using operating cash flow data to predict financial distress some extensions.pdf
《using operating cash flow data to predict financial distress some extensions.pdf》由会员分享,可在线阅读,更多相关《using operating cash flow data to predict financial distress some extensions.pdf(19页珍藏版)》请在冰豆网上搜索。
WileyandAccountingResearchCenter,BoothSchoolofBusiness,UniversityofChicagoarecollaboratingwithJSTORtodigitize,preserveandextendaccesstoJournalofAccountingResearch.http:
/www.jstor.orgAccountingResearchCenter,BoothSchoolofBusiness,UniversityofChicagoAccountingResearchCenter,BoothSchoolofBusiness,UniversityofChicagoUsingOperatingCashFlowDatatoPredictFinancialDistress:
SomeExtensionsAuthor(s):
CorneliusCaseyandNormanBartczakSource:
JournalofAccountingResearch,Vol.23,No.1(Spring,1985),pp.384-401Publishedby:
onbehalfofWileyAccountingResearchCenter,BoothSchoolofBusiness,UniversityofChicagoStableURL:
http:
/www.jstor.org/stable/2490926Accessed:
12-04-201522:
28UTCYouruseoftheJSTORarchiveindicatesyouracceptanceoftheTerms&ConditionsofUse,availableathttp:
/www.jstor.org/page/info/about/policies/terms.jspJSTORisanot-for-profitservicethathelpsscholars,researchers,andstudentsdiscover,use,andbuilduponawiderangeofcontentinatrusteddigitalarchive.Weuseinformationtechnologyandtoolstoincreaseproductivityandfacilitatenewformsofscholarship.FormoreinformationaboutJSTOR,pleasecontactsupportjstor.org.Thiscontentdownloadedfrom130.194.20.173onSun,12Apr201522:
28:
43UTCAllusesubjecttoJSTORTermsandConditionsJournalofAccountingResearchVol.23No.1Spring1985PrintedinU.S.A.UsingOperatingCashFlowDatatoPredictFinancialDistress:
SomeExtensionsCORNELIUSCASEY*ANDNORMANBARTCZAKt1.IntroductionRecently,financialstatementusersandregulatorsofpubliclyreportedfinancialaccountingdatahavearguedinfavorofthedisclosureofdetailedinformationonfirmscurrentoperatingcashflows(Harrisetal.1980,FASB1981,Smith1982,andThomas1982).TheFASBsuggeststhatsuchdisclosureswillallowuserstoassessbettertheamount,timing,anduncertaintyoffuturecashflows.Itstatesthatthegreatertheamountoffuturenetcashinflowsfromoperations,thegreatertheabilityoftheenterprisetowithstandadversechangesinoperatingconditions1981,p.11.Thepresumptionthathistoricaloperatingcashflowsenablebetterassessmentsoffuturecashflows,however,isbasedonintuitionratherthanonempiricalevidence(Griffin1982).Ourstudywasconductedtoassesswhetheroperatingcashflowdataandrelatedmeasuresleadtomoreaccuratepredictionsofbankruptand*AssociateProfessor,DartmouthCollege-theEditorsoftheJournalofAccountingResearchweresaddenedtolearnofthedeathofProfessorCaseyinOctober1984;tLecturer,HarvardUniversity.Theauthorsgratefullyacknowledgethehelpfulcommentsofpartici-pantsinaccountingworkshopsatseveraluniversities.SuggestionsmadebyE.Altman,P.Griffin,J.Horrigan,W.Mikkelson,T.Selling,J.Shank,C.Stickney,andP.Williamsonareappreciated,asisthecomputingassistancerenderedbyD.Bower,S.Hare,D.Roberts,andG.Peterson.J.Ohlsonkindlyprovidedmanyofthefinancialstatementsforthebankruptfirms.FundingforthisprojectwasprovidedbytheTuckAssociates,DartmouthCollege,andtheDivisionofResearch,HarvardBusinessSchool.AcceptedforpublicationSeptember1984.384CopyrightC),InstituteofProfessionalAccounting1985Thiscontentdownloadedfrom130.194.20.173onSun,12Apr201522:
28:
43UTCAllusesubjecttoJSTORTermsandConditionsUSINGOPERATINGCASHFLOWDATA385nonbankruptfirms.Recentstudies(GombolaandKetz1983andGombolaetal.1983)foundthatratiosbasedonoperatingcashflowloadonaseparatestatisticalfactor,suggestingthatoperatingcashflowvariablesmaybeusefulindescriptiveandpredictivestudiesinvolvingfinancialratios.Ourresultssuggestotherwise,atleastwithrespecttobankruptcyprediction.Inanotherstudy(CaseyandBartczak1984)wereportedthataccrual-basedmultivariatediscriminantmodelsforecastedcorporatebankruptcymoreaccuratelythananysingleoperatingcashflowratio.Thefocusofthepresentstudyisonthemarginalpredictivecontentoftheoperatingcashflowratios,incontrasttoourpreviousstudywhichexaminedtheirunivariatepredictivevalue.Thepoorperformanceoftheoperatingcashflowratiosinthepreviousstudydoesnotprecludetheirabilitytoenhancepredictivepowerwhenusedincombinationwithaccrual-basedratios.Previousaccountingstudieshavebeenunabletodocumenttheexistenceofincrementalinformationcontentinoperatingcashflowdata(Bowen,Burgstahler,andDaley1984).Unlikethepresentstudy,thesestudieshaveusedrestrictivesurrogatesformeasuringcashflowfromoperations,andthustheissueofwhetheroperatingcashflowdatahaveincrementalinformationvalueisunresolved.Insection2wereviewpriorbankruptcystudiesthathaveexaminedthepotentialusefulnessofoperatingcashflowdata.Section3describesthemethodologyemployedinthepresentstudy,includingselectionofsamplefirmsandanalysesappliedtothefirmsaccountingdata.Resultsarepresentedinsection4,followedbysomelimitationsofthestudyandpossibledirectionsforfuturerelatedresearchinsection5.2.PreviousResearchMostaccountingandfinancestudiesofcorporatefinancialdistressdefinecashflowasnetincomeplusnonworkingcapitalexpenses,sotheyomititemssuchaschangesincurrentassetsandcurrentliabilitieswhichmayhaveasignificantimpactonacompanysactualcashflowfromoperations.Thishasledtosuggestionsforbroadermeasuresofcashflows(e.g.,LargayandStickney1980andGombolaandKetz1983).Ourdefinitionofcashflowfromoperations(CFO)isessentiallythesameBankruptcywasselectedasthespecificformoffinancialdistressandasthecriterioneventforthreeprimaryreasons:
(1)thedirectandindirectcostsofbankruptcyaresignificantinrelationtothevalueofthefirm(Altman1983a):
(2)resultsfromthisstudycanbecomparedwithpreviousstudiesofbankruptcyprediction;(3)thecostsofdatagatheringandtheproblemofinterpretingtheeconomicsignificanceofotherevents(e.g.,loandefault)wereviewedasgreaterthanthebenefitsofusingeventswhoseoccurrencesarearguablylesssubjecttononeconomicfactorsthanisbankruptcy.Thus,whilebankruptcypredictionperseisnotthemainfocusofthisstudy,bankruptcyisnonethelessaconvenienteventforassessingoneaspectoftheinformationvalueofoperatingcashflowdata.Thiscontentdownloadedfrom130.194.20.173onSun,12Apr201522:
28:
43UTCAllusesubjecttoJSTORTermsandConditions386JOURNALOFACCOUNTINGRESEARCH,SPRING1985astheirs:
workingcapitalprovidedbyoperations,plusorminuschangesinthenoncashworkingcapitalaccountsexceptforshort-termindebt-edness(seasonalbankloans,nontradenotespayable,andthecurrentportionoflong-termdebt).ThisconceptofoperatingcashflowisalsoconsistentwiththerecentFASBExposureDraft1981.EmpiricalresearchontherelationshipbetweenCFOandfinancialdistresshasprovidedonlylimitedevidencethatsuchdataareusefulindiscriminatingbetweentroubledandhealthyfirms(table1).Noneofthestudieslistedtherewasvalidatedusingseparateholdoutsamplesofbankruptcompanies.Moreover,theproportionoffailedfirmsinallofthestudiesexceptthatofLargayandStickney19801,whichinvolvedonlyonecompany,wasclosetoorequalto50%.Thisislikelytoleadtoartificiallylowoverallerrorrates(Deakin1977).2Somebasicdifferencesbetweenourstudyandthoseofpreviousstudiesare:
(1)afocusonthepotentialmarginalimprovementinclassificationaccuracyusingCFO;
(2)alargernumberofsamplefirmsandasmallerproportionoffailedfirms;(3)avalidationoftheresultsbasedonaholdoutsample;(4)apreliminarycanonicalcorrelationanalysisoftheamountofvarianceintheoperatingcashflowratiosalreadyaccountedforbyaccrual-basedratios;and(5)theinclusionoffirst-orderinterac-tionsbetweentheoperatingcashflowandaccrual-basedratios.TheexpectationoftheFASB(andothers)thatthelevelofoperatingcashflowwillserveasausefulindicatorofthelikelihoodoffinancialdistressdoesnotderivefromanyformaltheory.Itismerelyconsistentwithanapparentshiftinpreferencebymanyanalyststousecashflowdataforassessingafirmsfinancialperformance(Hawkins1977).Thepresentstudyneitherproposesnortestsanytheoryoftheprobabilityoffinancialdistress.Whilesuchtheoriesdoexist(seeScott1981forasurvey;alsoEmeryandCogger1982),noonetheoryisgenerallyaccepted(BallandFoster1982)andnonegivesspecificattentiontoourmeasureofoperatingcashflow.3.Method3.1SAMPLEFIRMSThesamplefirmsarethesamecompaniesusedinourpreviousstudyonoperatingcashflows(CaseyandBartczak1984).Sixtyfirmswereselectedthathadpetitionedforbankruptcyduringtheperiod1971-82.Thebankruptfirmswereasubsetofthe105failedfirmsusedinapreviousstudy(Ohlson1980)andincludedfirmslistedbyDunandBradstreetandtheWallStreetJournalIndex.Criteriaforinclusionin2TypeIIerroristheclassificationofanonfailedfirmasfailed;TypeIerroristheclassificationofafailedfirmasnonfailed;overallerrorrateisaweightedcombinationofTypeIandTypeIIerrorrates.Thiscontentdownloadedfrom130.194.20.173onSun,12Apr201522:
28:
43UTCAllusesubjecttoJSTORTermsandConditionsUSINGOPERATINGCASHFLOWDATA387thepresentstudywere:
(1)thestockofthecompanywaspubliclytraded,and
(2)atleastfive,andinsomecasessix,yearsofpublishedfinancialstatementswereavailablepriortothefailuredate.Thefirstcriterionwasimposedinordertoundertakearelatedstudyinvolvingmarketvaluedata.Thesecondallowedfortheinclusionofatrendvariableintheanalyses.Wealsomadesurethatthedataforthefirstyearpriortofailurewereactuallypubliclyavailablepriortothebankruptcypetitiondate(Ohlson1980)byrequiringall