relativevaluationrolesofequitybookvalueandn

JournalofAccountingandEconomics25(1998)1Ð34RelativevaluationrolesofequitybookvalueandnetincomeasafunctionofÞnancialhealthMaryE.Barth!,WilliamH.Beaver!,*,WayneR.Landsman!GraduateSchoolofBusiness,StanfordUniversity,Stanford,CA94305,USAKenan-FlaglerBusinessSchool,UniversityofNorthCarolina,ChapelHill,NC27599-3490,USAReceived1September1996;receivedinrevisedform1June1998AbstractThisstudytestspredictionsthatpricingmultiplesonandincrementalexplanatorypowerofequitybookvalue(netincome)increase(decrease)asÞnancialhealthdecreases.Testsusingasampleof396bankruptÞrmsandtestsusingalarger,pooledsamplebothyieldinferencesconsistentwithpredictions.Findingsarerobusttoinclusionofcontrolsforindustry,size,return-on-equity,andvolatilityofequityreturns.Equitybookvalueandnetincomemultiplesandincrementalexplanatorypowervarypredictablyacrossthreeillustrativeindustries,selectedbasedonthelikelyextentofunrecognizedintangibleassets.(1998ElsevierScienceB.V.Allrightsreserved.JELclassiÞcation:M4;G33;C2Keywords:Capitalmarkets;Equitybookvalue;Financialhealth;Netincome;Valuation1.IntroductionAmajorinstitutionalfeatureofÞnancialreportingistherequirementthatÞrmsÕgeneralpurposeÞnancialstatementsincludebothabalancesheetandanincomestatement.ThisrequirementpresumeseachstatementfulÞllsdi¤erentrolesoratleastprovidesmeasuresorinformationincrementaltotheother.Althoughthebalancesheetandincomestatementbothprovidevaluation-relevantinformation,adistinctiveroleofthebalancesheetistofacilitateloandecisionsandmonitoringofdebtcontracts.ItfulÞllsthisrolebyproviding*Correspondingauthor.Tel.:[phone];fax:[phone];e-mail:[e-mail]0165-4101/98/$Ðseefrontmatter(1998ElsevierScienceB.V.AllrightsreservedPII:S0165-4101(98)00017-2informationonliquidationvalues,theamountavailabletodebtholdersintheeventofdefault(Watts,1974).Incontrast,theroleoftheincomestatementprimarilyisforvaluingequity.ItfulÞllsthisrolebyprovidinginformationabouttheÞrmÕsabnormalearningsopportunities,i.e.,unrecognizednetassets.Becauseliquidationvaluesandprobabilityofdefaulta¤ectequityvalues,theÞnancialstatementsÕdi¤eringrolessuggestthatforequityvaluationthebalancesheetincreasesinimportanceandtheincomestatementdecreasesinimportanceasÞnancialhealthdecreases.SpeciÞcally,intherelationbetweenmarketvalueofequityandthestatementsÕprimarysummarymeasures,bookvalueofequityandnetincome,wepredictthatasÞnancialhealthdecreasesthepricingmultipleonandincrementalexplanatorypowerofequitybookvalueincreaseandthepricingmultipleonandincrementalexplanatorypowerofnetincomedecrease.1Totestourpredictionsweusetwocomplementaryresearchdesignsthatfocusonintertemporalchangesandcross-sectionaldi¤erencesinÞnancialhealth,respectively.First,weidentifyasampleofÞrmsthatsubsequentlyÞlebank-ruptcy,asubstantialeconomiceventassociatedwithdebtdefault.ForeachoftheÞveyearsprecedingbankruptcy,weestimatetherelationbetweentheseÞrmsÕmarketvalueofequityandnetincomeandbookvalueofequity.WepredictandÞndthatovertheyearspriortobankruptcy,thecoe¦cientonandincrementalexplanatorypowerofequitybookvalueincrease,andthecoe¦cientonandincrementalexplanatorypowerofnetincomedecrease.Second,weestimatethevaluationrelationforalarger,pooledsampleofÞrmsthatlikelyvarycross-sectionallywithrespecttoÞnancialhealth.ClassifyingÞrmsintotwoÞnancialhealthcategoriesusinge¤ectivebondratings,apartitioningvariabledirectlylinkedtodebtdefault,andthusÞnancialhealth,wepredictandÞndforÞrmsclassiÞedasbeinglessÞnanciallyhealthythanotherÞrmsthatthecoe¦cientonandincrementalexplanatorypowerofequitybookvaluearehigher,andthosecorrespondingtonetincomearelower.AkeyfeatureofourpredictionsisthatweexpectasÞnancialhealthdecreasestheincrementalcoe¦cientonandexplanatorypowerofequitybookvalueandnetincomewillbeofoppositesign.Althoughnetincomereßectsunrecognizednetassets,itlikelyreßectstheirvaluewitherror.Asaresult,equitybookvaluecanreßectthatportionofunrecognizednetassetsnotreßectedbynetincomeinadditiontoreßectingliquidationvalue.Hence,equitybookvalueandnetincomedonotnecessarilyprovideaÔcleanÕseparationbetweenliquidationvalue1WeusethetermsÔnetincomeÕandÔearningsÕinterchangeably.Althoughtheempiricalmeasureofnetincomevariesacrossrelatedstudies,werefertothemallasnetincome.Ourmeasureofnetincomeisincomebeforeextraordinaryitems.Incrementalexplanatorypowerofequitybookvalue(netincome)isthedi¤erencebetweentheR2oftheregressionwithnetincomeandequitybookvaluevariablesandtheR2oftheregressionwithonlythenetincome(equitybookvalue)variables.Henceforth,werefertocoe¦cientsonandincrementalexplanatorypowerofequitybookvalueornetincomeastheirÔvaluationcharacteristicsÕ.2M.E.Barthetal./JournalofAccountingandEconomics25(1998)1Ð34andunrecognizednetassets,suchasresearchanddevelopment,brandnames,technologicalcorecompetencies,customerloyalty,andgrowthoptions.WithdecreasedÞnancialhealth,unrecognizednetassetsdiminishinvalue,andthecoe¦cientonandincrementalexplanatorypowerofnetincomefalls.Totheextentthatequitybookvalueisaproxyforunrecognizednetassets,itscoe¦cientandexplanatorypowerwillalsofall.However,equitybookvalueÕsc