金融学 外文文献 英文文献 外文翻译 担保的作用和个人担保贷款的关系.docx

金融学 外文文献 英文文献 外文翻译 担保的作用和个人担保贷款的关系.docx

《金融学 外文文献 英文文献 外文翻译 担保的作用和个人担保贷款的关系.docx》由会员分享，可在线阅读，更多相关《金融学 外文文献 英文文献 外文翻译 担保的作用和个人担保贷款的关系.docx（7页珍藏版）》请在冰豆网上搜索。

金融学外文文献英文文献外文翻译担保的作用和个人担保贷款的关系

金融学外文文献英文文献外文翻译担保的作用和个人担保贷款的关系

外文文献原文

MaterialSource:

HitotsubashiUniversityAuthor:

lichirouesugi

Roleofcollateralandpersonalguaranteesinrelationshiplending:

evidence

fromJapan'sSMEloanmarket

1Introduction

Akeyissueofinterestintherecentliteratureonfinancialintermediationhasbeentheroleofrelationshiplending.Relationshiplendingisparticularlycommoninthecaseofsmallbusinesslending,becausesmallbusinessestypicallyrelyonbankloansforasubstantialpartoftheirfinancingneedsbutalsotendtobeinformationallyopaque.Animportantissueinthiscontextistheuseofcollateral,whichisacommonfeatureofloancontractsbetweensmallfirmsandbanksaroundtheworld,andanumberoftheoreticalandempiricalstudieshaveexaminedwhyitissowidespreadandhowitrelatestotheincentivesforborrowersandlendersandtheborrower-lenderrelationship.Forinstance,ithasbeenarguedthatinthepresenceofinformationasymmetriesbetweencreditorsandborrowers,collateralmaymitigatetheproblemofadverseselection（Bester,1985;1987）and/ortheproblemofmoralhazard（Bester,1994;Boot,Thakor,andUdell,1991）.Collateralalsoaffectstheincentivesofcreditors,whowilluseiteitherasasubstitutefor（Manove,Padilla,andPagano,2001）orcomplementto（RajanandWinton,1995;Boot2000;LonghoferandSantos,2000）screeningandmonitoringefforts.Anotheraspectofcollateralthatstudieshaveconcentratedonisthatitspresencemaydependonthelengthandintimacyoftherelationshipbetweencreditorsandborrowers（Boot,2000;BootandThakor,1994;Sharpe,1990）.Existingempiricalresearchhasyettoreachdecisiveconclusionsaboutthenatureoftheserelationships.

Thispaperseekstocontributetotheexistingliteratureoncollateralusingauniquefirm-leveldatasetofthesmallandmediumsizedenterprise（SME）loanmarketinJapan.Explicitlydifferentiatingphysicalcollateral（suchasrealestate）andpersonalguaranteesbybusinessrepresentatives,weinvestigatehowtheuseofcollateralandpersonalguaranteesaffectstheincentivesofborrowers,lenders,andtherelationshipbetweenthem.Morespecifically,weexaminethefollowingthreeissues.First,weexaminewhetherriskierborrowersaremorelikelytoberequiredtoprovidecollateralorpersonalguarantees.Second,weinvestigatehowcollateralandpersonalguaranteesaffectbanks’monitoringofborrowers.Third,weexaminethe

1

correlationbetweentheuseofcollateralandpersonalguaranteesontheonehandandtheclosenessofborrower-lenderrelationshipsontheother.

Thedatasetweemployisbasedmainlyonthe“SurveyoftheFinancialEnvironment”（SFE）conductedbytheSmallandMediumEnterpriseAgencyofJapaninOctober2002.Inordertofocusonfirmsthatmostlydependonbankloansfortheirfinancing,welimitthesampletofirmssatisfyingthelegaldefinitionofanSMEinJapan.WethencombinetheSFEdataforeachSMEwithinformationontheirmainbankobtainedfromthebank’sfinancial

statementsinordertocontrolforlendercharacteristicsaswell.Furthermore,tocontrolfortheeffectofgovernmentcreditguaranteesoncollateralandpersonalguarantees,inthemainanalysisofthispaperweexcludefromthesampleallfirmsthatenjoyedanyformofgovernmentcreditguarantee.

Asaresultofthisscreeningprocess,weendupwithasampleof1,702firms.Ourmainfindingscanbesummarizedasfollows.Wefindthatfirms’riskinessdoesnothaveasignificanteffectonthelikelihoodthatcollateralisused.Thus,wecannotfindfirmevidencethattheuseofcollateralmitigatesmoralhazard.Wefind,however,thatbankswhoseclaimsarecollateralizedmonitorborrowersmoreintensively,andthatborrowerswhohavealong-termrelationshipwiththeirmainbankaremorelikelytopledgecollateral.Thesefindingssuggestthatcollateraliscomplementarytorelationshiplending.Incontrast,thecomplementaritybetweenrelationshiplendingandpersonalguaranteesisweaker.

Asfarasweknow,thisisthefirstempiricalstudythatsystematicallyexaminestheroleofcollateralandpersonalguaranteesinJapan’sSMEloanmarket.Thetwomaincontributionsofthepaperareasfollows.First,giventhatJapanisgenerallyconsideredtohavearelationship-basedfinancialsysteminwhichtherelationship-lender,themainbank,playsacentralroleincorporatefinancing（RajanandZingales,2003）,thestudyhelpstoimproveourunderstandingoftheroleofcollateralinrelationshiplendingandcomplementsexistingstudiesthatfocusontheUnitedStatesandEurope.Second,andmoreimportantly,bydistinguishingcollateralandpersonalguarantees,thestudydetectsanimportantroleofcollateralinrelationshiplendingthathasnotbeenremarkedonmuchbefore.Aswearguebelow,althoughatypicalSMEinJapanhasalong-termrelationshipwithitsmainbank,itactuallyengagesintransactionswithseveralbanks,whichisnotcommoninothercountries.Apossiblecorollaryofthisisthatbecauseoftheinformational

2

free-riderproblemitcreates,thispracticemayreducethemainbank’sincentivetoscreenandmonitorborrowers.Sincecollateraldefinestheorderofseniorityamongcreditors,usingcollateralmaymitigatethefree-riderproblemandenhancethemainbank’sscreeningandmonitoring.Thisincentiveeffectforthemainbankbecomestenuousforpersonalguarantees,becausepersonalguaranteesdonotdefinetheseniorityamongcreditors.Thus,ourworkprovidesempiricalevidenceonhowcollateralaffectsrelationshiplenders’incentives,andcomplementspreviousstudiesthatfocusontheproblemofborrowerincentives（moralhazardandadverseselection）.

Theremainderofthepaperisorganizedasfollows.Section2developsourempiricalhypotheseswhicharebasedonprevioustheoreticalmodelsandempiricalresearch.Section3describesthedataandvariablesthatareusedinthepaper,andexplainsourempiricalmodel.

Section4presentstheresultsofourempiricalanalysis,andSection5concludes.

2Empiricalhypotheses

2.1Borrowerriskiness

Muchoftheempiricalliteratureinthisfieldexaminestheoreticalpredictionsofasymmetricinformationmodelsontherelationshipbetweenriskandcollateral.Ifthebankcannotdiscernborrowers’riskiness（hiddeninformation）,thencollateralmayserveasascreeningdevicetodistinguishbetweenborrowersandtomitigatetheadverseselectionproblem（Bester,1985）.Thisfollowsfromtheobservationthatalower-riskborrowerhasagreaterincentivetopledgecollateralthanariskyborrower,becauseofhislowerprobabilityoffailureandlossofcollateral.Hence,thelower-riskborrowerwillchoosethecontractwithcollateral.

Ontheotherhand,ifthelendercanobservetheex-anterisk,butthereareinformationasymmetrieswithregardtoactionstakenbytheborroweraftertheloanisextended,collateralpotentiallyprovidesanincentivetomitigatemoralhazard.Thus,oppositetomodelsfocusingonhiddeninformation,thoseconcentratingonhiddenactionsuggestthatitisobservablyriskierborrowersthatwillpledgecollateral,becausecollateralinducesmoreeffortbytheborrower（Boot,Thakor,andUdell,1991）,orreducestheincentivesofstrategicdefault（Bester,1994）.

Becauseourdatabaseonlycontainsmeasuresoffirms’observedriskiness

3

（namely,creditscores）,wecouchourfirstempiricalhypothesisasfollows:

Hypothesis1（H1）:

Theuseofcollateralishigheramongobservablyhigher-risk（lowcreditscore）borrowersifthelenderrequirescollateralinordertomitigatetheextentofmoralhazard.

Alternatively,ifborrowerspledgecollateralasasignaloftheirunobservedhighcreditquality,thenthereisnegativeornorelationshipbetweentheuseofcollateralandthecreditscore.

Consistentwiththetheoryofmoralhazard,mostexistingempiricalstudies,includingBergerandUdell（1990;1995）,havefoundapositiverelationshipbetweencollateralandborrowers’ex-anterisk.Jiménez,SalasandSaurina（2006）

directlytesttheadverseselectionandmoralhazardhypothesesbyseparatingex-anteandex-postmeasuresofborrowerriskiness,namelydefaultspriortoandaftertheloanorigination.Theirresultssuggestthatalthoughobservedriskinessincreasesthelikelihoodthatcollateralisused,thereisalsoanegativeassociationbetweencollateralanddefaultaftertheloanhasbeengranted,whichisconsistentwiththeadverseselectionargument.

Itshouldbenotedthattheoriesofcollateralasasolutiontomoralhazardand/oradverseselectionproblemsassumecollateralisexternaltothefirm.

Unfortunately,ourmeasureoftheincidenceofcollateraldoesnotdistinguishbetweenfirm（inside）collateralandpersonal（outside）collateral.Hence,throughoutouranalysis,wewillassumethatcollateralismostlyinside,butallowforthefactthattheremayalsobesomeoutsidecollateral.Asforpersonalguarantees,theyclearlyrepresentoutsidecollateral.

2.2Screeningandmonitoringbythelender

Recentresearchoncollateralalsodiscusseshowcollateralaffectslenders’incentiveswithregardtoinformationproduction,thatis,thescreeningofborrowers’

qualityandthemonitoringoftheirperformance.Thesetheoriesoftheeffectofcollateralonlenders’incentivesapplytobothinsideandoutsidecollateral.Manove,Padilla,andPagano（2001）,forinstance,arguethat,frombanks’pointofview,collateralcanbeconsideredasasubstitutefortheevaluationoftheactualriskofaborrower.Thus,banksthatarehighlyprotectedbycollateralmayperformlessscreeningoftheprojectstheyfinancethanissociallyoptimal.