261.e中小企业成本控制研究 外文原文

'CharacteristicsofFirmswithMaterialWeaknessesinInternalControl:AnAssessmentofSection404ofSarbanesOxleyStephenBryanBabcockGraduateSchoolofManagementWakeForestUniversityStevenLilienBaruchCollegeCityUniversityofNewYorkAbstractThelegislationknownasSarbanesOxley(SOX)requiresfirmstoassesstheirinternalcontrolsoverfinancialreportingandtoreportmaterialweaknesses,asdefinedbythePublicAccountingOversightBoard.Baseduponearlyevidence,wefindthatfirmswithmaterialweaknessesare,onaverage,bothsmallerandworseperformersthantheirmatchedindustrycounterparts.Wealsofindthatfirmswithmaterialweaknesses,onaverage,havehigherbetas,suggestingahigherdiscountbythemarketforthesefirms.Fromamacro-economicview,thetotalmarketvalueoffirmswithreportedmaterialweaknessesisonly1.28%ofthemarketvalueoftheS&P500firms.Finally,althoughwedocumentnegativestockreturnsonthedateoftheannouncementofthematerialweakness,overanarrowinterval,thereturnsareinsignificant.Identifyingsmallfirmsthatcollectivelyconstituteaminorportionoftheeconomyataveryhighcosttoallpublicfirmsseemsoutofbalance.WhetherSOXwillyieldbenefitstocorporationsthroughbetteroperations,reducedcostofcapital,orothermeansremainstobeseen.Moreover,ifthesebenefitsmaterialize,whethertheyshould,ineffect,belegislatedisamatterofdebate.AlthoughsomemaintainthatSOXwillreduceearningsmanagement,firmscontinuetomanageearningsthroughproformaearnings.Furthermore,althoughtheSEC’sRegG,alsoenactedaspartofSOXlegislation,makesthe“manageddisclosuresystem”throughproformaearningsmoretransparent,itfailstoexercisecontroloverthereconciliationprocessthat19 firmsusebothtosetandtomeetanalystearningsestimate.IntroductionThegoalofthe“ThePublicCompaniesOxley(SOX),isrestorationofinvestorconfidence,whichhadbeenshakenbyEnronetal.Supposedly,improvedinvestorconfidencewillflowfromimproveddisclosuresoffinancialinformationandbettercorporategovernance,includinginternalcontroloveraccountingmeasurement,recognitionanddisclosurepresses.SOXisanattempttolegislatebetterdisclosureandinternalcontrol.OfSOX’smanyrequirements,themainoneforthisstudyisSection404,“EnhancedFinancialDisclosures,andManagementAssessmentofInternalControl.Section404requirescompaniestoassesstheeffectivenessoftheirinternalcontrolsoverfinancialreportingandreportthatassessmentintheirannualfilingswiththeSEC(Form10-K)1Firmsareencouragedtoincludethereportoninternalcontrolsinthe“glossy”annualreporttoshareholders,aswellasinthe10-K.Moreover,companies’externalauditorswillnotonlyauditthefinancialstatements,asbefore,butnowtheywillalsoauditmanagement’sassessmentoftheirinternalcontrolsoverfinancialreportingandperformtheirownindependentauditofthesamecontrolsystem.PriortoSOX,auditors’assessedinternalcontrolsbutonlyinconjunctionwithplanningtheaudit,i.e.,theydidnotformallyaudittheinternalcontrolsystem,norformallyreportontheresultsofthataudit.Althoughamajorpurposeoftheinternalcontrolauditisdeterrencefromfraudulentfinancialreporting,theramificationsofwhichincludeseveremarketreactionsandlegalliabilities,regulatorsalsomaintainthatbetterinternalcontrolwillhavebroaderfavorableimpactsoncompanies,suchasreducedcostofcapitalandbetteraccesstocapitalmarkets.Forinstance,tounderscorethepossiblepositivebenefitsofSection404,WilliamDonaldson,ChairmanoftheSECremarked,simplycomplyingwiththerulesisnotenough.[Companies]should,asIhavesaidbefore,makethisapproachpartoftheir...DNA.Forcompaniesthattakethisapproach,mostofthemajorconcernsaboutcompliancedisappear.Moreover,ifcompaniesviewthenewlawsasopportunities—opportunitiestoimproveinternalcontrols,improvetheperformanceoftheboard,andimprovetheirpublicreporting—theywillultimatelybebetterrun,moretransparent,andthereforemoreattractivetoinvestors.(SpeechattheNational19 PressClubonJuly30,2003).Notwithstandingtheallegedbenefits,theprivatecostsarenotinsignificantandarelikelytobecomeapermanentadditiontofirms’administrativeburden.Inasurveyof224firmsconductedbyFinancialExecutivesInternational(inJuly2004),respondentsrevealedthattheywillspendinthefirstyearofSOX,onaverage,anextra$3millioninordertocomplywithSection404.Thelargestcompanies(thosewithover$5,000millioninrevenues)willspendanaverageof$8million.Respondentsstatedthatauditfeesareexpectedtoincrease,onaverage,by53%inordertopayfortheattestationoverinternalcontrol.Theprofessionalaccountingfirmsareadmonishingclientsabouttheimportanceofsubsequentstepstoinsuresustainabilityofthecontrolsystemsothat“backsliding”willnotoccur.(DeloitteandTouche,2005).Additionally,thecostsassociatedwiththesestepsincludenotonly“out-of-pocket”costsbutalsotheindirectcostsofdivertingmanagerialattentionfromthecriticaldecisionsthatpertaintooperatingandinvestingactivitiesofthefirm.SOXalsodirectedtheSECtosetupthePublicCompanyAccountingOversightBoard(PCAOB),whichwaschargedwithdeterminingauditingstandardstoimplementSection404.InJune2003,theSECapprovedrulesimplementingSection404,andthefollowingyear,thePCAOBissuedstandardsforauditorattestationoffirms’internalcontrols.TheSECrulesandPCAOB’sAuditingStandardNo.2(“AnAuditofInternalControloverFinancialReportingPerformedinConjunctionwithanAuditofFinancialStatements”)requireboththemanagementreportandtheauditor’sreports.InadditiontoSection404,SOXcontainstherelatedSection302(“CorporateResponsibility”),whichrequiresacompany’sCEOandCFOtocertifyeachquarterlyandannualreport.Thecertificationstatesthatthefinancialstatementsarepresentedfairlyinallmaterialrespects;thattheydonotcontainuntruestatementsofmaterialfactorfailtostateamaterialfact;thattheCEOandCFOareresponsibleforestablishingandmaintainingcontrolsoverthefinancialreportingsystem;thattheCEOandCFOhaveevaluatedtheeffectivenessofthecontrols;thattheCEOandCFOhavedisclosedallsignificantdeficienciesandmaterialweaknessestotheauditorsandauditcommittee;andthattheCEOandCFOareresponsibleforestablishingandmaintaininginternalcontroloverfinancialreportingandforprovidingreasonableassuranceofthereliabilityofthefinancialreportingandforpreparingthefinancialstatementsinconformitywithGAAP.Finally,theCEOandCFOmustreportanychangesinthecompany’sinternalcontroloverthefinancialreportingduringthemostrecentquarter.DisciplinaryactionsagainstfirmsandtheabovenamedofficersarestipulatedinSection90619 (“White-CollarCrimePenaltyEnhancements”).Section906imposescriminalpenaltiesforofficerswhoprovidecertificationsknowntobeuntrue.TheSEChasgrantedpostponementofthereportsoninternalcontrols,asfirmshavebeenunabletomeettheinitialdeadlines.Specifically,acceleratedfilersmayamendtheir10-Kfilingswiththeinternalcontrolreportsupto45daysafterthedeadlineforthe10-K,beginningwithfiscalyearsendingonorafterNovember15,2004.(SEC,January21,2005)Togearupaheadofthedeadlines,firmsandtheirauditorsinstitutedexpandedprocedurestoassesstheirinternalcontrolsoverthefinancialreportingprocess.Incertaininstances,theexpandedprocedureshaveidentifiedmaterialweaknessesinadvanceofthedeadlinesfordisclosures.Overthepastseveralmonths,somefirmshavemadetheirfindingspublicin8-Kfilings.Somehavedonesoinconjunctionwith8-Kfilingsaboutaccountingrestatementsand/orchangesinauditors.Othershaveapparentlydonesovoluntarily.Thatis,forfirmswithamaterialweaknessthatisnotassociatedwithaconcomitantaccountingerrororauditorswitch,theredoesnotappeartobeacompellingreasonfordisclosingthematerialweakness,ifthefirmascertainsthatitcanremediatetheweaknessbythedisclosuredate.TheChiefAccountantoftheSEC,DonaldNicolaisen,notedthatitisessentiallytheresponsibilityofinvestorstoevaluatethenatureofeachSOXdisclosureofmaterialweaknesstomakeaninformedassessmentofitsimplications.TheChiefAccountantoftheSEC,DonaldNicolaisen,notedthatitisessentiallytheresponsibilityofinvestorstoevaluatethenatureofeachSOXdisclosureofmaterialweaknesstomakeaninformedassessmentofitsimplications.TheChiefAccountantoftheSEC,DonaldNicolaisen,notedthatitisessentiallytheresponsibilityofinvestorstoevaluatethenatureofeachSOXdisclosureofmaterialweaknesstomakeaninformedassessmentofitsimplications.SamuelDiPiazza,CEOofPWC,forecaststhatapproximately10%ofU.S.firmswilleitherdisclosematerialweaknessesorwilldisclosethattheyareunabletocertifytheirinternalcontrolproceduresbytheirreportingdate.Internalcontrolweaknesseswilllikelyclusteraroundindustryandfirmcharacteristics.Forinstance,smallfirmsandthoseinthetechnologysectorwilllikelybeheavilyrepresented(Bulkeley,2005).EarlyevidencethatinternalcontroldisclosurescouldhaverealeconomicconsequencescomesfromthedecisionbyMoody’sInvestorServices,whichinOctober2004,indicateditsapproachinevaluatingtheSOXreportsandindeterminingtheeffectsofmaterialweaknessesoncreditratings.EchoingtheabovequoteoftheChiefAccountant,Moody’sdoesnotbelieveallmaterialweaknessesininternalcontrolwillhaveseverenegativeeffectsonthefirm.19 Wearelessconcernedaboutmaterialweaknessesthatrelatetocontrolsoverspecificaccountbalancesortransaction-levelprocesses.Werefertothesematerialweaknessesas“CategoryA”materialweaknesses.Inmostcases,webelievethattheauditorcaneffectively“auditaround”thesematerialweaknessesbyperformingadditionalsubstantiveproceduresintheareawherethematerialweaknessexists.WeexpecttogivecompaniesreportingCategoryAmaterialweaknessesthebenefitofthedoubtandnottakeanyrelatedratingaction,assumingmanagementtakescorrectiveactiontoaddressthematerialweaknessinatimelymanner.Othermaterialweaknesses,whichwerefertoas“CategoryB”materialweaknesses,relatetocompany-levelcontrolsandmayresultinusbringingacompanytoratingcommitteetodeterminewhetheraratingactionisnecessary.(Moody’sInvestorServices,2004)Moody’scitesexamplesofwhattheywouldconsidertobeCategoryBweaknesses,suchasthe“toneatthetop”thatisnotconducivetoaneffectivecontrolenvironment.Specificindicatorscouldincludelackofconsistentpoliciesorlackofacompany-widecodeofconduct.Additionally,Moody’sstatesthattheCategoryBtypeweaknessesnotonlycastabadlightontheintegrityofthefinancialreportingprocessbutalsoonmanagement’sabilitytorunthebusiness.Ifamaterialweaknessmeansthatthereismorethanaremotelikelihoodofamaterialmisstatementoftheexternalfinancialstatements,whatdoesthatimplyaboutthequalityofinternaldatathatmanagementusestomakedecisions?PurposesofStudyTheabilitytodocumentinadirecttesttheeffectivenessofSOX’sgoalofrestoringinvestorconfidenceisdifficult.WecanhoweverattempttoassessSOXalonganumberofdimensions,fromwhichwecanbegintodrawcertaininferences.Ourconclusionsarelimitedfortwomainreasons.First,theimplementationdatehasjustrecentlypassed,subjecttodeferralsasmentionedpreviously.Second,certaineffectsofthelegislationarenotdirectlydetectableforthefollowingreasons.•Conceivablyafirmmayuncoveramaterialweaknessandprovideaneffectiveremedybeforeadisclosuredate(e.g.,fiscalyearend);•BeneficialexternalitiesmightaccruetothefirmbecauseofSOXbutthespecificationofthemodeltodetectandmeasurethesebenefitsandtheircausalfactorsis19 elusive.Forinstance,areportbyMcKinsey(2000)claimsthatmoneyspentoninternalcontrolhaspositiveeffectsbeyondmerecompliance.McKinseyclaimsthatinvestorsarewillingtopayasignificantpremiumforfirmswith“goodcorporategovernance,”ofwhichinternalcontrolsystemsareacomponent.However,measurementerrorstillsurroundscorporategovernance,althoughvariousmetricshavebeenofferedintheliterature(Gompers,2002).Giventheabovelimitingfactors,ourgoalsaretoidentifywithmaterialweaknessesandtomeasurestockmarketeffectsonandaroundthedisclosuredateofthematerialweakness.Weprovidetwocases(McAfeeInc.andEastmanKodak)bywayofillustrations.WedrawinitialconclusionsaboutthemandateofSection404andofferavenuesforfurtherresearch.RelatedliteratureSOXhasgeneratedmuchacademicinterest,particularlyamonglegalscholar.Forinstance,Ribstein(2002)notesthatgovernancefailures(particularlyrelianceoncontractingdevices)reinvigoratedlegislativeresponsestocorporateoversight,buthealsoarguesthatthenewlegislationwillhaveonlylimitedeffectivenessandwillnotnecessarilydoabetterjobthanmarketsthemselves,especiallygiventhehighcostofimplementation.Further,SOXrepresentsafederalinsertionintocorporategovernanceanditmaynotonlybeineffective,itcouldcauseharmtotheextentitreducesincentivesformanagerstoincreaseshareholdervalueandmisleadsmarketparticipantsintothinkingthatregulationhassolvedtheproblemsofcorporategovernanceandimprovedtheintegrityofthereportingprocess.Perino(2002)expressesconcernsoverSOXduetoinconsistenciesinthelegislation,whichheattributestotherushthroughthelegislativeprocess.Additionally,hearguesthatthedeterrenteffectoflongerpenaltiesandincreasedstatuteoflimitationsonsecuritiesfraudwillbeminimal.Ontheotherhand,moreresources(attheSEC)fordetectionmaybeeffective.Li,Pincus,andRego(2004)lookatmarketreactionsonsignificantdatesleadinguptopassageofSOXtotrytogaugewhetherthemarketsviewedthelegislationfavorably.TheyinferfromtheiranalysisthatmarketparticipantsviewedSOXfavorably,afindingthattheauthorsattributetoexpectedreducedearningsmanagementandenhancedcorporategovernance.Cohen,Dey,andLys(2005)provideevidencethatearningsmanagementdiddecreaseafterthepassageofSOX.Theyconcludethatthisrepresentsahighlevelofcorporatemisconduct,andnotjusta“fewbadapples.”Alongthesamelines,AllesandDatar(2003)argueforincreasedcontrolsoverthe19 reportingsystembecausemanagershavestrongincentivestomanageearnings:Buttoday,whenmanagersareunderpressuretocontinuallybeatanalysts’earningsexpectations,eitherbecauseofmarketdemandsorbecausetheirownoptions-basedincentivesmakeitprofitabletodoso,thepresumptionsurelyhastobethataccountingstandardsneedexplicitcontrolstoensurethattheyareimplementedappropriately.However,BryanandLilien(2004)showthatfirmsuseothermeansofmanagingearningsandmanagingexpectations,outsideoftheformaldisclosuresystem.TheycitetheexampleofEastmanKodak,whicheliminatedapproximately60%ofGAAPearningsover10consecutivequarters,throughproformaearnings,andconsequentlyvirtuallyeliminatedallearningssurprise.AlthoughtheSEC’sRegG,alsoenactedaspartofSOXlegislation,makesthe“manageddisclosuresystem”throughproformaearningsmoretransparent,itfailstoexercisecontroloverthereconciliationprocess(thatEKusedtoeliminatethe60%ofGAAPearningscomponents)firmsarebeingidentifiedashavingmaterialweaknesses,andweshowreactionstotheirannouncements.Wetieourfindingsintotheextantliteratureandpropose..Thus,itremainsunclearhowanassessmentofinternalcontrolwillhaveanyimpactonmanaged(Proforma)earnings.Whetherinternalcontrolshouldbeextendedtothe“RegGreconciliation”isdebatable.BrownandSivakumar(2003)showthatcertainnon-GAAPvariationsofearningsaremorevaluerelevant.Totheextentthatinvestorsrelymoreonnon-GAAPearnings,thenacaseforapplyinginternalcontroltothemanageddisclosuresystemcouldbemade.Gordon(2003)maintainsthattheimproveddisclosureprovisionsofSOXcouldbebeneficial,suchasthosepertainingtooff-balancesheetvehicles.However,Engel,Hayes,andWang(2004)showthatalthoughthebenefitsofbetterdisclosureandbettergovernancemayresultfromSOX,thecostofcompliancemayinducesomepubliccompaniestogoprivate.Theyfindevidenceconsistentwiththishypothesis.Bratton(2003)addressestherolethataccountingrulesmayhaveplayedincorporatefraud,particularlyEnron,andconcludesthattheassignmentofblamemoreappropriatelybedirectedattheauditprofession:Weaccordinglyneedtokeepourattentionfixedonthefactthattherecentgovernancebreakdownsoccurredentirelyontheenforcementside.Auditorshavestoppedactingasneutral19 reviewerstotakeupclientadvocacy.Enronwasanauditfailure,andnotacasewhereincomplete,ineffectiveorcorrupt[accounting]rulesfacilitatedafraud.(p.1055)Ourstudybringsfreshevidencetothedebate.Weshowwhichtypesoffirmsarebeingidentifiedashavingmaterialweaknesses,andweshowreactionstotheirannouncements.Wetieourfindingsintotheextantliteratureandproposedirectionsforcontinuesresearch.TypeofcontroldeficienciesAcontroldeficiencyexistswhenthedesignoroperationofacontroldoesnotallow,inthenormalcourseofperformingtheirassignedfunctions,managementoremployeestopreventordetectmisstatementsonatimelybasisDeloitteTouchesuggeststhattheemphasiswill,asapracticalmatter,resideondetectionratherthanprevention:“Nosystemofinternalcontrol,howeverwellplanned,willbefailsafefromtheoutset.Andnopersonorteamwillperformerror-free,especiallyintheearlyyearsofcompliance.Thekeyisnotsomuchpreventingerrorsasitisrecognizingandrespondingtothem.Awatchfuleyeshouldbemaintainedforpatternsofcontroldeficiencies,whichmayindicateanareaforadditionaltraining.”(DeloitteTouche“UnderControl”,page8.).Asignificantdeficiencyadverselyaffectsthecompany’sabilitytorecordorreportexternalfinancialdatareliablyinaccordancewithGAAP,suchthatthereismorethanaremotelikelihoodthatamisstatementofafirm’sfinancialstatementsthatismorethaninconsequentialwillnotbepreventedordetected.Significantdeficienciesarereportedtotheauditcommittee.FirmsarenotrequiredunderSection404todisclosesignificantdeficiencies.Materialweaknessesresultinmorethanaremotelikelihoodthatamaterialmisstatementofthefinancialstatementswillnotbepreventedordetected.Materialweaknessesthatexistatthefiscalyearendarereportedpublicly.Materialweaknesses(oneormore)willresultinaconclusionthatinternalcontroloverfinancialreportingisnoteffective.Judgmentsaboutthemagnitude(misstatementsthataremorethaninconsequentialversusthosethatarematerial)arerequiredtoclassifytheabovecategories.SampleWeperformedacomprehensivetextualsearchofpublicfilingsusingEdgar.Ourprimary19 searchtermwas“materialweakness”anditsvariations.Wescreenedtheresultingsample.Forinstance,somefirmsreportedsignificantdeficiencies,reportableconditions,orriskfactors,butthatthesedidnotrisetothelevelofmaterialweaknessesWealsoeliminatedrepeatobservationsthatoccurredwhenafirmdetectedamaterialweaknessandsubsequentlydisclosedstepstoremediatethematerialweakness.(Thiswouldcountasoneobservation).Weeliminatedfirmswithoutsufficientfinancialdata.Thisoccurredinsomeinstanceswhenfirmshaddelayedfilingsduetootherissuessuchasbankruptcy,auditorchanges,oraccountingerrors.Wehandcollectedfinancialdatasincedatabasesdidnotcoveraportionofthesample.Wealsohandcollectedstockpricessincefinancialdatabaseshavearchivaldatathatdidnotcorrespondtooureventwindows.Wedidnothandcollectpricedatausedintheestimationofbeta;ratherweusedCRSPmonthlyreturnsforallfirmsforwhichdatawereavailable.Wecontinuetoupdateoursamplewithnewlyidentifiedfirmsthathavematerialweaknesses,especiallysincetheimplementationdateonlyrecentlybecameeffective,subjecttothedeferral.Weconjecturethattheadditionaldatapointswillstrengthenourfuturelife.(SEC,2005)However,ourinitialsurveysoffilingsthatmeetourtextualsearchcriteriarevealthatmostfirmsarereportingthattheycannotyetdefinitivelysaythattheydonothavematerialweaknesses.Forinstance,SportsEntertainmentEnterprisesinits8-K(dated2/7/05)writes:Atthistime,duetotheongoingevaluationandtesting,noassurancescanbegiventhattherearemaynotbesignificantdeficienciesormaterialweaknessesthatwouldberequiredtobereported.Similarly,LogitechInternationalwrites:Ourevaluationandtestingisongoing,andtherecanbenoassurancethatwewillnotidentifysignificantdeficienciesormaterialweaknessesthatwouldrequireremediation.”(6-K,dated12/31/04)QMEDclaims:CompanymakesnorepresentationastowhethertheCompany"sauditorswillfindamaterialweaknessintheCompany"sinternalcontrols.”(8-K,dated12/6/04)Somefirms,however,declaredefinitivelythattheyhavenomaterialweakness.Forexample,OpswareInc.(8-K,2/3/05)states:therearenosignificantdeficienciesormaterialweaknessesinthedesignoroperationoftheCompany’sinternalcontrolsthatcouldadverselyaffecttheCompany’sabilitytorecord,process,summarizeandreportfinancialdata.19 ThereisnofraudinconnectionwiththeCompanyFinancialStatements,whetherornotmaterial,thatinvolvesmanagementorotheremployeeswhohaveasignificantroleintheCompany’sinternalcontrols.Theaboveexamplesaretypicalofthemostrecenthitsfromourtextualsearches.Thus,aftertheabovescreeningprocess,wehave,asofthiswriting,asampleof161firmswithmaterialweaknessesininternalcontroloverfinancialreporting.TheIllustrationofMcAfeeMcAfee(withtickersymbol“MFE”),formerlycalledNetworkAssociates,isasecuritysoftwarecompany,headquarteredinCalifornia,with2003salesof$936million,netincomeof$70.2million,andtotalassetsof$2,121million.Inan8-Kfiling,datedMarch9,2004,McAfeediscloseda$3.8millionrestatementofrevenueand$3.7millionrestatementofnetincome.DuringthepreparationandanalysisoftheRegistrant’s2003consolidatedfinancialstatements,theCompanyidentifiedandreportedtoPricewaterhouseCoopers,itsexternalauditors,andauditcommitteerequiredcorrectionstoitspreviouslyreportedorannouncedfinancialinformationrelatingtotheCompany’sbookingofinternationaldeferredrevenueandthemakingofmanualjournalentries.Intotal,theseadjustmentsincreased2003revenueby$3.8millionto$936.3million,2003netincomeby$3.7millionto$70.2millionand2003earningspershare,orEPS,by$0.03to$0.44(basic)and$0.43(diluted).TwodayslateronMarch11,2004,McAfeefiledanother8-KtodisclosethechangeinauditfirmsfromPWCtoDeloitte:PricewaterhouseCoopers(“PWC”)wasdismissedasNetworksAssociates,Inc.’s(the“Registrant’s”)independentaccountantsonMarch9,2004.Theauditcommitteehasauthorized,effectiveMarch10,2004,theengagementofDeloitte&ToucheLLP(“Deloitte”)asitsindependentaccountantstoaudittheRegistrant’sfinancialstatementsfortheyearendedDecember31,2004.Thesame8-Kfiling(ofMarch11,2004)alsodisclosedthatPWChadassertedthatMcAfeehadamaterialweaknessinitsinternalcontrolsystem.Inevaluatingthesecorrections,PWCdeterminedandreportedtoourauditcommitteethattheunderlyingcontrolissuesshouldbeconsideredamaterialweaknessunderstandardsestablishedbythePublicCompanyAccountingOversightBoardandtheRegistrantshouldinstituteadditionalrelatedcontrolprocedures.TheRegistrant’sauditcommitteehasdiscussedtheforegoingwithPWC,andtheRegistranthasbolsteredinternalcontrolsaroundtherecognitionofinternational19 revenuesaspartofitsquarterlyfinancialclosingprocessandthemanualjournalentryprocess.demandinghowtoautomaterevenuerecognitioncalculationssoastolimitthenumberofmanualadjustments,andengaginginadditionaltestingofitscontrolprocessesandprocedures.OnMarch12,2004,McAfeerestated(viaanother8-Kfiling)itsrevenuebyproductgroup,andonJuly27,2004,McAfeerestatedresults(alsoviaan8-Kfiling)foritsfirstquarterof2004.McAfeealsoassumeditscurrentname(previouslyitwasNetworkAssociates)onJune30,2004.OnMarch9,2004,PiperJaffraypublishedareportthatcalledtheinitialrestatement“embarrassing”forthecompany,buttheanalystsdidnotchangeratingsorestimates.OnMarch15,2004,KaufmannBrothersEquityResearchissuedareportthatnotedtherestatement,thematerialweakness,andthechangeinauditor.Theanalystcalledtheevents“ominous”andfurtherwrote:AnotherRedFlag:Itisdifficulttoascertainhowseriousthedismissalofthecompany’sauditorsisorwillbecome.Attheveryleast,theconfluenceoftheseeventsmaybeanunfortunatecoincidence,butwebelieveitisunwisetotreatitassuch.However,giventhatNetworkAssociatesremainsunderanSECinvestigation(fromissuesrelatedtopreviousmanagement),thatquestionshavearisenaboutaccountingcontrolissuesthathaveoccurredundercurrentmanagement,andthatthecompanydismisseditsauditorsallleadustorecommendfurthercautionontheshares.OnMarch23,2004,CitigroupSmithBarneywrote:RecentrestatementshavepartiallyeliminatedtheaccountingoverhangsincetheSEClaunchedaninvestigationin2002.However,thelackofclosureontheinvestigationandtherecentchangeinthecompany’sauditorsarethemainreasonsforourSpeculativeriskrating,andthefundamentalHoldrating.OnApril1,2004,CSFBinitiatedcoveragewithnomentionoftheaboveevents,ratherwithanoutperformratingandatargetpriceof$22(whenthecurrentpricewas$18).Thestockprice,appreciatedwellpastthistarget.ToconcludetheMFEexample,weunderscorethatMFEexhibitednotonlyamaterialweakness,butalsoanaccountingrestatementandanauditorswitch.Thesecompoundeventsoccurringatthesamefirmarenotatypicalthroughoutoursample.FindingsTheserviceindustries(whichincludecomputersoftware)andelectronicindustries(which19 includecomputerhardwareandotherbusinessequipment)togetheraccountfor36%ofthetotalsample,followedbyinstruments(8.7%)andfinancialservices(7.45%).Althougharangeofindustriesisrepresentedinthesampleoffirmswithmaterialweaknesses,thefirmsclusterintothesefewmaincategories.The“Big4”firmscollectivelyauditabouttwo-thirdsofthefirmswithmaterialweaknesses,andsmallerfirmsasagroupauditaboutone-third(34.16%).Showdescriptivestatisticsofthefirmswithmaterialweaknesses,aswellasdescriptivestatisticsoffirmsinthecorrespondingindustrieswithoutmaterialweaknesses.Thevariousmeasuresreportedinthetablearetakenfromthefinancialstatementsofthefiscalyearpriortotheannouncementofthematerialweakness.Asignificantportionofthesamplefirmsstatedpubliclythattheymustrestatetheirfinancialresultsasaresultofthematerialweaknessininternalcontroloverthereportingprocess.Oursamplecontainsonemajoroutlier—FannieMae,whichholdsover$1trillioninAssets.Givenourrelativelysmallsample,theInfluenceofFannieMaeleadsustoprovidesummarystatisticsbothwithandwithoutFannieMae.WeshowresultsinPanelsAandB,respectively.Theresultsindicatethatthesample(withFannieMae)hasmeanSalesof$1,134millionandmeanAssetsof$7,502million.TheratioofEBIT/TA(EarningsbeforeInterestandTaxesscaledbyTotalAssets)is-8.4%.TheratioofTL/TA(TotalLiabilitiestoTotalAssets)is62.6%,andtheaveragebeta(forfirmsforwhichwecouldobtainbeta)is1.448.MeanMarketCapis$1,249million.TheresultschangeinPanelB,whereFannieMaeisomittedfromthesample,particularlyformeanSales($805millioninsteadof$1,134million),meanAssets($1,239insteadof$7,502)million,andmeanMarketCap($731millioncomparedto$1,249million).Whencomparedtotheirrespectiveindustries,firmswithmaterialweaknessaresignificantlysmaller,exhibitworseperformance,andexhibithigherequityrisk.Forinstance,themeanSalesandmeanAssetsaresmallerforthematerialweaknessfirmsthanfortheirindustrycounterparts,especiallywhenFannieMaeisomittedfromthesample.WealsodocumentthatMarketCapissignificantlysmallerforthematerialweaknessfirms,bothwithandwithoutFannieMae.Toillustratemorebroadlytherelativemarketvalueofthematerialweaknessfirms,wefindthatthetotalmarketvalueoffirmswithreportedmaterialweaknessesis1.28%oftheHowever,oursummarymeasuresshowninTable3arebaseduponoriginallyreportedfinancialresults,notrestatedfinancialresults.marketvalueoftheS&P500firms(notshownin19 thetable).Forty-fivefirms(whichrepresent32%ofthesampleoffirmswithmarketcapdata)arebelowthe$75millionthresholdforacceleratedfilers.providedinPanelD.Thisprovidesfurtherevidencethatfirmswithmaterialweaknessesarenotonlysmallbutalsorepresentatinyproportionofequityvalueintheeconomy.Inaddition,theperformanceofmaterialweaknessfirmsisworsethantheirindustrycounterparts,asmeasuredbyEBIT/TA(-8.4%vs.-1.9%),andtheirequityriskissignificantlyhigher,asmeasuredbybeta(1.439vs.1.106).Wealsomeasurestockpricereactionaroundthedateoftheannouncementofthematerialweakness.Table4showsanaveragenegativerawreturnofslightlylessthan1%onthedateoftheannouncement,andthisresultissignificantlydifferentfromzeroatthe5%level(t-statis-1.99).Theresultsareverysimilarusingmarketadjustedreturns,wherethemarket,definedastheS&P500,issubtractedfromeachfirm’srawreturns.However,wealsonotethereturnsprecedingtheannouncementdate,althoughinsignificantlydifferentfromzero,are-0.0154twodaysbeforetheannouncementdateand+0.0115onedaybeforetheannouncementdate.Together,theresultsindicateconsiderablevariationinstockprices,and,overtheentireperiodshowninthetable(fromtwodaysbeforetoonedayaftertheannouncementdate),thereturnsareinsignificantlydifferentfromzero(-0.0090,withat-statisticof-0.92).Manyofthefirmswithmaterialweaknessesalsohadotherreportableeventsconcomitantwiththedisclosureofthematerialweakness.Forinstance,33firmshadrestatements,and38firmshadchangesinauditors(notshowninthetables).Noteworthy,12oftheauditorswitcheswerefromBig4firmstonon-Big4firms.Thus,71firms(44%ofthesample)hadadditionalreportableevents.Wetestedtoseeiffirmswithmultipleeventsdifferedfromfirmswithonlymaterialweaknesses.Asbefore,theresultsdependupontheinclusionorexclusionofFannieMae.IncludingFannieMae,onlyTotalAssetsaresignificantlydifferentbetweenthetwosub-samples(firmswithmultipleeventsarelargerintermsofassets).ExcludingFannieMae,wefindnoevidenceofanystatisticallysignificantsystematicdifferenceinanyofthesummarymeasuresshowninTable3.Insummary,firmswithmultipleeventsdisclosedtheirmaterialweaknessesinconjunctionwiththeauditorswitchorrestatement.Thereasons,however,fortheothersub-19 sampleoffirmstodisclosetheirmaterialweaknessesisnotreadilyapparent.Weconjecturethattheauditorsandmanagementrecognizedthatthematerialweaknesscouldnotberemediedinatimelyfashion(bythefirstmandateddisclosuredate)andthereforedecidedtodiscloseimmediately.Otherwise,adelayeddisclosurecouldinvitelitigation.EarningsManagementoutsideofSection404:TheCaseofEastmanKodakAccordingtoThomsonFinancial/FirstCall,EastmanKodak(EK)had“actualearnings”of$0.78pershareforthequarterendingDecember31,2004.Theconsensusestimateforthequarterwas$0.65,apositivesurpriseof$0.13(or20.0%).However,the“actualearnings”accordingtoGAAPwere-$0.04(dilutedEPSfromcontinuingoperations).FirstCallhadreported,not“GAAPactual,”butrather“proforma”EPS,whichtheynonethelesscalled“actual.”Firmsareallowedtoreportproformaearnings(thatremovecertainitemsfromearnings,usuallywhatfirmsdeemtobeunusual,non-recurring,ornon-operatingitems)aslongastheyarenotmisleadingandreconciletheproformaresultstoGAAPresultsunder“RegG.”Accordingly,EKprovidesthereconciliation,butFirstCalldoesnotreportthatitisdisseminatingproformaresults.GAAPactualnetearningsfromContinuingOpswere-$17million.EKaddedback$253milliontoarriveatitsproformaearningsfromContinuingOpsor$236million(or$0.78pershare).EKandmanyotherfirmshaveforalongtimeguidedtheinvestingcommunityusingproformanumbers.Thistypeofguidancerepresentsakindof“manageddisclosuresystem”thatfallsoutsideofGAAP.Moresignificantly,italsofallsoutsideofSection404,whichgovernstheintegrityofthe“GAAPactual”resultof-$0.04,notthe“proforma”resultof$0.78.Someliteratureclaimsthattheproformaresultsaremorevalue-relevant(BrownandSivakumar,2003).Iftrue,anaturalquestionarisesaboutspendingresourcestoimprovetheGAAPmeasurementanddisclosureprocess,particularlywithrespecttoearnings.EKalsoannouncedthatithadaccountingerrorsandmaterialweaknessesininternalcontroloverfinancialreporting.Inan8-Kfilingdated1/26/05,EKwrote:EKannouncedthematerialweaknessafteroursampleperiodandthereforeisnotincludedinthestatisticsreportedinthetables.WewillcontinuetoupdateoursampleonasystematicbasisTheerrorsareconfinedtoincometaxaccounting,andassuchtheydonotaffectthecompany"sbusinessoperations,"saidRobertH.Brust,Kodak"sChiefFinancialOfficer.19 “RememberthatKodakhasbeendevotingsignificantresourcesformorethanayeartoassessing,andstrengtheningasappropriate,itscontrolsinthecontextofitsSarbanes-OxleySection404review,”Brustsaid."Thissituationarisesfromtaxaccountingerrors,notmisconduct.Itinvolvescomplextaxrules,inmanycasesrelatingtoourrestructuringactionsoverseas,thatvarybycountry."Kodak"smanagement,inconjunctionwithexternalconsultants,iscurrentlyanalyzingitsincometaxaccounts,andadjustmentsmayresultfromthisreview,"Brustsaid.Asaresultoftheincometaxaccountingerrors,thecompanyhasdeterminedthatithasaninternalcontroldeficiencythatconstitutesa"materialweakness,"asdefinedbythePublicCompanyAccountingOversightBoard"sAuditingStandardNo.2.Consequently,managementwillbeunabletoconcludethatthecompany"sinternalcontrolsoverfinancialreportingareeffectiveasofDec.31,2004.Therefore,PricewaterhouseCooperswillissueanadverseopinionwithrespecttothecompany"sinternalcontrolsoverfinancialreporting.Anassessmentofthecompany"sinternalcontrolswillbeincludedinitsAnnualReportonForm10-K,whichwillbefiledinMarch.EK’sstockpriceroseonthedayofandinthedaysafterthemultipleannouncementsofnegativeearnings,accountingerrors,andweaknessesininternalcontrol.ThemarketsfocusedinsteadonEK’sproformaresultsToconcludetheexampleofEastmanKodak,weunderscorethatthemanageddisclosuresystemremainsuncheckedbySOXSection404.ItisevidentthatEKguidesanalystsusingproformaearningsandanalystsusetheguidedproformanumbersinsettingtheirexpectations.(BryanandLilien,2004)EventhoughSection404mayreducemanagementofGAAPearnings,firmswilllikelycontinuetoengagein“legal”earningsmanagementusingproformaearnings(solongasitisnotintentionallymisleading).Thus,themarketimplicationsofaccountingerrorsandmaterialweaknessesmaybediminishedsimplybecausefirmsemphasizealternativenumbersanyway.ConclusionsBaseduponearlyevidenceoffirmswithmaterialweaknesses,asdefinedbythePublicAccountingOversightBoard,wefindthatfirmswithmaterialweaknessesare,onaverage,bothsmallerandworseperformersthantheirmatchedindustrycounterparts.Wealsonoticeadichotomyinthesampleoffirmswithmaterialweaknesses—thosewithjustmaterialweaknessesandthosewithmultipleevents,suchasauditorchangesandrestatements.Even19 thoughthematerialweaknessfirmsaresmalleronaverage,thereareoutliers,suchasFannieMae.Thelargestfirmswithmaterialweaknesses,includingFannieMae,arefirmswithmultipleevents.Wealsofindthatfirmswithmaterialweaknesses,onaverage,havehigherbetas,suggestingahigherdiscountbythemarketforthesefirms.Finally,wedocumentnegativestockreturnsonthedateoftheannouncementofthematerialweaknessbutinsignificantreturnsoveranarrowinterval.Takentogether,thesefindingssuggestthatlargerfirmsonaveragehavetheresourcestodedicatetoeffectiveinternalcontrolswhereassmallerfirmsdonot,sincesmallerfirms,onaverage,arereportingmaterialweaknesses.However,compliancewithSOXhasrequiredandwillcontinuetorequiresignificantresources.Iferrorsarenotdiscovered,oriftheyarediscoveredpredominantlyinsmallerfirms,thenonecouldquestionwhatthebenefitsarebothtothefirmandtosocietyatlarge.Moreover,inthepast,beforeSOX,errorswerecaughtandfirms,bothlargeandsmall,wereforcedtorestate.Itcouldthereforebejustacaseoftheauditfunctionnotperforminguptosufficientlevelsinselectcasesandnotaneedtorevamptheinternalcontrolsystemforallpubliccompanies.Inaddition,thehighervolatility,assuggestedbybeta,suggeststhatthemarketdiscountsthesefirmswithahighercostofcapital,inpart,perhapsforgreaterriskassociatedwithlesseffectiveinternalcontroloverthefinancialreportingprocess.Finally,withrespecttoallegedbenefitsofSOX,wecannotdetermineormeasurewhetherbehind-the-scenesexternalitiesaccruetofirmsastheyimplementmoreinternalcontrolsoverfinancialreporting.AlthoughonestudydocumentsareductioninearningsmanagementsincethepassageofSOX,SOXdoesnotdirectlyaddressearningsmanagement.However,astheEKcaseillustrates,firmshaveothermeansoutsidetheformalreportingprocesstomanageearningsandexpectations.WhetherSOXcanandwilldeteroutrightfraudanymorethanthecapitalmarketsandothermonitoringandcontractingmechanismsisanongoingdebateintheliterature.Further,whetherSOXwilladdvaluetocorporationsthroughbetteroperations,reducedcostofcapital,orothermeansremainstobeseen.Atthisjuncture,thereisnodirectlymeasurableevidenceofthesebenefits.Withoutbeingabletodeterminebenefits,wequestionwhetherburdeningallpublicfirmswithSOXinternalcontrolcompliancewillhavebeenworththecost.Evenareductioninearningsmanagementisofquestionablevaluewhenfirmsdiscloseproformaearnings.Whatmayremainisanexpensiveidentificationofpredominantlysmallfirmswithinternalcontrolproblems.Inasurveyof224firmsconductedbyFinancialExecutives19 International(inJuly2004),respondentsrevealedthattheywillspendinthefirstyearofSOX,onaverage,anextra$3millioninordertocomplywithSection404.Thelargestcompanies(thosewithover$5,000millioninrevenues)willspendanaverageof$8million.Respondentsalsosaidthattheywillspendapproximately73%ofthemoneyonnon-internalauditstaff,suggestinganextrawealthtransferfromshareholderstotheexternalauditors.Infact,respondentsstatedthatauditfeesareexpectedtoincrease,onaverage,by53%inordertopayfortheattestationoverinternalcontrol.Futureresearchdirectionsincludesurveysoffirms,notunliketheworkbyFEI(2004),toattempttomeasure,notonlythecosts,butalsowhatfirmsperceiveasbenefits.Additionally,whilewecontinuetocollectdataonareal-timebasis,theevidencethusfarcontinuestosupporttheinitialfindingsonthefirmcharacteristics.Withmoreobservations,however,includingfullfinancialdata,furtheravenuesforresearchincludeexploringlinkagesbetweenmaterialweaknessfirmsandothermeasuresofcorporategovernance.Wealsodonotobservethusfarthatanyoftheidentifiedmaterialweaknessespertaintotheestimationprocessesthatfirmsmustutilizetomeasurecertainaccountingnumbers.Wethereforedonotknowiforhowimprovedinternalcontrol,including“toneatthetop,”linkstobetterestimatesandthereforereducedearningsmanagement.Drillingdownintotheexactnatureofthematerialweaknesseswouldthereforebeuseful.ReferenceAlles,MichalG.andSrikantDatar.HowDoYouStopTheBooksFromBeingCooked?AManagementControlPerspectiveonFinancialAccountingStandardSettingandtheSection404RequirementsoftheSarbanes/OxleyAct.InternationalJournalofDisclosureandGovernance,Version4:December2003.Bulkeley,William.M.andRobertTomsho.Kodak,getauditors"adverseview.WallStreetJournal,January27,2005.Bratton,WilliamW.,Enron,Sarbanes-OxleyandAccounting:RulesversusPrinciplesversusRents.VillanovaUniversitySchoolofLaw"sLawReviewSymposiumIssue:LessonsfromEnron;HowdidCorporateandSecuritiesLawFail?Volume48,Number4,2003.Brown,awrence,andK.Sivakumar.ComparingtheValueRevenueofTwoOperatingIncomeMeasures.ReviewofAccountingStudies8(December2003):561-572.Bryan,Stephen,andStevenLilien,TheManagedDisclosureSystem,CPAJournal,March2004.Cohen,DanielA.,AiyeshaDev,andThomasZLys.TrendsinEarningsManagementandInformativenessofEarningsAnnouncementsinthePre-andPost-SOXPeriods.Northwestern19 University,February,2005.DeloitteandTouche.UnderControl:SustainingCompliancewithSarbanes-OxleyinYearTwoandBeyond,2005.Engel,Ellen,RachelHayes,andXueWang.TheSarbanes-OxleyActandFirms"Going-PrivateDecisions.UniversityofChicago,May2004.FinancialExecutivesInternational.SurveyonSOXSection404Implementation.July2004.Gompers,PaulA.,JoyIshii,andAndrewMetrick.CorporateGovernanceandEquityPrices.QuarterlyJournalofEconomics,February2003.Gordon,JeffreyN.GovernanceFailuresoftheEnronBoardandtheNewInformationOrderofSarbanes-Oxley.ColumbiaLawSchool,March2003.Heflin,Frank,andCharlesHsu,TheImpactoftheSarbanes-OxleyActof2002ontheUseofNon-GAAPEarningsMeasures.PurdueUniversity,January2004.Li,Haidan,MortonPincus,SonjaOlhoftRego.MarketReactiontoEventsSurroundingtheSarbanes-OxleyActof2002.UniversityofIowa,November2004.McKinsey&Co.McKinseyInvestorOpinionSurveyonCorporateGovernance.June2000.Moody’sInvestorServices.Section404ReportsonInternalControl:ImpactonRatingsWillDependonNatureofMaterialWeaknessesReported,October2004.Perino,MichaelA.Enron’sLegislativeAftermath:SomeReflectionsontheDeterrenceAspectsoftheSarbanes-OxleyActof2002;ColumbiaLawSchool,TheCenterforLawandEconomicStudies,WorkingPaperNo.212,October2002.Ribstein,LarryE.Marketvs.RegulatoryResponsestoCorporateFraud:ACritiqueoftheSarbanes-OxleyActof2002.IllinoisLawandEconomicWorkingPapersSeries,WorkingPaperNo.LE02-008,September2002.SecuritiesandExchangeCommission.ExemptiveOrderonManagement"sReportonInternalControloverFinancialReportingandRelatedAuditorReport:FrequentlyAskedQuestions,January21,200519 19'