盈余质量与盈余的定价模式的影响外文翻译.docx

盈余质量与盈余的定价模式的影响外文翻译.docx

《盈余质量与盈余的定价模式的影响外文翻译.docx》由会员分享，可在线阅读，更多相关《盈余质量与盈余的定价模式的影响外文翻译.docx（11页珍藏版）》请在冰豆网上搜索。

盈余质量与盈余的定价模式的影响外文翻译

原文：

EarningsQualityandthePricingEffectsofEarningsPatterns

EarningsPatternsandEarningsQuality

Previousresearchonpricingeffectsofearningspatternshasconsideredincreasingearnings,earningsthatmeetorexceedanalystforecasts,andsmoothearnings.Theresearchismotivatedbytheobservationthatmanagersappeartofocusonmaintainingsuchpatterns.Webuildonthisobservationbyexploringthenotionthatmanagers’accrualschoiceseitherreinforceorundercutthepricingeffectsofearningspatterns.Inthissection,wedescribepreviousresearch（section2.1）,discusswhythethreeearningspatternsweconsidermightbeeconomicallydistinct,asopposedtomanifestationsofasingleunderlyingconstruct（section2.2）,anddiscussthelinksbetweenearningspatternsandearningsquality（section2.3）.

1、Previousresearchonpricingeffectsofearningspatterns

Beginningwiththeobservationthatmanymanagersappeartostrivetoreportsteadilyincreasingearnings,Barth,ElliottandFinn[1999]documentrewardstosuchbehavior:

firmswithincreasingearningshavehigherprice-earningsmultiplesthanotherfirms,aftercontrollingforgrowthandrisk.Thecontrolforgrowthisbasedonbothincreasesinbookvalueofequityandanalystearningsgrowthforecasts,andthecontrolforriskisthevarianceofthemostrecentsixyears’percentageearningschanges.Theyattempttoruleoutthepossibilitythatfirmswithlongpatternsofincreasingearningsalsoshareavaluationrelevantfactorthatisknownbeforetheearningspatterndevelops;resultsaremixedalthoughtheauthorsconcludethattheweightoftheevidencedoesnotsupporttheexistenceofsuchafactor.Barthetal.donotconsiderwhetherthequalityofearningsmightaffecttheirresults,althoughtheynotethatifearningsincreasesareobtainedviaearningsmanagementwhichinvestorsdiscernanddiscount,theirtestswillbebiasedagainstfindingpricingeffectsassociatedwithpatternsofincreasingearnings（seetheirnote6）.

KasznikandMcNichols[2002]andBartov,GivolyandHayn[2002]focusonmarketeffectsofmeetingorexceedinganalystforecasts.KasznikandMcNicholsdocumentthatconsistentlymeetingorexceedinganalystforecastsisassociatedwithahighermultiplerelatingearningstoprice,andthispricingeffectisduetotwodistinctsub-effects:

higherfutureearningsandapremiumassociatedwiththeactofmeetingorexceedingforecastsforatleastthreeyears.Theyinterprettheirresultsasindicatingthatconsistentlymeetingorexceedinganalystexpectationsconnoteslowerriskand,therefore,ahigherearningsmultiple.However,theyalsonotetwoopenissues.First,theydonotinvestigatewhyinvestorswouldassociatelowerriskwithconsistentlymeetinganalystforecasts.Second,theydonotinvestigatetheroleofmanipulatingeitherexpectationsofearningsorearningsthemselves.Webelieveouranalysesshedlightontheseissues;specifically,wepositthatinvestorspricefirmsthatmeetorexceedanalystearningsexpectationsandhavehighearningsqualityhigherthanfirmsthatmeetthefirstbutnotthesecondcondition.

UsingadifferentresearchdesignthanKasznikandMcNichols,Bartovetal.[2002]documenthigherquarterlyabnormalreturnsinquarterswherefirmsmeetorexceedforecasts.Severaloftheirtestsdistinguishbetweenshiftsinanalystforecastsoverthequarterandtheearningssurprise,basedonIBESreportedearningscomparedtothemostrecentIBESforecast.Theyinterpretthefindingthatthereturntoanearningssurprisegreatlyexceedsthereturntoshiftsinanalystforecastsasrationalizingexpectationsmanagement—themarketpenaltyfordampeninganalystexpectationsisswampedbytherewardtoafavorableearningssurprise.Subsequenttestssupportthisinterpretation.

Mostdirectlyrelevanttoouranalyses,Bartovetal.testwhetherthepremiumformeetingorexceedinganalystforecastsissmallerwhenthereisevidenceofeitherorbothexpectationsmanagementandearningsmanagement.Theydefinetheformerasthecombinationofnegativeanalystforecastrevisionsandzeroorpositiveearningssurprises,andthelatterasthepresenceofunexpectedaccrualsbasedeitherontheJones[1991]modelasmodifiedbyDechow,SloanandSweeney[1995]orondefiningnormalaccrualsasworkingcapitalaccrualsplusdepreciationandamortization.Theyconcludethatbothexpectationsmanagementandearningsmanagementareassociatedwithstatisticallyreliablebutsmalldecreasesinthemarketpremiumtomeetingorexceedinganalystexpectations,andthatthe“economicallyminor”（p.198）effectofearningsmanagementmightbeduetolowpower.Weextendtheseanalysesbyprovidingdirecttestsofwhetherthepricingeffectsofconsistentlymeetingorexceedinganalystforecastsareafunctionofearningsquality,aninversemeasureofearningsmanagement.

ThomasandZhang[2002]reporthigherprice-earningsmultiplesforfirmswithsmoothearnings.Theirmeasuresofearningssmoothnessarederivedfromthestandarderrorsofregressionsofearningsontimeandfromthestandarddeviationsofseasonallydifferencedquarterlyearningspershare.Theyconditiononintrinsicearningsvolatility（proxiedbytheover-timevariabilityofanalystearningsforecasts）,tocapturethevaluationeffectsofmanagerialinterventionstosmoothearnings.Resultsindicatethatbothlowerintrinsicearningsvolatilityandearningssmoothnessareassociatedwithhigherprice-earningsratiosandthatthisrelationisdueatleastinparttoanassociationbetweenlowerriskandlowerintrinsicearningsvolatility.

LikeThomasandZhang,Hunt,MoyerandShevlin[2000]reportthatearningsmultiplesincreasewithearningssmoothness.Theirmeasuresofsmoothnessarederivedfromaseparationofearningsintocashandtotalaccruals,andtotalaccrualsintoitsdiscretionaryandnondiscretionarycomponents,usingtheJones[1991]modelasmodifiedbyDechow,SloanandSweeney[1995].Controllingforthestandarddeviationofcashflows（presumablyasaproxyforunsmoothedearnings）Huntetal.assesstheincrementaleffectsonpricingmultipleswhen

（1）thestandarddeviationofcashplusnondiscretionaryaccrualsissmallrelativetothestandarddeviationofcashflowsand

（2）thestandarddeviationofreportednetincomeissmallrelativetothestandarddeviationofcashplusnondiscretionaryaccruals.Theformermeasureisinterpretedassmoothnessinducedbynondiscretionaryaccrualsandthelatter,assmoothnessinducedbydiscretionaryaccruals.Huntetal.concludethatdiscretionarysmoothnessisassociatedwithalargerpositivepricingeffectthannondiscretionarysmoothness,andthatdiscretionarysmoothnessisalsoassociatedwithgreaterearningspersistence.

Takentogether,priorresearchsuggeststwoempiricalregularitieswhichweexplore.First,thereisaconsistentpositivepricingeffectassociatedwithearningsthataresmooth,earningsthatincrease,andearningsthatmeetorexceedanalystforecasts.Previousresearchhasconsideredthesepatternsoneatatimeandunconditionally.Weconsidertheeffectsofeachearningspatternconditionalontheothers,toanalyzewhethertheeffectsareinfactdistinct,asopposedtovaryingmanifestationsofasingleunderlyingfactor.Second,thereis（mixed）evidencesuggestingthatthepricingeffectsofearningspatternsmightbeconditionalonwhetherearningsaremanaged.Theeffectsofearningsmanagementappeartobesmall,perhapsbecauseoflowpowertests.Weprovidedirectassessmentsoftheeffectsofearningsquality（aninversemeasureofearningsmanagement）.

2、Relationsamongearningspatterns

Apatternofincreasingearningssimplyimpliesthat,overtime,eachsuccessiveearningsnumberexceedsthatofthepreviousperiod.Thereisnonecessaryimplicationof,orconflictwith,smoothness,measuredeitherintermsofovertimeearningsvariability（i.e.,earningsmighttrendupwardeithersmoothlyorbyvaryingamountseachperiod）orintermsoflowreportedincomevariabilityrelativetointrinsicincomevariability.Thereisalsononecessaryimplicationofmeeting,ornotmeeting,analystexpectations.Intermsofmanagement’sbehavior,earningsincreasescanbeachievedbysomecombinationofintrinsicgrowthandearningsmanagementviaeitherorbothaccrualsmanipulationsandrealactions.Dependingonmanagement’splanninghorizonandtheunderlyingeconomicsofthefirm,earningsincreasescanbeplannedandmanipulatedoverlongperiodsbysystematicallybuildingandthenusing“reserves”ofaccruals—thatis,someportionofunmanagedcurrentperiodearningscanbedelayed,viaaccrualsmanagement,andincorporatedintolaterearnings,asneeded.However,onceanearningsnumberisannounced,itbecomesafixedtargetthatmustbeexceededifthepatternofincreasingearningsistobemaintained.

Incontrast,patternsofexceedinganalystforecastscanbeachievedbyintrinsicperformance,earningsmanagementandexpectationsmanagement.Evidenceofexpectationsmanagementisprovidedby,forexample,Bartovetal.[2002]andMatsumoto[2002].2Totheextentmanagerscananddoguideanalyststowardaforecastthatwillensureazeroorpositiveearningssurprise,theovertimepatternofearningsthatmeetorexceedanalystforecastscantakeanyformofincreasesanddecreases,canexhibitanylevelofsmoothness,andcanbeentirelyunmanaged.Thus,thereisnonecessaryrelationbetweenmeetingorexceedinganalystforecastsandanyotherearningspattern,orbetweenthisearningspatternandearningsquality.

Finally,earningssmoothnessistypicallyassessedbycalibratingearningsvariabilityovertime,relativetothevariabilityofso