Government assisted earnings management in China

2024-10-18 来源：威能网

Available online at www.sciencedirect.com

JournalofAccountingandPublicPolicy27(2008)262–274

GovernmentassistedearningsmanagementinChinaqXiaoChena,Chi-WenJevonsLeeb,c,*,JingLidbSchoolofEconomicsandManagement,TsinghuaUniversity,China

A.B.FreemanSchoolofBusiness,TulaneUniversity,NewOrleans,LA70118-5669,USA

cGraduateInstituteofBusiness,ZhejiangUniversity,ChinadGraduateSchoolofBusiness,ColumbiaUniversity,USA

aAbstract

AsthesocialistsysteminChinaembracesthemarketeconomy,ithascreatedmanyconﬂictsofinterestsandcollusionbetweenﬁrmsanddiﬀerentlayersofgovernments.ThecentralgovernmentinChinasetsregulationstoensurethequalityofﬁrmslistedinthecapitalmarket,whilelocalgovern-mentsengageininter-jurisdictionalcompetitionformorecapital,andtheirinterestsarealignedwithlistedﬁrmsthroughthestringentIPOquotasystem.ThispaperexamineshowlocalgovernmentsinChinahelplistedﬁrmsinearningsmanagementtocircumventthecentralgovernment’sregulation.Weﬁndthatlocalgovernmentsprovidesubsidiestohelpﬁrmsboosttheirearningsabovethereg-ulatorythresholdofrightsoﬀeringanddelisting.Moreover,thiscollusionbetweengovernmentandlistedﬁrmsinearningsmanagementexistsmainlyinﬁrmscontrolledbylocalgovernments.Ó2008ElsevierInc.Allrightsreserved.

JELclassiﬁcation:C71;G32;G38;H71;M41

Keywords:Earningsmanagement;Collusion;Localgovernment;Governmentsubsidy;Regulation;State-ownedenterprise(SOE)

WeappreciatecommentsfromKevinChen,JosephP.H.Fan,ZhaoyangGu,DonHerman,CharlesLee,TerryShevlin,T.J.Wong,membersofworkshopsatTsinghuaUniversity,andparticipantsatthe2003AmericanAccountingAssociationAnnualMeetingsinHawaii,2004ChinaAccountingResearchIncubatoratHongKongUniversityofScienceandTechnologyand2004AAANZConferenceinAliceSpring,Australia.*Correspondingauthor.Address:A.B.FreemanSchoolofBusiness,TulaneUniversity,NewOrleans,LA70118-5669,USA.Tel.:+1(504)8628485;fax:+1(504)8656751.E-mailaddress:jevons.lee@tulane.edu(Chi-WenJevonsLee).0278-4254/$-seefrontmatterÓ2008ElsevierInc.Allrightsreserved.doi:10.1016/j.jaccpubpol.2008.02.005

qX.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274263

1.Introduction

Earningsmanagementisawell-documentedaccountingphenomenonwheremanagersusejudgmentinﬁnancialreportingorstructuretransactionstomisrepresenttheunderly-ingeconomicperformanceofacompanyortoinﬂuencecontractualoutcomesthatdependonreportedaccountingnumbers.1Inthispaperweidentifyaspecialcaseofearningsman-agementinlistedﬁrmsinChinawherelocalgovernmentsprovidesubsidiestoassistlocallistedSOEs(state-ownedenterprises)tomeetthecentralgovernment’sregulatoryrequire-mentsforrightsoﬀeringanddelisting.

ChinaestablishedstockmarketsinShanghaiandShenzheninearly1990stofacilitateeconomicreform.MostlistedﬁrmsaretransformedfromSOEs.However,thecentralgov-ernmentdesignstheregulatorystructureinordertomaintainastronggripthroughastringentquotasystem.Accordingly,followingtheirinitialpublicoﬀering(IPO),listedﬁrmsareanxioustoraisemorecapitalfromthemarket.TheprimarysourceofadditionalﬁnancingforﬁrmslistedontheChinesestockexchangeistheissuanceofstockrights(rightsoﬀering).2Inaddition,maintainingthelistingstatusensuresaccesstothecapitalmarket.Therefore,astheChineseSecuritiesRegulatoryCommission(CSRC)setstoughregulationsfortheproﬁtabilityrequirementsforrightsoﬀeringanddelisting,listedﬁrmshavestrongincentivestomanipulateaccountingearningstomeettheregulatoryrequire-ment.PriorliteraturehasdocumentedtherampantearningsmanagementphenomenoninChineselistedﬁrmsstimulatedbyCSRC’sregulations(ChenandYuan,2004;Hawetal.,2005).Thesestudiesalsoconcludethatrealtransactionsthroughnon-operatingearningsorbelow-the-lineitemsaretheprimarymeansofearningsmanagementinChina.

Inthispaper,weparticularlyfocusongovernmentsubsidies,whichcompaniesmustlistasoneofthebelow-the-lineitemsontheirincomestatements.SpeciﬁcinvestigationoftheroleofgovernmentsubsidiesinearningsmanagementwarrantsincreasedinterestasitcastslightontheentangledrelationshipbetweenSOEsandvariouslayersofgovernmentinChina’semergingcapitalmarket.EarningsmanagementinChinaisnotsimplyaman-agementchoice,butinvolvescollusivemanipulationbylocalgovernmentsagainstthecen-tralgovernment.Byexamininggovernmentsubsidies,wealsoshowthatearningsmanagementinSOEsmayarisefromtheconﬂictofthepolitician’sroleasthegovernmentagentandastheeconomicagentoftheﬁrm’slargeshareholder.

Thecentralgovernment’sprimaryobjectiveinregulatingthestockmarketistoprotecttheinvestorsandensurethequalityoflistedcompanies.LocalgovernmentsaredelegatedwiththerightsofownershipandtheappointmentofCEOsinSOEsundertheirownjuris-dictions.3UndertheIPOquotasystem,eachlocalgovernmentcanselectonlylimitednumbersofﬁrmstogopublic.HencelistedSOEsareusuallythelargestcompaniesintheir

SeeSchipper(1989),HealyandWahlen(1999),andDechowandSkinner(2000)fordetailsandreferencesonearningmanagement.2Rightsoﬀering,oﬀeredtoexistingshareholdersatadiscount,isdiﬀerentfromSEOs(seasonalequityoﬀering).SEOisopentogeneralpublicanddoesnotinvolveadiscount.RightsoﬀeringisanexperimentalstepforSEO.Notuntil2001wastheSEOallowedinChina.3China’sgovernmentadministrationhasﬁvelevels:(1)central;(2)provincial;(3)prefecture;(4)county;and(5)township.InourpaperwerefertoprovincialgovernmentsaslocalgovernmentsbecausetheIPOquotasareallocatedtoprovinciallevelonly.Besidesﬁrmscontrolledbylocalgovernments,asmallpercentageoflistingquota(about5%)isremainedwithcentralgovernment.Eventhere,astateadministrativebureauisincontrolofthelistedcentralSOEsonbehalfofthecentralgovernment.

1264X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274

localesandhavegreatinﬂuenceonthelocaleconomy.Atthesametime,theﬁscaldecen-tralizationreforminthe1990sinduceslocalgovernmentstocompetewitheachotherformobilecapitalinordertodeveloplocaleconomies.4Thestockmarketprovideslocalgov-ernmentsanopportunitytoattractmoreequitycapitalintotheirownjurisdictionswith-outincreasinglocalpublicexpenditure.Thereforelocalgovernmentsandlistedﬁrmsarealignedthroughacommoninterestinraisingmorecapitalfromstockmarkets.Byhelpinglistedﬁrmstoacquireadditionalcapitalfromthestockmarket,thelocalgovernmentreapsthebeneﬁtsofnewinvestmentsandeconomicgrowth.Moreover,governmentoﬃ-cialsmayhavepersonalincentivestosupportlistedﬁrmsastheperformanceoflocalecon-omyalsoaﬀectstheirpromotionthroughthebureaucraticranks(Li,1998).

Localgovernmentshavealternativewaystosupportlistedﬁrms,suchasgrantingtax-ationpreferenceorfavoringlistedﬁrmsintheprojectapprovalprocess.However,thegovernmentsubsidyisamoredirectandconvenientwaytohelplistedﬁrmsinearningsmanagement.TheﬁscalpolicyongovernmentsubsidytoSOEsissetbytheChinesecen-tralgovernment.Accordingtothepolicy,theprimarygoalofsubsidyistodevelopstra-tegicareassuchasagriculture,publicutilities,andhigh-techindustries,andtohelpﬁnanciallydistressedSOEs.LocalgovernmentshavethediscretiontodeterminewhethertograntsubsidiestoSOEsintheirownjurisdictions.Iflocalgovernmentshavenoincentivestoengageinearningsmanagementoflistedﬁrms,weexpectthatonlyﬁrmsthatmeettheﬁscalpolicyrequirementswillreceivegovernmentsubsidy.Wehypothesizethatgiventheirstrongincentivestohelplocallistedﬁrms,localgovernmentsmayusesubsidytohelplistedﬁrmsinneedofearningsmanagementtosurpasstheregulationthreshold.

Oursampleconsistsofalllistedﬁrmsfrom1994to2000andcoversthreemajorregu-lationperiods:thenoregulationregime(1994–1995),thestringentregulationregime(1996–1997)andtherevisedregulationregime(1998–2000).WeshowthatﬁrmsaremorelikelytoreceivesubsidiesfromgovernmentsiftheirROE-before-subsidyisslightlybelowtheregulationhurdlesofrightsoﬀeringanddelisting.Weﬁndevidenceofearningsman-agementthroughgovernmentsubsidiesinboththestringentregulationandrevisedregu-lationregimes,butnotinthenoregulationregime.Wefurtherexaminetheownershipstructure,andﬁndthatgovernmentassistedearningsmanagementexistsmerelyinthoseSOEscontrolledbylocalgovernments,andthattheSOEscontrolledbycentralgovern-mentandthejointventureﬁrmscontrolledbymultinationalﬁrmsdonotgetsuchfavorfromthelocalgovernment.

Governmentsubsidiestostateenterprisesarenotsurprisinginasocialisteconomyandoftencitedasanissueof‘soft-budgetconstraint’underthegovernment’sownership(BaiandWang,1998).Ourresultssuggestthatthemeddlinghandofgovernmentstillexistsintheemergingcapitalmarket,eventhoughmanyeconomistsviewtheSOEsgoingpublicasanimportantprocessofprivatizationintheeconomyreforminChina.Weobservethatthecentralgovernmentstronglycontrolsthestockmarketthroughthequotasystemandtightregulations,andthatlocalgovernmentscolludewithlistedﬁrmsfortheinterestoftheirownjurisdictions.Thegovernment’s‘‘visiblehand”inﬂuencestheresourcealloca-tioninthestockmarket.

Theinter-jurisdictionalcapitalcompetitionthroughtaxcompetitionhasbeenstudiedinthelocalpubliceconomicliteraturefollowingTiebout(1956).

4X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274265

Ourstudyhasimplicationsforglobalinvestors,whoareincreasinglyinvestinginemergingmarketslikeChina,andotherusersofﬁnancialreportsofChineselistedcom-panies.Ourresultshighlighttheimportanceofunderstandinginstitutionalfactorswhenanalyzingtheﬁnancialinformationintheemergingmarket(BushmanandSmith,2001).TounderstandtheeconomicbehaviorofChineseﬁrmsandChinesestockmarket,onemustﬁrstunderstandtheroleofgovernmentanditsinﬂuenceontheﬁrm’seconomicactivities.

Theremainderofthispaperisorganizedasfollows:Section2describestheinstitutionalbackgroundoftheSOEreformandtheCSRC’sregulationofrightsoﬀeringanddelisting.Section3developsthegovernmentassistedearningsmanagementhypothesis.Section4presentstheempiricalresearchdesignandresults.Thesummaryandconclusionarepre-sentedinSection5.

2.Institutionalbackground

TheeconomicreforminChinashiftsSOEﬁnancingfromthegovernmenttothemar-ket.TheChinesestockmarketwasestablishedasavehicletohelpﬁnanceSOEsinanopenmarketandtograntSOEsmoreautonomyasindependenteconomicentities.5Asanemergingmarketinatransitionaleconomy,theChinesestockmarketexhibitstworatheruniquecharacteristics:(1)atightquotasystemdesignedtoassureastablemarketinitsinfancy;and(2)awiderangeofstocktransactionsrequirevariousgovernmentapprovalsandareregulatedbymultiplelayersofgovernmentalagencies.ThecentralgovernmentsetsannualquotasforIPOsinthestockmarket.6Theannualquotaisallocatedamongtheprovincesandstateadministrativebureausthathavetheauthoritytodictatethedis-tributionofsharesforpublicoﬀeringsamongSOEsintheirjurisdictions.Eachapprovalforpublicoﬀeringwentthroughacomplicatedadministrativereviewprocess,withanemphasisontheprospectiveproﬁtability.LocalgovernmentsmakemaximumuseoftheirIPOquotasbyrestructuringlargeandproﬁtablelocalSOEstotakethempublic.Mostlistedﬁrmswerecarve-outsfromtheirparentSOEsandthegovernmentstillcontrolsthemajorityoftheirshares.

Thisstudyfocusesonrightsoﬀeringsanddelistingregulations.ForlistedﬁrmsinChina,theissuanceofstockrightsistheprimarysourceofadditionalﬁnancingaftergoingpublic.7In1993theCSRCguidelinesrequiredtwosuccessiveyearsofproﬁtforaﬁrmtobeeligibletomakearightsoﬀering.Thecriteriabecamemorestringentin1994byrequir-ingrightsoﬀeringcompaniestohavethreesuccessiveyearsofpositiveearningsandanaveragereturnonequity(ROE)of10%.TheROErequirementwastightenedin1996,requiringa10%ROEineachofthethreepreviousyears.The1996guidelinessigniﬁcantlyreducedthenumberofeligiblecompanies,butledtotheoutbreakoframpantearnings

SeeLee(1987)foranalysisofmakingcapitalmarketasaninstitutionalinfrastructureforeconomicdevelopment.SeeLee(2001)fordetailsandreferencesoneconomictransitionandSOEreform.6TheIPOquotasystemwasexplicitlyabolishedin2001,buttheIPOmarketisstillundertightcontroloftheCSRC.Oursampleperiodfrom1994to2000isstillunderperiodwithIPOquota.7Stockrightsoﬀeringsareoﬀeredtoexistingshareholdersatadiscount,diﬀerentfromSEOs(seasonalequityoﬀering).SEOisopentogeneralpublicanddoesnotinvolveadiscount.RightoﬀeringisanexperimentalstepforSEO.Notuntil2001wastheSEOallowedinChina.

5266X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274

management.Inearly1999theCSRCreducedtheminimumROEto6%,whichalleviatedsomeofthemotivationforearningsmanagement.

Noregulationsconcerningdelistingwereissueduntil1998,whentheCSRCissuedtheSpecialTreatment(ST)andParticularTransfer(PT)regulations.TheSTregulationrequiresanylistedﬁrmwithtwosuccessiveyearsoflossesorwithanassetvaluepersharethatislessthanthestock’sfacevaluebedesignatedasanSTﬁrmonthestockexchange.Moreover,theregulationrequiresanylistedﬁrmwithlossesforthreesucces-siveyearsbedesignatedasaPTﬁrm,whichcanbedelistedfromthestockexchangebytheCSRC.83.Governmentassistedearningsmanagement

TheCSRC’srigidregulationbasedonaccountingnumbershascreatedaclusterofﬁrmswithROEsthatareclosetotheregulatoryborderline,astrongsignofwide-spreadearningsmanagement.9Withrigidrule-basedaccountingstandards,earningsmanagementthroughaccountingmethodchoiceanddiscretionaryaccrualsisrareinChina.Priorstud-ieshaveconcludedthatrealtransactionsinvolvingnon-operatingearningsorbelow-the-lineitemsistheprimarymeansofearningsmanagementinChina(ChenandYuan,2004;Hawetal.,2005).

MostlistedﬁrmsareformerSOEsselectedandreorganizedbylocalgovernments.Hence,earningsmanagementisnotsimplyamanagerialbehavior,butajointeﬀortoflocalgovernmentandﬁrmmanagers.Ascontrollingshareholders,localgovernmentshaveseveralincentivestohelplistedﬁrmsachievetheaccountingnumbernecessarytomeettheCSRC’sregulatoryrequirementsforrightsoﬀeringsanddelisting.First,theﬁs-caldecentralizationthattookplaceinthe1990scontinuestoprovidelocalgovernmentswithstrongincentivestoincreasetheirrevenuebasesbystrengtheningtheirlocalecon-omies(QianandWeingast,1994).SincelistedSOEsareusuallylargeandproﬁtableﬁrmsinlocaljurisdictions,theirsustainedproﬁtabilityandgrowtharecrucialtotheirlocaleconomies.

Second,listedﬁrmsareusuallycommittedtohelpimprovelocalwelfarebecausetheIPOquotasystemrestrictsthenumberofﬁrmsgoingpublicineachregionsoﬁrmsthatdesiretogopublicmustcompetewithotherﬁrmsforthefavorabletreatmentandcloutthatlocalgovernmentsareabletoprovide.Lawrence(1999)documentedthatlistedﬁrmsinChinawereaskedtotakeoverailingSOEs,tograntsubstantiallong-termloanstoparentSOEs,toinvestingovernmentinfrastructureprojectswithnewcapitalraisedfromthestockmarket,oreventopayoutdividendswhentheirlocalgovernmentshavecash-ﬂowproblems.Thesecommitmentscanbeinformalorenforcedthroughpoliticalreciprocity.

Third,thesuccessoflistedﬁrmsisgenerallyconsistentwiththepersonalinterestsoflocalpoliticians(Li,1998).GaininganIPOquotaisoftenviewedasasignofpoliticalcloutandsuperbperformanceforapolitician,whileadelistingwillbringgreatdisgracetoapolitician’sreputation.Asanagentofthecontrollingshareholder,agovernmentoﬃ-ThetradingpricevolatilityonSTﬁrmsislimitedwithin+5%daily.ThepriceincreaseinaPTsharecannotbemorethan5%foranytradingdaytopreventinsidermanipulation,andthepriceofaPTshareisallowedtodeclineunlimitedly.ThePTsharecanonlybetradedonFridays.9SeeJiang,1998;ChenandYuan,2004;andHawetal.,2005.

8X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274267

cialoftensitsonthecorporateboardofdirectorsandappointsafavoredsubordinatetobethechiefexecutiveofalistedﬁrm.Moreover,theoﬃcialcansecureanincomestreamfromthelistedﬁrmsthroughbribery(ShleiferandVishny,1993).

Localgovernmentshavealternativewaystosupportlistedﬁrms,suchasgrantingtax-ationpreferenceorfavoringlistedﬁrmsintheprojectapprovalprocess.However,govern-mentsubsidiesareamoredirectandconvenientwaytohelplistedﬁrmsinearningsmanagement.TheMinistryofFinanceofthecentralgovernmentregulatesthesubsidypolicy.Accordingtothepolicy,theprimarygoalofsubsidyistodevelopstrategicareassuchasagriculture,publicutilities,andhigh-techindustries,andtohelpﬁnanciallydis-tressedSOEs.Thelocalgovernmenthasthediscretiontodeterminethetimingandamountofsubsidy,andthereforecanuseitverypreciselyandpromptlyforthepurposeofearningsmanagement.

4.Researchdesignandempiricalanalysis

OursampledataincludesalllistedﬁrmsinChinaduringtheperiod1994–2000.10TheﬁnancialdataisavailablefromGeniusdatabase,developedbytheGeniusCompanyinChina.11Thetotalsampleincludes4437ﬁrm-yearobservationsafterdeletingobservationswithmissingorextremevalues.Chineseaccountingregulationsrequireﬁrmstolistgov-ernmentsubsidyasaseparateitemontheirincomestatements.Regionalmacroeconomicsdata,suchasunemploymentrates,ﬁscaldeﬁcitsandtheprivatemarketdevelopmentindex,arefromtheChinaEconomyStatisticalYearbook.

Totestthehypothesisofwhethergovernmentsubsidyisusedtohelplistedﬁrmsmeettheregulatorythreshold,weusealogisticmodeltoestimatetheprobabilityoflistedﬁrmsreceivinggovernmentsubsidies:

ProbðSUB¼1Þ¼

11þeÀz

ð1Þ

whereSUBisadummyvariable,withSUB=1forreceivingsubsidyandSUB=0other-wise.zisasetofexplanatoryvariables,includingkeyearningsmanagementincentivevari-ablesandothercontrolvariables.

Webackouttheﬁrm’spre-subsidyROE(PROE)12anduseittoidentifytheﬁrmswithearningsmanagementincentives.Thekeyindicatorvariablesforearningsmanagementincentivesinclude:INCIPIENT_RIGHTindicatesﬁrmswithPROEclosetoreachingtherightsoﬀeringROEthreshold;CLEAR_RIGHTindicatesﬁrmswithPROEclearlyqualiﬁedforrightsoﬀering;INCIPIENT_LOSSindicatesﬁrmswithPROEclosetoreach-

Ourdatatruncatedto1994–2000.Before1994,thenumberoflistedﬁrmsistoosmall(onlyabout40)andtherewasnosubsidyreceivedbytheselistedﬁrms.After2000,aseriesofnewregulationswereissued.Beforethereﬁnedregulationin2001,theCSRCacceptedthesubsidywhenevaluatingtheﬁrmsROE.The2001reﬁnedROEbenchmarkrequiresthatthetargetROEistheROEbasedonnetproﬁtortheROEexcludingnon-operatingproﬁts,whicheverislower.Subsidyisincorporatedifitisgrantedsuccessivelyforatleastthreeyears.Oursampleperiodfrom1994to2000isnotaﬀectedbythenewregulation.11TheGeniusdatabaseiswidelyusedbyacademicresearchersinChina.12Accordingtotaxlaw,thesubsidyisnottaxableinChina.Thereforethesubsidyislistedasanaftertaxitemintheincomestatement.Tobackoutthepre-subsidyROE,wesimplydeductthesubsidyfromnetincome.

10268X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274

ingpositiveproﬁtability;andCLEAR_LOSSindicatesﬁrmswithPROEclearlyinloss.ForthedeﬁnitionofclosenesstoROEbenchmark,weuse1%tobeconsistentwithﬁnd-ingsofpriorresearch.13BothﬁrmsinINCIPIENT_RIGHTandINCIPIENT_LOSShavestrongincentivestomanageearningsandthereforeareexpectedtohaveahigherproba-bilityofreceivingsubsidyfromgovernment.FirmsinCLEAR_RIGHThavetheleastincentivetomanageearningsupward,andtheyarelesslikelytoreceivesubsidyfromgov-ernment.FirmsinCLEAR_LOSSmightbeignoredbylocalgovernmentsasan‘‘unwor-thychild.”

Inthemodelwealsoincludeanumberofvariablestocontrolforpolicyfactors,solvencyneeds,sizeeﬀect,andregionaldiﬀerences.WeusethedummyvariablesINDUS-TRYandTRADEtocapturethepolicyeﬀect.INDUSTRYrepresentsﬁrmsinhigh-tech,publicutility,oragriculturalindustries,andTRADEindicatesﬁrmsinvolvinginterna-tionaltrade.ThevariableDEBT,measuredbytheleverageratio,indicatesthesolvencyneedoflistedﬁrms.ThevariableSIZEistocapturethesizeeﬀectthatlargerﬁrmshavegreaterinﬂuenceinlocaleconomiesandaremorelikelytoreceivesubsidyevenwithouttheneedofmeetingROErequirements.ThevariableNUMcontrolsforthetotalnumberoflistedﬁrmsinthesameregion,asaproxyforthebudgetconstraintofthelocalgovern-ment.ThevariableGDP(percapitaGDP)controlsforthewealtheﬀectoftheregion.Thewealthierregioncanaﬀordmoresubsidies.ThevariableFISDEF(percapitaﬁscaldeﬁcitintheregion)controlsforthebudgettightnessofthelocalgovernment.Thewealthierlocalgovernmentcanbemoregenerous.Ourdeﬁnitionofregionincludesalltheprovin-cialleveljurisdictions.

Asummaryofabovevariablesislistedbelow:ROEitreturnonequityforﬁrmiinyeart;

PROEitthepre-subsidyROE,deﬁnedasnetincomeminusthesubsidydeﬂatedbycom-monequity;

INCIPIENT_LOSSit1ifÀ1%Ã*INCIPIENT_RIGHTit1ifRÃtÀ1%minimumROErequirementforrightsoﬀering

MODEST_RIGHTit1ifR*minimumROErequirementforrightsoﬀering

*CLEAR_RIGHTit1ifPROEit>RÃtþ1%;0otherwise;whereRindicatestheminimum

ROErequirementforrightsoﬀering

DEBTitleverageratio,deﬁnedaslong-termdebttoasset

INDUSTRYi1forﬁrmsinhigh-techandprotectionindustry(publicutilityandagricul-tural);0otherwise

TRADEi1forﬁrmswithinternationaltrade;0otherwiseNUMittotalnumberoflistedﬁrmsinthesameregionSIZEitnaturallogofthetotalasset

GDPitnaturallogofpercapitaGDPinﬁrmi’slocatedregioninyeart

Fordetail,seeAharonyetal.,2000;ChenandYuan,2004;Hawetal.,2005,thesestudiesreportthatﬁrmswithROEmarginally(usually1%or1.5%)abovetheregulatorylevelsareidentiﬁedasbeingsuspectedofearningsmanagement.

13X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274269

4.1.Empiricalresults

Table1providesthedescriptivestatisticsofgovernmentsubsidiesfrom1994to2000.PanelAshowsthenumberandthepercentageofﬁrmsreceivingsubsidiesineachyear.In1994,onlyabout5%ofthecompaniesobtainedsubsidies.From1995to1997,theper-centageincreasedtoaround20%.Thepercentagejumpedtoalmost50%after1998.Gov-ernmentsubsidyhasbecomeanincreasinglyimportantcomponentinlistedﬁrm’snetincome.PanelBillustratesthedollaramountofsubsidyandtheratioofsubsidyovernetincome.Themeanandmediansubsidygrantedtolistedﬁrmswereabout560and215mil-lionRMBrespectively,andaccountedfor12.27%and5.46%,respectively,ofthenetincomeoftheﬁrms.PanelCliststhefrequencyofsubsidiesandshowsthat33.79%ofﬁrmsneverreceivesubsidiesand42.26%ofﬁrmsreceivesubsidiesonlyonceortwice.Thisdiscretionaryfeaturemakessubsidyaneﬀectivetoolforassistingearningsmanagement.

Table2presentstheunivariatetestresultbycomparingthecharactersofﬁrmswithsubsidyversusthosewithoutsubsidy.T-testisappliedtocontinuousvariablesandv2testtobinaryvariables.PanelAshowsthatﬁrmswithsubsidiesarelarger,moreleveraged,lessproﬁtable,andinwealthierregionscomparedtoﬁrmswithoutsubsidy.Hence,budgetconstraintandeconomicenvironmentaﬀectgovernmentsubsidyinpredictableways.PanelBshowsthatﬁrmsclosetomeetingtheROEthresholdforrightsoﬀeringsandﬁrmswithsmalllosseshaveahigherfrequencyofreceivingsubsidiesthanotherﬁrms;andﬁrmsclearlymeetingtheROEthresholdforrightsoﬀeringshavealowerfrequencyofreceivingsubsidies.

ThemultivariatelogisticregressionresultsarepresentedinTable3.InTable3,weﬁrstperformthetestforthewholesampleperiodof1994–2000.ThecoeﬃcientsonINCIPI-Table1

Descriptivestatisticsofgovernmentsubsidies

1994

1995

1996

1997

1998775377

48.65%

1999839461

54.95%

20001006469

46.62%

Total44371641

36.98%

PanelA:PercentageoflistedﬁrmsreceivinggovernmentsubsidiesTotalnumberofﬁrms271312527707Firmswithsubsidy1470109141

5.17%22.44%20.68%19.94%

MeanSDMinP2550.57

1.16

Median214.515.46

P75728.3817.53

Max5070.37129.00

PanelB:DescriptivestatisticsofgovernmentsubsidiesSubsidy(millionRMB)559.16806.570.00Subsidy/netincome(%)12.2717.670.00Frequency

Numberofﬁrms

Percentageofﬁrmsinthetotalsample(%)

PanelC:Frequencyofreceivingsubsidiesoverthewholesampleperiod036733.79124222.28221719.98315314.094544.975343.136131.20760.55

270X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274

Table2

Descriptivestatisticsofallvariablesforsubsidyandnon-subsidyﬁrms(1994–2000)Variable

FirmswithsubsidyMean

PanelA:ContinuousvariablesaDEBT/ASSET0.442PROE0.066TOTAL_ASSET11.510NUM4.987GDP9.184FISDEF2.375Variable

PanelB:BinaryvariablesbINCIPIENT_LOSSCLEAR_LOSS

INCIPIENT_RIGHTCLEAR_RIGHTINDUSTRYTRADE

aFirmswithoutsubsidyMean0.4140.08911.3153.9058.9432.388

SD0.1590.1260.7502.5820.5710.419

TstatisticscSD0.1680.1190.8453.0610.6490.344

2.76**2.89**3.96***12.56***6.45***1.11

FrequencyofreceivingsubsidiesWithvalueof1(%)86.6736.7971.1531.2238.1643.43

Withvalueof0(%)36.8239.9335.3044.2636.7236.31

v2c7.87***0.5454.66***39.98***0.294.18***PanelApresentsthet-testresultofthemeandiﬀerenceofcontinuousvariablesbetweensubsidyandnon-subsidygroup.bPanelBpresentsthev2teststatisticsofthediﬀerencesinthefrequenciesofreceivingsubsidyingroupsdeﬁnedbythevalueofbinaryvariables(0or1).c******,,indicatethesigniﬁcanceatalevelof0.1,0.05and0.01respectivelyusingonetailedtest.Bold-typehighlightssigniﬁcantﬁndings.

ENT_RIGHTandINCIPIENT_LOSSaresigniﬁcantlypositive,suggestingthatlistedﬁrmsclosetomeetingtherightsoﬀeringrequirementandﬁrmswithsmalllosshavehigherprobabilityofreceivingsubsidy.ThecoeﬃcientsofCLEAR_RIGHTandCLEAR_LOSSaresigniﬁcantlynegative,suggestingthatlistedﬁrmsclearlyqualiﬁedforrightsoﬀeringrequirementandﬁrmsthatclearlyfallsigniﬁcantlyshortoftheregulatoryearningsrequirementsarelesslikelytoreceivesubsidy.Thecontrolvariablesaregenerallyconsis-tentwithourpredictions.Policyfactorshavelittlepredictionpowertosubsidydecision.Hence,localgovernmentsseemmoreinterestedinplayingtheearningsmanagementgamethanfollowingthecentralgovernment’spolicyintentions.

Wealsoperformthelogisticregressiontestoverthethreediﬀerentregulationregimesrespectively.Basedonregulationchanges,wedivideoursampleintothreeperiods:from1994to1995,whenonlygentleregulationontherightsoﬀeringwasinplace;from1996to1997,whentherightsoﬀeringregulationbecametougherwithaminimum10%ROErequirementandnoregulationondelistingexisted;andfrom1998to2000,whenthenewregulationregardingdelistingwasputinplace.Theresultsshowthatinthenoreg-ulationperiodthereisnoevidenceofearningsmanagementthroughsubsidy;inthetoughrightsoﬀeringregulationperiod,onlyﬁrmswithincentivestomeetrightsoﬀeringregula-tionaremorelikelytoreceivegovernmentsubsidy;andinthelaterperiodwhenbothdelistingandrightsoﬀeringregulationswereenacted,ﬁrmswitheitherrightsoﬀeringordelisting-avoidanceincentivesaremorelikelytoreceivesubsidy.Thisdynamicdevelop-mentofsubsidydecisionsfurtherconﬁrmsourhypothesis.

X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274

Table3

LocalgovernmentsubsidiesandearningsmanagementaModel:ProbðSUB¼1Þ¼1þ1eÀzSUB:1forcompanyreceivingsubsidy,0otherwise;zconsistsofasetsofexplanationvariablesbVariableINTERCEPTINCIPIENT_LOSSINCIPIENT_RIGHTCLEAR_LOSSCLEAR_RIGHTINDUSTRYTRADEDEBTPROEASSETNUMGDPFISDEFObservations

a271

Sign1994–2000À5.473(0.000)1.690(0.032)1.394(0.000)À0.716(0.000)À0.395(0.000)À0.110(0.215)0.174(0.127)0.480(0.000)À1.557(0.000)0.160(0.000)0.007(0.001)0.418(0.000)À0.385(0.000)4437

1994–1995À35.269(0.000)

1996–1997À4.549(0.006)12.288(0.967)1.201(0.000)À0.463(0.147)À0.573(0.000)0.087(0.512)0.009(0.957)0.446(0.188)À0.799(0.216)0.053(0.433)0.104(0.001)0.388(0.169)À0.081(0.698)1233

1998–2000À4.309(0.009)2.157(0.045)1.721(0.000)À0.237(0.049)À0.113(0.004)0.217(0.083)0.692(0.000)1.554(0.000)À0.047(0.947)0.017(0.787)À0.026(0.317)0.462(0.060)À0.123(0.521)2615

++ÀÀ+++

1.448(0.415)À1.446(0.225)À0.858(0.321)À0.548(0.202)À0.202(0.627)2.438(0.011)À4.570(0.083)0.047(0.783)0.035(0.741)4.470(0.000)À1.220(0.180)581

Themodelwasestimatedonthewholesampleperiodfrom1994to2000andonthethreesubperiods.1994–1995indicatestheregimeofnoregulation.Inthis,thedeﬁnitionsofINCIPIENT_LOSSandINCIPI-ENT_RIGHTarethesame.1996–1997indicatesthetoughrightsoﬀeringregulationregime,and1998–2000indicatestherevisedregulationregime.bPvalueofv2statisticsforeachcoeﬃcientisreported.

4.2.Additionalevidence:Governmentownershipandsubsidy

Theincentivesoflocalgovernmentstocolludewithlistedﬁrmsdependontheclose-nessoftherelationshipbetweenthelistedﬁrmsandthelocalgovernments.Weexpectthatﬁrmscontrolledbythelocalgovernmenttohavetheclosestrelationshipwiththelocalgovernment.Townshipandvillageenterprises(TVEs)andjointventureshavemodestrelationshipswiththelocalgovernment.Thosecontrolledbythecentralgovern-menthavethemostdistantrelationships.Table4presentstheregressionresultsforﬁrmsbasedonownerships.WeﬁndthatwhenthelistedSOEsarecontrolledbylocalgovernments,SUBissigniﬁcantlyassociatedwiththeROErequirement.Ontheotherhand,whenthelistedSOEsarecontrolledbythecentralgovernment,SUBisnotsig-niﬁcantlyassociatedwiththeROErequirement,butissigniﬁcantlyassociatedwiththepolicyinitiativesetbythecentralgovernment.Forjointventures,localgovernments

272X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274

Table4

CorporateownershipandgovernmentsubsidydecisionaModel:ProbðSUB¼1Þ¼1þ1eÀzSUB:1forcompanyreceivingsubsidy,0otherwise;zconsistsofasetsofexplanationvariables

CorporateownershipLocalSOEsbObservations

SubsidyfrequencyRegressionresults

(1)

INTERCEPTINCIPIENT_LOSSINCIPIENT_RIGHTCLEAR_LOSSCLEAR_RIGHTINDUSTRYTRADEDEBTPROEASSETNUMGDPFISDEF

aCentralSOEsc20139.30%(2)À10.192(0.052)

JointVenturesd23036.68%(3)À16.586(0.024)

TVEse23531.91%(4)À14.904(0.000)

377137.20%

À4.881(0.000)1.619(0.039)1.406(0.000)À0.735(0.002)À0.412(0.000)À0.067(0.487)0.005(0.969)0.443(0.063)À1.583(0.000)0.133(0.005)0.006(0.004)0.415(0.000)0.438(0.000)

0.473(0.674)1.248(0.177)À1.348(0.004)1.197(0.031)1.609(0.002)3.730(0.003)1.566(0.558)0.235(0.228)0.004(0.688)À0.528(0.403)0.064(0.919)0.527(0.489)À0.483(0.555)0.897(0.035)0.047(0.904)1.813(0.069)À0.133(0.915)À2.840(0.206)0.238(0.243)À0.099(0.313)0.838(0.476)0.827(0.280)

1.020(0.007)0.280(0.700)0.637(0.143)1.105(0.014)1.229(0.042)0.337(0.757)

1.147(0.000)0.314(0.006)1.224(0.179)À1.170(0.149)

Themodelwasestimatedonthepooledsampleover1994–2000fordiﬀerentﬁrmtypes.Pvalueofv2statistictestisreportedforeachcoeﬃcient.bLocalSOEsarelistedSOEsthatareundercontroloflocalgovernment.cCentralSOEsarelistedSOEsthatareundercontrolofcentralgovernment.dJointventuresarerestructuredSOEsthathaveforeigninvestors.eTVEsaretownshipandvillageenterprisesinwhichlocalgovernmenthasownership.ForregressionofTVEs,variablesPROEareexcludedasthehighcorrelationbetweenthesubsidydecisioniscloselyrelatedtotheperformance.

oﬀersubsidiestoﬁrmsclearlyabovetherightsoﬀeringROEthreshold.Hence,thesub-sidiestojointventuresaremorelikelyservedasaninducementtoattractgood-perform-ingjointventureﬁrmsthanasaninstrumentforearningsmanagement.FinallywealsoﬁndmoderatelysigniﬁcantevidencethatlocalgovernmentsassisttheTVEstocrossovertherightsoﬀeringthreshold.

X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274273

5.Conclusion

ThestockmarketinChinawasdevelopedtofacilitateeconomicrestructuringaswellastopromoteeconomicdevelopment.Duringtheeconomictransition,localgov-ernmentsservemultipleroles:providersofpublicservice,agentsofChina’scentralgovernmenttomonitorlistedﬁrms,andmajorshareholdersoflistedﬁrms.Thiscom-plicatedrelationshipgivesrisetoauniquephenomenonintheChinesestockmarket:localgovernmentsintimatelydancewithlistedﬁrmstothetunesetbythecentralgovernment.

Thetightquotasystemleftalmostalllistedﬁrmsundercapitalizedandhungryfortheprivilegeofrightsoﬀering.TheoverwhelmingdemandofequityﬁnancingforcedtheCSRCtousetheROErequirementtoraisethethresholdforrightsoﬀering.Withrigidrule-basedaccounting,thelistedﬁrmshavelimitedinstrumentsforearningsmanagement.Asaresult,localgovernmentslentahelpinghandbyprovidingsubsidiestoboosttheROEsofcertainlistedﬁrms.Ourempiricalstudysuggeststhatwhenlocalgovernmentsheldmajorityownership,governmentsubsidycanbeusedtohelplistedﬁrmstomanageearningstocircumventtheCSRC’sROErequirements.

Intheprocessofeconomictransition,multiplelayersofgovernmentwereinvolvedintheoperationofSOEs.Weﬁndevidenceofmoralhazardinthelocalgovernment’sroleasanagentforthecentralgovernment.Inanemergingmarket,theregulatorplaysmulti-ple,sometimesconﬂictingrolesinthemarket,duetothelackofthesophisticatedinterme-diariesandeﬀectivecorporategovernancemechanisms.Therefore,ourstudyalsohasimplicationsforregulatorsinChinatodetectandpreventoratleastbeawareofthecol-lusiveanddynamicactionsoflocalgovernmentsandlistedﬁrmsastheyattempttocir-cumventthecentralgovernment’ssecuritiesregulations.Infact,thecentralgovernmentinChinahasbecomegraduallyawareoflocalgovernments’involvementinearningsman-agement,andin2001CSRCreﬁnedtheROEbenchmarktoexcludetheeﬀectoftransitorynon-operatingproﬁts.References

Aharony,J.,Lee,C.J.,Wong,T.J.,2000.FinancialpackagingofIPOﬁrmsinatraditionaleconomy,thecaseofB-sharesandH-sharesinChina.JournalofAccountingResearch33(1),103–125.

Bai,C.E.,Wang,Y.J.,1998.Bureaucraticcontrolandthesoftbudgetconstraint.JournalofComparativeEconomics26(1),41–61.

Bushman,R.M.,Smith,A.J.,2001.Financialaccountinginformationandcorporategovernance.JournalofAccountingandEconomics32(1–3),237–333.

Chen,K.C.W.,Yuan,H.,2004.Earningsmanagementandcapitalresourceallocation,evidencefromChina’saccounting-basedregulationofrightissue.TheAccountingReview79(3),645–665.

Dechow,P.M.,Skinner,D.J.,2000.Earningsmanagement,reconcilingtheviewsofaccountingacademics,practitioners,andregulators.AccountingHorizons14(2),235–250.

Haw,I.M.,Qi,D.,Wu,W.,Zhang,W.,2005.MarketconsequencesofearningsmanagementinresponsetosecurityregulationsinChina.ContemporaryAccountingResearch22(1),95–140.

Healy,P.M.,Wahlen,M.J.,1999.Areviewoftheearningsmanagementliteratureanditsimplicationsforstandardsetting.AccountingHorizons13(4),365–384.

Jiang,Y.,1998.Anempiricalstudyonearningsmanagementinlistedcompanies—AnanalysisofthresholdonEPSandROE.WorkingPaper,ShanghaiUniversityofEconomicsandFinance.

Lawrence,S.,1999.Toomanymothers-in-law.FarEasternEconomicReview(2/19/1999),12–13.

Lee,C.J.,1987.Accountinginfrastructureandeconomicdevelopment.JournalofAccountingandPublicPolicy6(2),75–86.

274X.Chenetal./JournalofAccountingandPublicPolicy27(2008)262–274

Lee,C.J.,2001.Financialrestructuringofstate-ownedenterprisesinChina,thecaseofShanghaiSunvePharmaceuticalCorporation.Accounting,OrganizationandSociety26,673–689.

Li,D.,1998.ChangingincentivesoftheChinesebureaucracy.AmericanEconomicReview88(2),393–397.Qian,Y.,Weingast,B.R.,1994.China’stransitiontomarkets,market-preservingfederalism,Chinesestyle.JournalofPolicyReform1,149–188.

Schipper,K.,1989.Commentaryonearningsmanagement.AccountingHorizons3(4),92–106.Shleifer,A.,Vishny,R.,1993.Corruption.QuarterlyJournalofEconomics108(3),599–617.

Tiebout,C.M.,1956.Apuretheoryoflocalexpenditures.JournalofPoliticalEconomy64,416–424.

因篇幅问题不能全部显示，请点此查看更多更全内容

查看全文

全部栏目

Government assisted earnings management in China