内部控制的弱点和股权成本

InternalControlWeaknessandCostofEquity:EvidencefromSOXSection404DisclosuresMariaOgnevaLeventhalSchoolofAccountingUniversityofSouthernCaliforniaLosAngelesCA90089-1441K.R.Subramanyam*LeventhalSchoolofAccountingUniversityofSouthernCaliforniaLosAngelesCA90089-1441K.RaghunandanSchoolofAccountingFloridaInternationalUniversityMiamiFL33199July2006*Correspondingauthor.Theauthorsgratefullyacknowledgehelpfulcommentsandsuggestionsfromtwoanonymousreferees,DanDhaliwal(editor),SurajSrinivasan,SarahMcVayandtheparticipantsofthe2005HaskellandWhiteCorporateGovernanceConferenceatCaliforniaStateUniversity–Fullertonandthe2006mid-yearFinancialAccountingandReportingSection(FARS)meetingatAtlanta.INTERNALCONTROLWEAKNESSANDCOSTOFEQUITY:EVIDENCEFROMSOXSECTION404DISCLOSURESAbstractSection404oftheSarbanes-OxleyAct(SOX)requireseverycompanytoreportontheeffectivenessofinternalcontrolsoverfinancialreporting.Section404hasarguablybeenthemostcontroversialprovisionofSOX,withmanyregistrantscomplainingthatthehighcostofcomplianceoutweighsitsbenefits.Incontrast,theSECandothershavearguedthattheSection404provisionsarebeneficialtothecapitalmarketsandwilleventuallyreducethecostofcapital.Inthispaper,weexaminetheassociationbetweenimpliedcostofequityandinternalcontroleffectivenessforfirmsthatfiledSection404reportswiththeSEC.Wefindmarginallyhighercostofequityforfirmsdisclosingmaterialweaknessininternalcontrolsthanforasampleoffirmsdisclosingnomaterialweaknesses.Thedifferencesincostofequitydisappearaftercontrollingforfirmcharacteristicsassociatedwithfirmsdisclosingmaterialweaknesses.Overall,ourresultsareconsistentwithinternalcontrolweaknessidentifiedunderSection404notbeingdirectlyassociated,onaverage,withhigherimpliedcostofequity.1INTERNALCONTROLWEAKNESSANDCOSTOFEQUITY:EVIDENCEFROMSOXSECTION404DISCLOSURESI.INTRODUCTIONSection404oftheSarbanes-OxleyActof2002(SOX)requiresmanagerstoacknowledgeresponsibilityformaintainingadequateinternalcontrolsoverfinancialreportingandprovideanauditorattestedassessmentoftheireffectivenessintheannualreport(10Kfiling),highlightingdiscoveredmaterialweaknesses(henceforth,ICweaknesses).Section404hasbeenextremelycontroversial.ManycompaniesmaintainthatthehighcostsofcomplyingwithSection404arenotcommensuratewithperceivedbenefitsandhavecalledformodificationorrepealofthesection(e.g.,ABA2005;FEI2005;Microsoft2005;Powell2005).RegulatorsinsistthattheinternalcontrolrequirementsofSection404wouldresultinhigherqualityfinancialreporting,whichinturnshouldlowercostofcapital(U.S.HouseofRepresentatives2005;U.S.Senate2004;Donaldson2005b).FinancialadvisorsmaintainthatinvestorsshoulddemandahigherriskpremiumforfirmswithICweaknessandcreditratingagenciesnotethatICweaknesswillbeanimportantconsiderationintheratingprocess(Wilfert2005;Moody’sInvestorService2004;FitchRatings2005).Inlightofthisdebate,itisimportanttoinvestigatewhetherICweaknessisassociatedwithcostofcapital.Accordingly,ourpaperusesalargesampleofSection404disclosurestoexaminewhetherreportedICweaknessesareassociatedwithhigherimpliedcostofequity.ICweaknesscanpertaintospecificaccountingissues(suchasrevenuerecognitionorinventoryaccounting)orbroadercontrolissues(suchasadequacyandtrainingofpersonneland“toneatthetop”).Accordingly,thereareatleasttworeasonstoexpecthighercostofequityforfirmswithICweakness.First,ICweaknessescanresultinpooraccountingquality,therebyincreasinginformationrisk.Informationriskhasbeentheoretically(e.g.,EasleyandO’Hara22004;Lambertetal.2006)andempirically(Francisetal.2004a2004b)linkedtohighercostofequity.Second,ICweaknessmaybesymptomaticofpoormanagementcontrolsingeneral,whichcouldincreasebusinessriskandthecostofequitytotheextenttheriskissystematic.Westudy2,025firmsthatfiledSection404reportswiththeSECovertheperiodfromNovember2004toJanuary2006;ofthese,252firmsfiledadverseopinions(i.e.,acknowledgedoneormoreICweaknesses)and1,773firmsfiledcleanreports.Ourprimarytestsuseimpliedcostofequitywhichweestimatebyinvertingtherelationbetweenstockpriceandanalysts’earningsforecasts.Tocontrolformacro-economiceffects,weemployapurelycross-sectionaldesignbyestimatingcostofequityasofJune2004foralloursamplefirms.Becauseoferrorinmeasuringimpliedcostofequity,weuseseveralalternativeproxieswithdifferentunderlyingmodelsandassumptions,aswellasdifferentdata,toestimatecostofequity.Tofurtherbuttressourevidence,wealsoexaminepatternsinvariousunderlyingdriversofcostofequity(riskattributes)thattheorysuggestsmaybeassociatedwithICweakness,includingearningsqualityproxies,informationasymmetryproxiesandmarketriskattributes.Inunivariateanalysis,wefindthatmostofourimpliedcostofequityproxiesare,onaverage,significantlyhigherforICweaknessfirmsthanforcontrolfirms(i.e.,thosethatfilecleanreports).However,themagnitudeofthedifferencesismodest,withmeandifferencerangingfrom26to126basispointsdependingonthemeasureused.Wefindsimilarresultsforvariousriskattributes.Forexample,meansignedandabsolutediscretionaryaccruals,probabilityofinformedtrading(PIN),betaandreturnvolatilityareallsignificantlyhigherforICweaknessfirms,suggestinghigherriskasso