资产负债表的稳健性与债务契约(Sunder,April2009)

Balance Sheet Conservatism and Debt Contracting JayanthiSunderaShyamV.SunderbJingjingZhangcKelloggSchoolofManagementNorthwesternUniversityApril2009___________________________________________________________________________aNorthwesternUniversity,6245JacobsCenter,2001SheridanRoad,Evanston,IL60208.E-mail:[e-mail];Phone:[phone].bNorthwesternUniversity,6226JacobsCenter,2001SheridanRoad,Evanston,IL60208.E-mail:[e-mail];Phone:[phone].cNorthwesternUniversity,6218JacobsCenter,2001SheridanRoad,Evanston,IL60208.E-mail:[e-mail];Phone:[phone].WethankAnneBeatty,ThomasLys,DarrenRoulstone,SugataRoychowdhury,workshopparticipantsatNorthwesternUniversity,OhioStateUniversity,FloridaStateUniversity,FARS2009conferenceandespeciallySudiptaBasu(discussant).WealsothanktheAccountingResearchCenterattheKelloggSchoolofManagementforfinancialsupport.BalanceSheetConservatismandDebtContractingAbstractWestudytheroleofcumulativeconservatisminassetvalues(balancesheetconservatism)onprivatedebtcontracting.Wefocusonbalancesheetconservatismtoisolateitseffectfromconditionalconservatismwhichhasbeenstudiedinthepriorliterature.Wehypothesizethatbalancesheetconservatismprovideslendersgreaterconfidenceinthecollateralvalueofthefirm’sassetsandreducestheriskintheloan(AssetValueHypothesis).Second,wehypothesizethatbalancesheetconservatismconstrainsfutureconditionalconservatismsuchthatdebtcontractingefficiencyishighonlywhenthebalancesheetconservatismisnothigh(ConstraintHypothesis).Usingasampleofbankloanswestudyinterestspreads,dealsize,covenantintensityandcovenantslackandfindresultsconsistentwithourhypotheses.Ourstudyshedslightonthescreeningandmonitoringroleofbalancesheetconservatismindebtcontracting.Balance Sheet Conservatism and Debt Contracting 1.IntroductionLendersrelyuponfinancialstatementsforscreeningandmonitoringofborrowers.Priorresearchhasprovidedevidenceofthelinkagesbetweenborrowerfinancialreportingchoicesanddebtcontracting(seesurveysbyHolthausenandWatts,2001;Fields,Lys,andVincent,2001;andadiscussionbySloan,2001).Thisstudyfocusesonhowconservativeaccountingchoicesinborrowers’financialstatementsimpactscontracttermsinprivatedebtcontracts.Theevidencebuildsuponinsightsfromrecentliteraturewhichhasexaminedasimilarquestion(forexample,Beatty,WeberandYu,2008;Zhang,2008;FrankelandLitov,2007;Nikolaev,2007).Theprimarilyfocusofthesestudiestoexaminehowongoingconditionalconservatismfacilitatesmonitoringoftheborrower.Incontrast,inthisstudywefocusonhowconservativeassetvaluesontheborrowers’balancesheetimpactthesettingofboththeinitialcontracttermsandthepost-loanmonitoringtermsbylenders.1Wedefine“balancesheetconservatism”asthecumulativeconservatisminassetvaluesanditincludestheeffectsofbothconditionalandunconditionalongoingconservatisminperiodspriortotheloancontractingyear.Therefore,balancesheetconservatismresultsindownward-biasedestimatesofassetvalues.Weconjecturethattheroleofbalancesheetconservatismindebtcontractingcouldbetwofold.First,balancesheetconservatismcouldprovideimportantinformationforscreeningtheborrowers.Thedownward-biasedassetvalueestimatescouldprovidevaluableinformationtolendersaboutthecollateralvalueoftheassetsofthefirmandtheriskofnon-realizationof1Ingenerallendersareinterestedintheassessmentofliquidationvaluesofasset-basedcollateralasreflectedonthebalancesheetandtheabilityoftheborrowertomakeperiodicinterestpaymentsasreflectedintheincomestatementandcashflowstatements.Ourfocusisprimarilyonthedebtcontractingeffectsoftheborrower’sbalancesheetvaluesofassets.2loanedamounts.Forexample,Watts(2003)highlightstheroleofconservativeassetvaluesinalleviatingtheconcernoflenderswithrespecttopreservationofassetvaluesintheeventofpotentialrepaymentproblemsoftheborrower.2However,basedonBallandShivakumar(2006)itisnotclearwhetherbalancesheetconservatismwouldaffectdebtcontractingaboveandbeyondpastconditionalconservatism.3Basedontheirargument,totheextentthatthebalancesheetconservatismisdrivenbypastunconditionalconservatism,anyknownbiasinassetvaluescanbeundone.Thissuggeststhatanyeconomicroleofbalancesheetconservatismindebtcontractingwouldbelargelysubsumedbyconditionalconservatismoftheborrower.However,whiletheeffectsofunconditionalconservatismcouldbeinverted,informationasymmetriesbetweentheborrowerandlendercouldmakeithardforthelendertocompletelyachievethisinversion.Thustheultimateeffectofbalancesheetconservatism,throughdownwardbiasedassetvaluation,ondebtcontractingremainsanopenempiricalquestion.Toexploretheseeffectswerelatetheborrowers’levelofbalancesheetconservatismatthetimeofloaninitiationonthecostofdebt,accesstodebtandlevelofmonitoringtermssetbythelender.Weexpectthatifdownwardbiasedassetvaluesarevaluabletothelender,borrowerswithhigherbalancesheetconservatismwouldhavealowercostofdebt,largerloansize,lowerexantemonitoringprovisions(measuredasnumberofcovenantsandslackincovenants).Alternatively,ifunderstatedassetvaluesmerelyaddnoisethenweexpectthatitwouldbecontractingneutralormayeven