Governance Structure and the Weighting of Performa

Electroniccopyavailableat:=8753241WorkingPaperWPNo601July,2005GOVERNANCESTRUCTUREANDTHEWEIGHTINGOFPERFORMANCEMEASURESINCEOCOMPENSATIONAntonioDávila*FernandoPeñalva***ProfessorofAccountingandControl,IESEandGraduateSchoolofBusiness,StanfordUniversity**ProfessorofAccountingandControl,IESEIESEBusinessSchool–UniversityofNavarraAvda.Pearson,21–08034Barcelona,Spain.Tel.:[phone]Fax:[phone]CaminodelCerrodelÁguila,3(Ctra.deCastilla,km.5,180)–28023Madrid,Spain.Tel.:[phone]Fax:[phone]Copyright©2005IESEBusinessSchool.Electroniccopyavailableat:=875324GOVERNANCESTRUCTUREANDTHEWEIGHTINGOFPERFORMANCEMEASURESINCEOCOMPENSATIONAbstractWeempiricallyexaminehowgovernancestructureaffectsthedesignofexecutivecompensationcontractsand,inparticular,theimplicitweightsoffirmperformancemeasuresinCEOcompensation.WefindthatcompensationcontractsinfirmswithhighertakeoverprotectionandwheretheCEOhasmoreinfluenceongovernancedecisionsputmoreweightonaccounting-basedmeasuresofperformance(returnonassets)thanonstock-basedperformancemeasures(marketreturns).Inadditionaltests,wefurtherfindthatCEOcompensationinthesefirmshaslowervarianceandahigherproportionofcash(versusstock-based)compensation.WefurtherfindthatCEOs’incentives(measuredaschangesinCEOannualwealth,whichincludeschangesinthevalueofCEOs’equityholdings,inadditiontoyearlycompensation)donotvaryacrossgovernancestructures.ThesefindingsareconsistentwithCEOsinfirmswithhightakeoverprotectionandwheretheyhavemoreinfluenceongovernancebeingabletoinfluencethecontracting.Keywords:CEOcompensation,executivecompensation,performancemeasures,compensationcontracts.GOVERNANCESTRUCTUREANDTHEWEIGHTINGOFPERFORMANCEMEASURESINCEOCOMPENSATION*1.IntroductionGovernancestructureshaveastronginfluenceonCEOcompensation.Theydeterminethecompany’sexposuretothemarketforcorporatecontrolthroughtheirpolicydecisions(Jensen,1993;HermalinandWeisbach,2003)andthereforehowcontractsinfluenceCEObehavior.InternalgovernancebodiesarealsodirectlyresponsibleforthedesignofCEOcompensationcontractsandthiscontractingprocessisoneofcompanydirectors’maintasks.Coreetal.(1999),usingsurvey-basedcompensationdata,provideinitialevidenceontherelationshipbetweengovernanceandCEOcompensation.TheirfindingsindicatethatgovernancestructuressubjecttomoreinfluencefromtheCEOarecorrelatedwithhigherlevelsofCEOcompensation.Theyalsofindthatthesegovernancestructuresarecorrelatedwithworsestockreturnsandworseoperatingperformance.Theauthorsarguethatthisevidenceisconsistentwiththepresenceofagencycostsassociatedwithweakergovernance,wheretheCEOexertshisbargainingpowertoextractrentsattheexpenseofshareholders.WhileexistingevidenceindicatesthatcertaingovernancestructuresallowCEOstoextractexcessrents,itremainsunknownhowtheserentsmightbeextracted.TheobjectiveofthisstudyistoexaminewhetherthedesignofCEOcompensationcontractsvariesacrossgovernancestructures.Wefocusononeaspectofcontractdesign:theimplicitweightonvariousperformancemeasures.CEOsmaywanttoincreasetheweightthatthecontractputsonmorecontrollablemeasures,thusreducingvariabilityinactualcompensation,raisingthelikelihoodofachievingtargetobjectives,andincreasingtherentsthattheCEOisabletocapture.FromtheCEO’spointofview,controllablemeasureshaveabettersignal-to-noiseratio,whichreducestheuncertaintyaboutoutcomes,andamoretransparentassociationbetweencertaintypesofeffortandmeasuredoutcome,thusmakingiteasierfortheagenttoexertthetypeandlevelofeffortrequiredtoachievetheobjectives.Fromtheshareholders’pointofview,thesemeasuresareincomplete—leavingoutcertaintypesofeffortrelevanttothevalueofthefirm—andincongruent—weightingthevarioustypesofeffortdifferentlyfromwhatshareholderswouldotherwisechoose.Thus,thesecontractsfavorcontrollablemeasuresattheexpenseofnoisiermeasuresthatenhancethecompletenessandcongruencyofthecontract._____________*WeappreciatethecommentsfromRamjiBalakrishnan,DanCollins,JoanEnricRicart,andGeorgeFoster,andfromparticipantsatIESEBusinessSchoolandIowaUniversityseminars,andatEuropeanAccountingAssociationandAmericanAccountingAssociationmeetings.2Wefocusontwomainclassesofperformancemeasures:accounting-basedmeasuresofperformance(returnonassets)andstock-basedmeasures(stockreturns);andweexaminewhetheraccounting-basedmeasures—arguablymorecontrollableonaverage—havehigherimplicitweightsand/or,conversely,whetherstock-basedmeasures—lesscontrollable—receivelowerweightsasgovernancefavorsCEOinfluence.PreviousaccountingliteraturehasstudiedtheimplicitweightsofperformancemeasuresonthedesignofCEOcompensationcontracts(Lambert,2001),suchastherelevanceofaccountingnumbers(LambertandLarcker,1987;Sloan,1993)andnon-financialperformancemeasures(Bushmanetal.,1996;Ittneretal.,1997;DávilaandVenkatachalam,2004)beyondstockreturns.TheevidenceindicatesthattheimplicitweightsvarywiththeabilityofthevariousmeasurestocongruentlyreflectaCEO’smulti-dimensionaleffortwhileminimizingtheriskimposeduponhim.Accordingly,theweightonaparticularmeasureincreasesasitsrelativenoisedecreases(BankerandDatar,1989)and