BigTech和貨幣政策傳導：基于中國的微觀經(jīng)濟層面證據(jù)

BISWorkingPapers

No1084

BigTechcreditand

monetarypolicytransmission:micro-levelevidencefromChina

byYipingHuang,XiangLi,HanQiuandChanghuaYu

MonetaryandEconomicDepartment

March2023

JELclassification:E52;G21;G23.

Keywords:Financialtechnology,banklending,monetarypolicytransmission.

ISSN1020-0959(print)

ISSN1682-7678(online)

BISWorkingPapersarewrittenbymembersoftheMonetaryandEconomicDepartmentoftheBankforInternationalSettlements,andfromtimetotimebyothereconomists,andarepublishedbytheBank.Thepapersareonsubjectsoftopicalinterestandaretechnicalincharacter.TheviewsexpressedinthemarethoseoftheirauthorsandnotnecessarilytheviewsoftheBIS.

ThispublicationisavailableontheBISwebsite().

?BankforInternationalSettlements2023.Allrightsreserved.Briefexcerptsmaybe

reproducedortranslatedprovidedthesourceisstated.

1

BigTechCreditandMonetaryPolicyTransmission:Micro-LevelEvidencefromChina*

YipingHuangtXiangLi主HanQiu§ChanghuaYu聞

Abstract

ThispaperstudiesmonetarypolicytransmissionthroughBigTechandtraditionalbanks.BycomparingbusinessloansmadebyaBigTechbankwiththosemadebytraditionalbanks,it?ndsthatBigTechcreditampli?esmonetarypolicytransmissionmainlythroughtheex-tensivemargin.Speci?cally,theBigTechbankismorelikelytograntcredittonewborrow-

erscomparedwithconventionalbanksinresponsetoexpansionarymonetarypolicy.The

*Forcomments,discussion,andsuggestions,wethankGeneAmbrocio,GuidoAscari(discussant),ChristophBasten,ChristianeBaumeister,JonathanBenchimol(discussant),SallyChen(discussant),MarcoDiMaggio(discussant),ThomasDrechsel,ZuzanaFung′aˇcov′a,AndreasFuster,LeonardoGambacorta(dis-cussant),EmiliaGarcia-Appendini,AlexandraGutsch,JiayinHu(discussant),YiHuang,BoreumKwak,WeiLi(discussant),ChangMa,AakritiMathur,MrinalMishra,StevenOngena,MelinaPapoutsi,MalteRieth,MatthiasRottner,AlessandroSardone,ChristophSchult,LauraSolanko,IlhyockShim,RubenSta?a,GregorvonSchweinitz,YongjieZhang,andotherscholarsattheChinaFinancialResearchConference;theAsianFAConference;theBankofEnglandworkshoponAdvancedAnalytics:NewMethodsandApplicationsforMacroeconomicPolicy,theCentralBankResearchAssociation2022AnnualMeeting,the9thAnnualConferenceofMITGolubCenterforFinanceandPolicy,BISresearchmeeting,theECBChinaExpertNetworkWorkshop2022;andseminarsattheUniversityofZurich,BankofFinland,andHalleInstituteforEconomicResearch.Anyremainingerrorsareoursalone.

tChinaCenterforEconomicResearch,NationalSchoolofDevelopment,andInstituteofDigitalFinance,PekingUniversity.YiheyuanRoad5,Beijing,100871,China.Workphonenumber:+86106275-4798.Email:

yhuang@

主HalleInstituteforEconomicResearch,Martin-Luther-UniversityHalle-Wittenberg,andInstituteofDigitalFinance,PekingUniversity.KleineMaekerstrasse8,Halle(Saale),06108,Germany.Workphonenumber:+493457753-805.Email:

xiang.li@iwh-halle.de

sBankforInternationalSettlements.78th?oor,TwoInternationalFinanceCentre,8FinanceStreet,Central,HongKong.Workphonenumber:+8522982-7100.Email:

han.qiu@

"ChinaCenterforEconomicResearch,NationalSchoolofDevelopment,andInstituteofDigitalFinance,PekingUniversity.YiheyuanRoad5,Beijing,100871,China.Workphonenumber:+86106275-8935.Email:

changhuayu@

BigTechbank’sadvantagesininformation,monitoring,andriskmanagementarethepo-tentialmechanisms.Inaddition,theusageofBigTechcreditisassociatedwithastrongerresponseof?rms’salesinresponsetomonetarypolicy.

Keywords:FinancialTechnology;BankLending;MonetaryPolicyTransmissionJELCodes:E52;G21;G23

1

1Introduction

Financialtechnology(FinTech)hasbeenamajorphenomenonintherecentdevelopmentof?nancialmarkets.DuringtheCOVID-19crisis,FinTechhasplayedanunprecedentedlyprominentroleinstabilizingandreignitingtheeconomy(

CoreandDeMarco

2021

,

Kwan

etal.

2021

,

BaoandHuang

2021

,

FuandMishra

2021

).Byde?nition,FinTechisabroadconceptthatreferstotheuseoftechnologyinproviding?nancialservices(

FSB

2019

).Whatmakesitstandoutinthelonghistoryof?nancialinnovationisthatthedisruptionthistimehasbeeninitiatedbyplayersoutsidethe?nancialmarketsratherthanwithintheoldsystem.Digitalplatformsformarketplacelendingandcreditissuedbybigtechnologycompanies(BigTech),suchasAntGroup,Amazon,orMercadoLibre,haveposedseriouschallengestothelendingmodeloftraditional?nancialintermediaries(

Bootetal.

2021

).

Figure

1

showsthatBigTechcredithasovertakencreditissuedbydecentralizedplatformsinrecentyears.BigTechcreditaccountsfor2%-3%ofgrossdomesticproduct(GDP)incountrieslikeChinaandKenya.TheseBigTechcreditsareparticularlyimportantformicro,small,andmedium-sizedenterprises(MSMEs),whicharethebackboneofentrepreneurshipandeconomicgrowth.Asoftheyear2018,MSMEsaccountfor99.8%ofestablishments,79.4%ofemployment,and68.2%ofsalesintheChineseeconomy.Armedwithinformation,distribution,andmonitoringtechnologiesbuiltintotheecosystemofBigTechdigitalplat-forms,BigTechlendersareabletoreducerelianceontraditionalcollateralandthuscovermoreborrowersthathavebeenunservedorunderservedbytraditional?nancialinstitutions(

PetersenandRajan

1994

,

BergerandUdell

1995

,

Cornellietal.

2022

).BigTechcredithasbecomeatopconcernforeconomicpolicymaking(

Carstensetal.

2021

,

Adrian

2021

).Asrecognizedby

Philippon

(

2016

)and

Lagarde

(

2018

),thedisruptionbyFinTechbringsa“bravenewworld”formonetarypolicymakersandrequiresre-evaluationofthee?ectivenessofmonetarypolicytransmissionthroughthesenewlenders.Despitetheburgeoninglitera-

2

tureonFinTech,littleisknownaboutitsimplicationsformonetarypolicytransmission.

1

ThispaperbridgesthisgapbyexploringmonetarypolicytransmissionmechanismsthroughBigTechandconventionalbanks.

Total,USDbn

800

600

400

200

0

2013201420152016201720182019

PlatformCreditBigTechCredit

percapita,USD

150

100

50

0

2013201420152016201720182019

PlatformCreditBigTechCredit

Figure1:GlobalFinTechCredit

Datasource:

Cornellietal.

(

2020

).

WeemployauniquedatasetcoveringthefullborrowinghistoryofsampledMSMEsfromamajorBigTechlenderandtraditionalbanksinChina.WeaccessedcreditdatafromtheAntGroup,oneofthedominantBigTechcompaniesbothdomesticallyandinternationally,andmatchwiththeseMSMEs’borrowinghistoryfromtraditionalbanks.OurdatasetcoversmonthlyobservationsofbothBigTechcreditandbankcreditto?rmsfromJanuary2017toDecember2019.Combinedwithvariationsinmonetarypolicy,ourdatasetprovidesanideallaboratoryforinvestigatingmonetarypolicytransmissionmechanismsthroughBigTechlendersandtraditionalbanks.The?ndingsbasedontheevidencefromChinamayshedlightonregulatoryandmonetarypoliciesinothercountriesaswell.

Ouridenti?cationstrategyfocusesontheextensivemargin,capturedbyanewlending

relationshipbetweenabankanda?rm,andtheintensivemargin,capturedbynewlyissued

1See

Allenetal.

(

2021

)forasurveyofFinTechresearchandpolicydiscussion.

3

loanstoa?rmthathasalreadyborrowedfromthebank.WeexploretherelativeresponseofBigTechlendingtochangesinmonetarypolicy,comparedwithtraditionalbanklending.Aftercontrolling?rms’demandforcredit,ourestimatescapturetheimpactofmonetarypolicythroughthecreditsupplyofdi?erenttypesofbanks.Inaddition,weexaminetherealimpacton?rmsofBigTechcreditrelativetoconventionalbankloansbycomparingsalesgrowthinresponsetochangesinmonetarypolicy.

Themain?ndingsofthepaperarethefollowing.We?ndthatBigTechloanstendtobesmaller,andBigTechbanksgrantcredittomorenewborrowers,comparedwithconventionalbanks,inresponsetoexpansionarymonetarypolicy.Inotherwords,whenmonetarypolicyeases,BigTechlendersaremorelikelytoestablishnewlendingrelationshipswith?rms,comparedwithtraditionalbanks.BigTechbanks’advantagesininformation,monitoring,andriskmanagementarethepotentialmechanisms.Comparedwithtraditionalbankloans,BigTechlendingampli?esmonetarypolicytoalargerextentfor?rmsthathaveonlinebusinesses,ratherthan?rmsthathaveonlyo?inebusinesses,andwhenBigTechlendingiscomparedwithsecuredbankloans,ratherthanunsecuredbanksloans.However,BigTechandtraditionalbankcreditsto?rmsthathavealreadyborrowedfromthesebanksrespondsimilarlytomonetarypolicychanges.Overall,BigTechcreditampli?esmonetarypolicytransmissionmainlythroughtheextensivemarginrelativetotraditionalbankloans.Inaddition,monetarypolicyhasastrongerimpactontherealeconomythroughBigTechlendingthantraditionalbankloans.

Thisstudyrelatestothreebranchesoftheliterature.First,wecontributetotheliter-atureonmonetarypolicytransmissionbyfocusingonanewplayer,BigTechlenders,andcomparingtheirresponsestomonetarypolicywiththoseoftraditionalbanks.Thebanklendingchannelofmonetarypolicy(

BernankeandBlinder

1988

,

1992

,

KashyapandStein

1995

)dependsoncross-sectionalheterogeneityinvariousdimensions,includingliquidity,size,incomegap,leverage,andmarketpower(

KashyapandStein

2000

,

Brissimisetal.

4

2014

,

Drechsleretal.

2017

,

Gomezetal.

2021

,

Wangetal.

2021

).Therisktoleranceandriskexposureof?nancialintermediationmayamplifymonetarypolicyshocks,asisfoundby

Coimbraetal.

(

2022

)and

DiTellaandKurlat

(

2021

).Heterogeneityinlenders’techno-logicalcharacteristicsisamissinglinkintheliterature.

2

Recently,

Hasanetal.

(

2020

)and

Hasanetal.

(

2022

)examinetheroleofregionalFinTechpenetrationandbanks’in-housetechnologydevelopmentinthee?ectivenessofmonetarypolicy.

DeFioreetal.

(

2022

)studyBigTech’sresponsetomonetarypolicybasedoncross-countryannualdataandmodeltheroleofBigTechasfacilitatingmatchingbetweensellersandbuyers.

Zhou

(

2022

)emphasizestheroleofsocialnetworkinhelpingFinTechenhancethetransmissionofmonetarypolicytothemortgagemarket.

ThekeyinnovationofourstudyisthatwefocusonthemonetarytransmissionmechanismthroughBigTechlendingrelativetotraditionalbanklendingbyexploringquasi-loan-leveldatabetweenMSMEsandtwotypesoflenders,BigTechandtraditionalbanks.Theevi-dencethatBigTechlendingampli?esmonetarypolicyalsoaddstotherecentliteraturethatinvestigatestheroleofnonbanksinmonetarypolicytransmission(e.g.,

Elliottetal.

2019

,

Chenetal.

2018

).

Second,ourstudyisrelatedtotheburgeoningstudiesontherelationshipbetweenFin-Techlendersandbanks.WecontributetotheliteraturebydirectlycomparingthelendingbehaviorsofthesetwotypesoflenderstothesameMSMEborrowersthroughthelensofauniquedataset.Assummarizedin

Stulz

(

2019

),

Bootetal.

(

2021

),

Thakor

(

2020

)and

Bergetal.

(

2022

),therecentwaveof?nancialtechnologiesisnewandhasbroughtanabun-danceofdataandcodi?cationofsoftinformation.Thesedevelopmentshavestrengthenedscreeningandmonitoring,whichrationalizetheempirical?ndingthatcomparedwithbanks,

FinTechlendersrelymoreonhardinformation.Ontheonehand,manystudiesexamine

2Therearestudiesfocusingon?rms’technologyadoptionanditse?ectonmonetarypolicy,buttheyarelimitedtonon-?nancial?rms.Forinstance,

Consoloetal.

(

2021

)?ndthat?rms’informationtechnologyinvestmentweakensthecreditchannelofmonetarypolicytransmission,and

FornaroandWolf

(

2021

)studytheimpactofmonetarypolicyon?rms’technologyadoptiondecisions.

5

whetherFinTechlendingsubstitutesfororcomplementsbanklending.Forinstance,usingU.S.mortgagelendingandpersonalcreditdata,

Buchaketal.

(

2018

),

DiMaggioandYao

(

2021

),and

DolsonandJagtiani

(

2021

)showthatFinTechlendersusedi?erentinformationtosetinterestratesrelativetobanksandaremorelikelytoservenonprimeconsumers.UsingconsumerlendingdatafromLendingClubandbanksintheUnitedStates,

Jagtiani

andLemieux

(

2018

)and

Hughesetal.

(

2022

)showthatFinTechpenetratesareasthatareunderservedbybanks.

Surietal.

(

2021

)and

ErelandLiebersohn

(

2022

)?ndthatFinTechcouldimprove?nancialaccessandresilience.

GopalandSchnabl

(

2022

)documentthatFinTechlenderssubstitutedforthereductioninbanklendingtosmallbusinessafterthe2008?nancialcrisis.

Tang

(

2019

)and

Beaumontetal.

(

2022

)showthatFinTechlendingsubstitutesbanklendingforinfra-marginalbankborrowersbutcomplementsbanklendingwithrespecttosmallloans.

Liuetal.

(

2022

)comparesyndicatedloansbyaBigTechlenderandatraditionalbankinChinaand?ndthatBigTechloanstendtobesmaller,havehigherinterestrates,andarerepaidfarbeforematurity.

Buchaketal.

(

2021

)useChinesedatatoshowthatFinTechfacilitatestheinterestrateliberalizationofbanksthroughcompetitionindeposit-likeproducts.Otherrecentstudies,suchas

PierriandTimmer

(

2022

),

Linetal.

(

2021

),

Kwanetal.

(

2021

),

Heetal.

(

2021

),

Hasanetal.

(

2022

),and

Modietal.

(

2022

),focusontechnologyadoptionbybanksandexamineitsimpactonlending.Although

Stulz

(

2019

)highlightsthespecialroleofBigTechcredit,thereislittleevidenceonthedi?erenceincorporatelendingbetweenBigTechlendersandbanks,inparticulartheirresponsestomonetarypolicyshocks.Thisstudy?llsthisgapintheliterature.

Third,thispaperalsocontributestotheliteratureon?nancialinnovationandeconomicgrowth,byhighlightingtheimpactofBigTechcrediton?rmperformance.Manystudiesfocusonthereale?ectsoftheinnovationsofnon-?nancial?rms,suchas

Akermanetal.

(

2015

),

Beaudryetal.

(

2010

),and

Autoretal.

(

2003

).Thesestudiesdwarfthoseontech-nologicalinnovationinthe?nancialsector,whichmayspureconomicgrowth.Forinstance,

6

Becketal.

(

2016

)showthatbankinginnovationisassociatedwithhighergrowthincountriesandindustrieswithbettergrowthopportunities.

GortonandHe

(

2021

)?ndthatbankinginnovationcontributestoeconomicgrowthbyallowingbankstoo?erlongermaturityloanstotherealsectorwithhigherproductivity.Bycontrast,researchonthereale?ectsofFin-TechorBigTechcreditisquitelimited.

Chenetal.

(

2022

),

E?caetal.

(

2021

),

Ahnertetal.

(

2021

),and

Becketal.

(

2022

)documentthataccesstoFinTechcreditreducessalesvolatil-ity,increasessalesgrowth,andspurs?rminvestmentandentrepreneurship.Inthisstudy,weprovidefurtherevidencetoshowthat,comparedwithtraditionalbanklending,BigTechcreditincreasesMSMEs’salesgrowthinresponsetochangesinmonetarypolicy,echoingtherealimpactofmonetarypolicyasin

GertlerandGilchrist

(

1994

).

Therestofthepaperisstructuredasfollows.Section

2

describestheinstitutionalbackgroundofBigTechcreditinChina,thedataconstruction,andthevariablesusedinthepaper.Section

3

illustratestheidenti?cationstrategiesandreportstheempiricalresults.Section

4

providesfurtherdiscussion.Section

5

concludes.

2DataandVariables

ChinahasgraduallybecomealeadingplayerinBigTechcredit.AccordingtoboththetotalandpercapitaBigTechcreditofthetopsixcountriesfrom2013to2019(seeFigure

A1

intheappendix),China’sBigTechcredithasdominatedothercountriessince2017.Ontheonehand,aidedbyadvantagesininformation,technology,distribution,andmonitoringbuiltintoBigTechplatforms’ecosystems,BigTechcompanieshaveaccesstomillionsofunservedandunderservedcreditusersatverylowcost,particularlyMSMEs.Ontheotherhand,thegovernment’sregulatorytoleranceintheearlystagedevelopmentofFinTechhasplayedanimportantroleinsupportingtherapidexpansionofBigTechcredit(see

Chui

2021

).DoesBigTechcreditsubstitutefororcomplementtraditionalbanklendingto?rmssinceboth

7

typesofcreditprovidersmayfacethesamepoolofpotentialcreditusers?IsBigTechcreditmoreresponsiveto?nancialmarketconditions,suchasthemonetarypolicystance,partic-ularlyindevelopingcountrieslikeChina?China’sBigTechcreditdi?ersfromthatofothercountriesinmanydimensions.Oneimportantdi?erenceisthatunlikeintheUnitedStatesandotheradvancedeconomies,BigTechlendinginChinaisdominatedbybusinesslend-ingratherthanmortgagelending.WillBigTechcreditreduce?rms’,particularlyMSMEs’,?nancialconstraintsandboosttheirgrowth?

Toaddressthesequestions,weusedatafromthebiggestBigTechcreditproviderinChina,MYBank.MYBankisownedbytheAntGroup,whichisana?liatecompanyoftheAlibabaGroupandoperatesvirtuallywithoutphysicalbranches.Sinceitslaunchin2015,MYBankhasfollowedthesamerulesandpoliciesoftheChinaBankingandInsuranceRegulatoryCommission(CBIRC)astraditionalbanks.

3

MYBankmainlyserveshouseholdsandMSMEssuchase-commercesellersandQRcodeo?inemerchants.TheAntGroupownstheworld’slargestdigitalpaymentplatform,Alipay,whichiseasytoaccessandusebybothmerchantsandcustomers.Bothe-commercesellersandQRcodeo?inemerchantsleavedigitalfootprintswhentheyuseAlipaytosettleonlineoro?inetransactions.Armedwiththisinformationandanadvancedriskmanagementmodel,MYBanko?ersloanswitha“contact-freefeature,”withoutanyvisitstophysicalbankbranches,underaso-called“310”model.Thatis,MYBankpromisesthecompletionofuserregistrationandloanapplicationwithin3minutes,moneytransfertoanAlipayaccountwithin1second,and0humanintervention.MoreinstitutionalbackgroundonMYBankandotherBigTechlendersinChinacanbefoundin

Frostetal.

(

2019

),

Huangetal.

(

2020

),

Hongetal.

(

2020

),

Hauetal.

(

2021

),

Gambacortaetal.

(

2022

),and

Liuetal.

(

2022

).

3TheChinaBankingRegulatoryCommission(CBRC)wastheagencythatregulatedthebankingsectorinChina.InApril2018,itwasmergedwiththeChinaInsuranceRegulatoryCommission(CIRC)toformtheCBIRC.

8

DepositRatio(%)

60

40

20

0

2015201620172018201920202021

MYBankTraditionalBanks

Profitability(%)

1

.5

0

?.5

2015201620172018201920202021

MYBankTraditionalBanks

CapitalAdequacyRatio(%)

20

15

10

5

0

2015201620172018201920202021

MYBankTraditionalBanks

NPLRatio(%)

2

1.5

1

.5

0

2015201620172018201920202021

MYBankTraditionalBanks

Figure2:MainIndicatorsforMYBankandTraditionalBanksSources:AnnualReportofMYBank;CBIRC.

Therearesimilaritiesanddi?erencesbetweenMYBankandtraditionalbanks.Both

typesofbanksareregulatedbytheCBIRC,attractdeposits,andlendtocreditusers.Theymayhavedi?erentlendingmodels.Traditionalbanksusuallyrequirein-personinteractionandinspectiontoissueloansandthereforetaketimetoapproveloanapplications.MYBankissuesloansveryquicklybyusingvarioussoftandhardinformationfromtheAntGroupanditsparentcompany,theAlibabaGroup.Therepaymentschedulecouldbedi?erenttoo.LoansfromMYBankcanberepaidearlywithoutanycost(

Liuetal.

2022

).Figure

2

showsthemain?nancialindicatorsforMYBankandothertraditionalbanksfrom2015to2021,includingthedeposit-to-assetratio,pro?tabilitycalculatedastheratioofnetincometoassets,capitaladequacycalculatedastheratioofcapitaltorisk-weightedassets,andtheratioofnonperformingloans(NPLs)toassets.The?gureshowsthataftertheyearofitslaunch,2015,MYBankhastendedtodependlessonexternal?nanceviaattracting

9

deposits,haveaslightlylowercapitaladequacyratiothantraditionalbanksonaverage,buthavelowerpro?tabilityandNPLratio.Lowerpro?tabilitymaybeassociatedwithhighercompetitioninthecreditmarket,andthelowerNPLratiowouldimplythatMYBankmayhavebetterriskmanagementviaabundantinformationandadvancedtechnologies.

2.1DataConstruction

MYBankservesbothhouseholdsand?rmsinChina.Forourpurpose,wemainlyfocusonMYBank’sentrepreneurialcustomers.Weexplorehowmonetarypolicya?ectscreditexpansionandcontractiondi?erentlythroughMYBankandtraditionalbanks.Bothonlineando?ineentrepreneurialcustomerssettletransactionsviaAlipayandleavetheirdigitalfootprintsontheecosystemoftheAntGroup.Moreover,thebusinessactivitiesofonlinemerchantsonthedigitalplatformsoperatedbytheAlibabaGroupprovideadditionalin-formationforMYBanktoevaluatetheriskofthesemerchants.MYBank’slendingmodelmightrespondtomonetarypolicyquitedi?erentlycomparedwithtraditionalbanks.

DuetoMYBank’sdataregulationpolicy,weobtaineda10%randomsampleofits?rmcustomersfromJanuary2017toDecember2019.Wedroppedinactive?rmsbythefollowingcriter