《国际清算银行-通货膨胀对家庭资产负债表的异质性影响(英)-2023.12_市场营销策划_重点报告20.docx》由会员分享,可在线阅读,更多相关《国际清算银行-通货膨胀对家庭资产负债表的异质性影响(英)-2023.12_市场营销策划_重点报告20.docx(38页珍藏版)》请在三一办公上搜索。
1、OBISBISWorkingPapersNo1152TheHeterogeneousImpactOfInfIationonHouseholds7BalanceSheetsbyClodomiroFerreira,JoseMiguelLeiva,GaloNuho,AlvaroOrtiz,TomasaRodrigo,andSireniaVazquezMonetaryandEconomicDepartmentDecember2023JELclassification:G51,D31,E31Keywords:inflationinequality,netnominalpositions,nominalw
2、agerigiditiesBISWorkingPapersarewrittenbymembersoftheMonetaryandEconomicDepartmentoftheBankforInternationalSettlements,andfromtimetotimebyothereconomists,andarepublishedbytheBank.Thepapersareonsubjectsoftopicalinterestandaretechnicalincharacter.Theviewsexpressedinthemarethoseoftheirauthorsandnotnece
3、ssarilytheviewsoftheBIS.ThispublicationisavailableontheBISwebsite(www.bis.org).BankforInternationalSettlements2023.Allrightsreserved.Briefexcerptsmaybereproducedortranslatedprovidedthesourceisstated.ISSN1020-0959(print)ISSN1682-7678(online)TheHeterogeneousImpactofInflationonHouseholds7BalanceSheetsW
4、e are especially grateful to Klaus Adamz Riccardo Cioffiz Cesaire Mehz Gonzalo Paz-Pardo, Anna Rogantini-Piccoz and GilIes Saint-Paul for their suggestions/ as well as participants at several seminars and conferences. The opinions and analysis are the responsibility of the authors, and therefore, do
5、 not necessarily coincide with those of BBVA, Banco de Espana, the BIS or the Eurosystem. All errors are our own.ClodomiroFerreira2,JoseMiguelLeivaSee Ross Sorkin (2022).,GaloNuno2z3,4,AlvaroOrtiz1,TomasaRodrigo1landSireniaVazquez11BBVAResearch2BancodeEspana3BIS4CEMFIDecember2023AbstractWeidentifyan
6、dstudyanalyticallythreekeychannelsthatshapehowinflationaffectswealthinequality:(i)thetraditionalwealth(Fisher)channelthroughwhichinflationredistributesfromlenderstoborrowers;(ii)anincomechannelthroughwhichinflationreducestherealvalueofstickywagesandbenefits;and(iii)arelativeconsumptionchannelthrough
7、whichheterogeneousincreasesinthepriceofdifferentgoodsaffectpeopledifferentlydependingontheirconsumptionbaskets.Wethenquantifythesechannelsduringthe2021inflationsurgeinSpainusingdetailedandhigh-frequencyclient-leveldatafromoneofthemaincommercialbanks.Theunexpectednatureandtemporaryperceptionoftheinfl
8、ationshockinthisparticularperiodcloselymapsontotheassumptionsbehindourtheoreticaldecomposition.Resultsshowthatthewealthandincomechannelsareoneorderofmagnitudelargerthantheconsumptionchannel.Middle-agedindividualswerelargelyunaffectedbyinflation,whileolderonessufferedthemost.Wefindsimilarresultswhenu
9、singrepresentativesurveysonhouseholdszwealth,income,andconsumption.Keyzuords:inflationinequality,netnominalpositions,nominalwagerigidities.JELclassification:G51,D31,E31.1 IntroductionBenBemanke,formerChairmanofFederalReserve,statedthatzzThedifferencebetweeninflationandunemploymentisthatinflationaffe
10、ctsjusteverybody.Inflationhasasocial-widekindofimpact.IWhileinflationmayaffecteveryone,itdoesnotaffecteveryoneinthesameway.Differencesinwealthcomposition,salaryorconsumptionpatternsmayleadtoquitedifferentoutcomesfordifferentindividuals.Theaimofthispaperistocastsomelightonthedifferentchannelsthroughw
11、hichinflationaffectswealthinequality.Thepaperleveragestheunexpectednatureofthe2021surgeinaggregateinflationinSpain(andtheEuro-Areamorebroadly),aswellasthebroadperceptionofitsshort-liveddynamics,inordertoanalyticallycharacterizeshowashocktoinflationimpactsapersonswealthintheshortrun.Wederiveananalyti
12、calexpressiondecomposingtheimpactofinflationonindividualwealthintothreedifferentchannels,inthespiritofthesufficientstatisticsapproachofAuclert(2019).Tothisend,weconsiderasurpriseone-offincreaseingdsandservicesprices,whichisheterogeneousacrosssectorsbutdoesnottriggeraresponseofassetpricesorinterestra
13、tesonimpact.First,thezzwealth(Fisher)channel“captureshowinflationredistributesrealwealthfromlenderstoborrowersbychangingthevalueofnominalassetsandliabilities.ThischannelhasalreadybeenstudiedbyDoepkeandSchncider(2006),Mehetal.(2010)AdamandZhu(2016),Auclert(2019),Caoetal.(2021)orPallotti(2022).Theimpa
14、ctofinflationonwealthisfullycapturedbythenetnominalposition(NNP),definedasthedifferencebetweennominalassetsandnominalliabilities.Second,inflationreducestherealvalueofnominalincome,inwhatwecallthezyincomechannel,z.Nominalincomesourcessuchaswages,pensions,orunemploymentbenefitsareoftensticky.WagesinSp
15、ain,forinstance,aretypicallyupdatedatannualfrequencyorevenlowerthanthat.1This is in line with the evidence presented by Bihan et al. (2012) and Barattieri et al. (2014)z who find that the pattern of nominal wage changes appears to be in line with the staggered contracting model of Taylor (1980), wit
16、h the hazard for wage contracts peaking at 12 months both in France and the US.Whileinflationreducestherealincomeandwealthofallagentsinthesameproportionrelativetotheirpre-shocklevels,theabsoluteimpactwillnaturallybehigherthehigherthenominalincomeis.Third,inflationdoesnottypicallyaffectallpriceshomog
17、eneously.Thepricesofsomegoodsorservicesrisemorethanthoseofothers.GiventhatindividualsconsumedifferentbasketsOfgoodsandservices,anincreaseinpricesthatisheterogeneouslydistributedacrossthesegoodswillimpactagentsdifferently.Wedemonstratehowthis/zrelativeconsumptionChanneriSproportionaltoapersonsconsump
18、tionexpendituremultipliedbytheratiobetweentheindividualinflationrate,computedusingthebasketofeachparticularindividual,andtheeconomy-wideinflationrate,computedusingthebasketoftheaverageconsumer.Thisconnectsthischanneltotheemergingliteratureonzzinflationinequality”(HobijnandLagakos,2005;KaplanandSchul
19、hofer-Wohl,2017;ArgenteandLee,2021;Jaravel,2018;Jaravel,2021)zwhichanalyzeshowdifferentindividualsexperiencedifferentinflationrates.Forthoseagentswhoconsumemoreofthegoodsandservicesthatexperiencethelargestpriceincreases,thatisthoseagentswithhigherindividualinflationrates,thischannelwillleadtoareduct
20、ionintheirwealth,astheyneedtodevotealargershareofresourcestopayfortheirconsumptionbasket.Conversely,thoseagentsexperiencinglowerindividualinflationrateswillenjoyanincreaseintheirwealth,astheyneedtodevotelessresourcestopayfortheirconsumption.Next,weturntotheempiricalquantificationofthesethreechannels
21、withinthecontextofthe2021inflationsurgeinSpain.Threefeaturesofthisparticularepisoderenderitanalmostidealeventforourestimationexercise.First,thesurgewascompletelyunexpectedasevidentfromdifferentmeasuresofinflationexpectationsattheendof2020.Second,evenwhenthesharpincreaseininflationwaswellunderway,itw
22、as(incorrectly)perceivedtobeveryshort-lived.Third,thesurgearrivedafteralongperiodofverylowinflationrates,makingitmorelikelythathouseholds,firms,unions,andotheragentswere,atleastduringthefirstphase,stillinattentiveandfacingsignificantcostsandinertiatoadjusttheirbehavior?Forourquantification,weemploya
23、novelproprietarydatasetfromBancoBilbaoVizcayaArgentariaSA(BBVA).BBVAisalargeglobalbankwithastrongpresenceinSpain.Thedatasetincludestheuniverseofbankaccounts,withinformationaboutdepositsandcurrentaccounts,aswellasmortgages,consumerloans,creditandpre-paidcards.Italsoincludestheuniverseofcardtransactio
24、nscollectedfromBBVAcardholdersandBBVA-operatedpoint-of-saleinSpain,originallycollectedinCarvalhoetal.(2021)(andfurtherexpandedinBudaetal.z2022),togetherwithothermeansofpaymentsuchascash,directdebitandtransfers,whichallowsustoaccountformonthlypaymentstoutilities(i.e.electricity,water,gas,andtelecommu
25、nication.).See, for example, Reis (2006).The advantages of using actual transaction and accounts data instead of representative surveys in our setting are threefold: first, and as discussed in Kaplan and Schulhofer-Wohl (2017)z actual transactions and income do not suffer from a problem of miss-repo
26、rting, or lack of memory. Second, unlike both representative surveys available for the purpose of comparison, our client data contains both transaction (i.e. consumption spending) as well as income and detailed balance sheet information for the same individual. Third, client transactions and balance
27、 sheet positions can be observed at a weekly frequency, while the relevant surveys are released with an annual or lower frequency.Threemainresultsemerge.First,boththewealthandtheincomechannelsare,onaverage,oneorderofmagnitudelargerthantherelativeconsumptionchannel.Thisimpliesthatheterogeneityinconsu
28、mptionbasketsplaysasmallerrolethanthesetwo,traditional,channels.Thereasonisthatthedispersioninindividualinflationratesacrossthepopulationin2021wasnotlargeenoughtogeneratesignificantlossesorgains.Second,themagnitudesoftheincomeandwealthchannelsareequivalentinabsolutevalue,thoughtheincomechannelleadst
29、oinflationreducingrealwealthofallhouseholds,whereasthewealthchannelincreasestherealwealthofdebtorswhilereducingthatofcreditors.Theresultisthatmiddle-agedindividuals,whohavelargenegativeNNPsduetomortgages,wereroughlyunaffected,orevenbenefited,bythe2021inflation,whileoldpeopleexperiencedthelargestdecl
30、ineinrealwealth,astheyhavelargepositiveNNPs.Reassuringly,thesetworesultsareconfirmedifweusedatafromtworepresentativesurveys,publiclyavailable,thathavebeenextensivelyemployedbothinthepolicyandresearchdomains:theconsumerexpendituresurvey(theEncuestadePresupuestosFatniliares(EPF),andtherepresentativeco
31、nsumerfinancesurvey(theEncuestaFinancieradelasFarnilias(EFF).The EPF is a comprehensive household expenditure survey carried out with an annual frequency by the Spanish National Statistical Institute (INE). The EFF instead is a representative household survey collecting detailed information on house
32、holds balance sheets.Third,despitetheincreaseininflation,WeobservenosignificantchangeineitherNNPorincomefordifferentagegroupsfromend-2020toend-2022.Individualinflation,however,displayedhighervariabilityacrossagegroupsin2022,whichsuggeststhattherelativeconsumptionchannelmayhaveplayedalargerroleinthat
33、year.We do not recompute aggregate results for 2022 as the conditions for the decomposition are not satisfied in 2022z with inflation being now both expected and perceived to be more persistent.Theapproachwefollowinourpaperisquitestraightforward:itisbasedonaverygeneralsetofassumptions.Whilethisisapp
34、ealingasafirstanswertothisimportantquestion,anumberofrecentworksgobeyondoursimpleframeworkinanumberofimportantdimensions.Within the policy domain, the International Monetary Fund (IMF (2023) has recently applied our framework to a wide sample of countries, including Kenya, Mexico, Senegal, Colombia,
35、 Finland, and France, showing how the estimates of the different channels, and their relative importance, areFirst,Cantoetal.(2023)studymoney-metricwelfaremovements,ratherthanwealthinequality,andcombinecross-sectionalandtimeseriesdatatoestimatetheresponsetoidentifiedshocks,whichallowsthemtoassesswhe
36、therdifferentshockstoinflationcarrydifferentimpacts.Second,Pallottietal.(2023)measurewelfareeffectsforseveralcountriesintheEuroarea,whilealsoconsideringrents,dividends,andcapitalgains.Ontopofpartialequilibriumeffects,theyimposefurtherassumptionswhichallowthemtoanalyzegeneralequilibriumeffects.Theyfi
37、ndthatgeneralequilibriumeffectsareanorderofmagnitudesmallerthanpartialequilibriumones.WeviewourpaperasabuildingblocktoCantoetal.(2023)andPallottietal.(2023).Weimposeaminimumsetofassumptions,carefullymappedtoaparticularevent,inordertoquantifythecontemporaneouseffectofunexpectedinflationonindividualsr
38、ealwealth.Importantly,giventheempiricallysoundassumptions,ourestimatedeffectscanbeinterpretedasemergingfromgeneralequilibrium.Thelengthoftheperiodforwhichthisisavalidapproximationdepends,crucially,onthedynamicsofexpectationsandhowtheseshapebehavioralresponsesofhouseholdsandfirms(intermsofconsumption
39、baskets,portfoliocomposition,etc.),aswellasofthefiscalandmonetaryauthorities.Tracingoutthedynamicimpactofdifferent(andsizeable)inflationaryshocksoncetheseadjustmentsstarttotakeplace,isoutsidethescopeofthispaper.Doingsorequiresimposingmorestructureinordertocharacterizehowequilibriumpricesaredetermine
40、d,aswellastheroleofpolicyinstrumentsinshapingthedistributionalimpactofinflation.Thispapercontributestotheemergingliteratureanalyzingtheheterogeneousconsequencesofinflationacrossthepopulation.Inadditiontotheempiricalreferencesalreadycited,anumberofrecentpapershaveanalyzedthere-distributiveeffectsofmo
41、netarypolicyingeneral-equilibriummodelswithheterogeneousagentsthroughtheFisherchannel(seeNunoandThomas,2022orFerranteandPaustian,2019zamongothers),wagestickiness(Hagedornetal.z2019zAuclertetal.z2020)andtherelativeconsumptionchannel(Cravinoetal.z2020;Olivietal.z2023).Wecontributetothisliteraturebypre
42、sentingatheoreticalframeworkthatallowsustoquantifytherelativeimportanceofthedifferentchannelsinthedata.Inarecentpaper,Yang(2023)combinesthesethreechannelsinaheterogeneous-agentNewKeynesianmodelandanalyzesoptimalmonetarypolicy.contingenttotheeconomicstructureofeachcountry.2 Theunexpectedincreaseofinf
43、lationin2021Inthissection,weanalyzethesurgeininflationobservedintheyear2021.Wefocusonthisyearfortlreereasons.First,inflationexperiencedalargeincrease,quiteheterogeneousacrossgoodsandservices:itwasconcentratedonfoodandenergyitems.Second,thisincreasewas,toalargeextent,unexpected:neitherhouseholds,fina
44、ncialmarketsnorprofessionalforecastershadanticipateditbytheendof2020,oreventhemiddleof2021.Third,theincreasewas(wrongly)perceivedtobeshort-lived,quicklyrevertingtorelativelylowinflationlevelsin2022.Thesethreereasonsmaketheepisodein2021idealtostudytheimpactoflarge,temporaryinflationshocksonhouseholdb
45、alancesheets.Inflationsurgedin2021.Figure1presentstheevolutionoftheHICPgeneralinflation,togetherwithbroadcategories,inSpainsinceJanuary2020.Inflationsurgedin2021,withenergyandfooditemsexperiencingalargerincreasethanservicesormanufactured(industrial)goods.ThisinflationrateismuchlargerthanthatinDecemb
46、er2020(-0.6%)oreveninJune2021(2.5%),reflectingthesurgeininflationexperiencedovertheyear2021.WethencollectpriceindicesandinflationfiguresfordifferentspendingcategoriesdirectlyfromtheNationalStatisticalInstituteinSpain(INE).Focusingon2021,Column(a)ofTable2inappendixAshowstheannualgrowthrateoftheHarmon
47、izedIndexofConsumerPrices(HICP)inSpaininDecember2021,whichwas6.6%.Theincreaseinpriceswasstronglyasymmetric.Ontheonehand,housing,water,electricitygasandotherfuelsincreasemorethan20%andtransportationmorethan10%,reflectingthelargeriseinthepriceofoilandgasaftertheCovidcrisis.Ontheotherhand,communication
48、sdeclinedby-0.3%andclothingandfootwearincreasedonly0.7%.Column(b)displaystheweightsin2021employedtocomputeaggregateinflation,whichINEcomputesusingEPFresponses.Foodandnon-alcoholicbeverages,housingandtransportrepresentalmosthalfoftheconsumptionexpendituresoftheaverageindividual.Thesurgeininflationwas
49、largelyunanticipated.Table1displaysdifferentindicatorsofinflationexpectationsforSpain.TheseincludetheSUrVeyofProfessionalForecasters(SPF)ztheECBMacroeconomicProjections,theECBConsumerExpectationsSurvey(CES)andtheinstantaneousforwardratesderivedfrominflation-linkedswaps(ILS).Thefirsttwoindicators,theSPFandECBprojections,forecastSpanishinflation.TheSPFisasurveyofbanksandeconomicinstitutions.TheECBprojectionsarebasedondifferenteconometricmodels.TheCESisasurveyof