersonassociatedwiththeFederalReserveSystem. 1Introduction Bankssitatthecenterofthesavings-investmentprocess.Butwhatdoesitmeanthatbanksareatthe“center”ofthesavings-investmentprocess?Toaddressthisquestion,weaskwhethertheequityanddebtpricesoflargenonbankfirmscontaininformationaboutthefuturestateofthebankingsystem(“thestateofthebankingsystem”).Welookatnormaltimesandduringthefinancialcrisisof2007-2009.Wefindthattheequityanddebtpricesoflargenonbankfirmsdoindeedembedinformationaboutthestateofthebankingsystem.Theamountofinformationembeddedinpricesvariesovertime,andoverequityanddebt.1 Alargeliteraturestudiesfinancialcrisesasinformationeventsinwhichshort-termdebttransitsfrombeinginformation-insensitivetoinformationsensitive–acrisis.ForareviewofthisliteratureseeDang,GortonandHolmström(2020).Afinancialcrisisisasystemicevent,thesolvencyoftheentirebankingsystemisthreatened.BenBernankemadethispointinhistestimonybeforetheFinancialCrisisInquiryCommission(2012).HesaidthatduringSeptemberandOctoberof2008“...outofthe13mostimportant financialinstitutionsintheUnitedStates,12wereatriskoffailurewithin 1Thereisalargeliteratureestablishingthatfirmoutcomesdependonconditionsinthebankingsector,includingbankfinancingconstraints,competition,andprofitability(Paravisini,2008;ClaessensandLaeven,2005;ChavaandPurnanandam,2011).Atthemicrolevel,individualbanksthattightentheirloansupplyhaverealeffectsonfirminvestmentandemploymentdecisions(Bassett,Chosak,DriscollandZakrajsek,2014;Chodorow-Reich,2014;Castro,Glancy,IonescuandMarchal,2022).Anearlierstrandofthesameliteratureusedaggregatebankdata,includingOwensandSchreft(1991)andLownandMorgan(2002,2006). aperiodofaweekortwo”(p.354).Weshowthatduringafinancialcrisis,thereisakindofinformationregimeswitchforcorporateassetsaswell.Wefindthatbothequityanddebtpricesarealwaysinformativeaboutthebankingsystem.Butduringthefinancialcrisisof2007-2009debtpriceswereabout50percentmoreinformativethanequityprices.Weshowthatthisdifferencewasinpartduetoinvestors’fearsthatbanksmightnotbeabletorefinancethefirms’debt. Weproceedbyestimatingpriceinformativenesscorrespondingtotherelativeprecisionofthesignalaboutfuturestatescontainedinassetprices.Inpractice,aspecificcombinationofR2statisticsfromlinearregressionsofchangesinassetpricesonchangesinstatesexactlyidentifiesrelativepriceinformativeness(DavilaandParlatore,2022).Inoursettingwestudywhetherasinglenonbankfirm’sassetpricesareinformativeabouttwounknownstates:thefirmstate—measuredasthefirm’sfutureearningsasinDavilaandParlatore(2022)—andthestateofthebankingsystem(definedbelow).Ourcalculationsareanalogoustoanexternalobserverupdatingherprioraboutthestateofthebankingsystemafterobservingchangesinanonbankfirm’sassetprices.Weshowthattheobservercanidentifytheinformationcontentaboutthestateofthebankingsystem.U