Insider trading restrictions and top executive compensation

内容需要下载文档才能查看 内容需要下载文档才能查看

JournalofAccountingandEconomics56(2013)91–112

ContentslistsavailableatSciVerseScienceDirect

JournalofAccountingandEconomics

journalhomepage:http://wendang.chazidian.com/locate/jae

Insidertradingrestrictionsandtopexecutivecompensation$

DavidJ.Denisa,n,JinXub

ab

KatzGraduateSchoolofBusiness,UniversityofPittsburgh,Pittsburgh,PA15260,USA

KrannertSchoolofManagement,PurdueUniversity,403WestStateStreet,WestLafayette,IN47907-2056,USA

articleinfo

Articlehistory:

Received7July2011

Receivedinrevisedform8April2013

Accepted19April2013

Availableonline29April2013JELclassification:G18G32G34

Keywords:

InsidertradingrestrictionsExecutivecompensationInsiderownership

abstract

Theuseofequityincentivesissignificantlygreaterincountrieswithstrongerinsidertradingrestrictions,andthesehigherincentivesareassociatedwithhighertotalpay.Thesefindingsarerobusttoalternativedefinitionsofinsidertradingrestrictionsandenforcement,andtopanelregressionswithcountryfixedeffects.Wealsofindsignificantincreasesintopexecutivepayandtheuseofequity-basedincentivesintheperiodimmediatelyfollowingtheinitialenforcementofinsidertradinglaws.Weconcludethatinsidertradinglawsareonechannelthroughwhichcross-countrydifferencesinpaypracticescanbeexplained.

&2013ElsevierB.V.Allrightsreserved.

1.Introduction

Recentstudiesdocumentsubstantialcross-countryvariationinboththelevelofexecutivecompensationandtheuseofequity-basedincentivesfortopexecutives(see,e.g.,Murphy,1999,forthcoming).However,theunderlyingfactorscontributingtotheseobserveddifferencesremainthetopicofactiveinvestigation.Forexample,althoughConyonetal.(2011)findthathigherlevelsofpayforU.S.CEOsrelativetotheirU.K.andE.U.counterpartscanbeexplained(atleastinpart)bytheirhigherstockandoptionincentives,theirfindingsleaveopenthequestionofwhyincentivesaresomuchhigherforU.S.CEOs.Indeed,Conyonetal.suggestthat“researchersshouldshifttheireffortstowardbetterunderstandingthereasonfordifferencesinincentivesbetweenCEOsintheU.S.andCEOsintheU.K.andotherpartsoftheworld”.

Weanalyzewhethercountry-levelrestrictionsinsidertradingcontributetocross-countrydifferencesintopexecutivecompensation.Specifically,weanalyzeseveral(notmutuallyexclusive)channelsthroughwhichcompensationandinsidertradingrestrictionsmightberelated.Onepossibilityisthatinsidertradingrepresentsaformofcompensationfortop

Wethankananonymousreferee,LeonceBargeron,JohnCore(theeditor),MelanieCao,MartijnCremers,DianeDenis,MaraFaccio,SteveKaplan,SeoyoungKim,KenLehn,DavidLesmond,TimLoughran,HenryManne,JohnMcConnell,DarrenRoulstone,ShawnThomas,andMarkWalkerforhelpfuldiscussionsandcomments.Inaddition,thepaperhasbenefitedfromcommentsfromworkshopparticipantsatCarnegieMellonUniversity,HongKongUniversityofScienceandTechnology,MichiganStateUniversity,NanyangTechnologicalUniversity,NationalUniversityofSingapore,PennStateUniversity,SingaporeManagementUniversity,TempleUniversity,TulaneUniversity,UniversityofToronto,the2009NorthernFinanceAssociationmeetings,the2009StateofIndianaFinanceConferenceandthe2010AmericanFinanceAssociationmeetingsfortheirhelpfulcomments.XuacknowledgesthefinancialsupportfromthePurdueUniversityCenterforInternationalBusinessEducationandResearch.SpecialthanksareduetoLuisMarquesforpartofthedatausedinthepaper.ClarkeBjarnasonprovidedableresearch

assistance.n

Correspondingauthor.Tel.:+14126481708.

E-mailaddresses:djdenis@katz.pitt.edu(D.J.Denis),xu68@purdue.edu(J.Xu).0165-4101/$-seefrontmatter&2013ElsevierB.V.Allrightsreserved.http://wendang.chazidian.com/10.1016/j.jacceco.2013.04.003

☆

92D.J.Denis,J.Xu/JournalofAccountingandEconomics56(2013)91–112

executives.Thus,wheninsidertradinglawsarerestrictive,equilibriuminthemarketforcesfirmstoincreasetheleveloftopexecutivecompensation.Asecondpossibilityisthatthestrengthofinsiderlawsaffectstheoptimaluseequityincentives.insidertradingrepresentsaformofequityincentives(e.g.,Manne,1966),restrictingsuchtradingmightleadfirmstosubstituteothertypesofequityincentivesintheexecutivecompensationcontract.Similarly,wheninsidertradinglawsareweaker,firmsmightchoosetouserelativelyfewerincentivesinpaypackagessoastoavoidtrading-relatedagencyconflicts(see,forexample,BaimanandVerrecchia,1995).Becausegreaterequityincentivesexposetopexecutivestogreaterrisk,theincreaseduseofequityincentiveswheninsidertradingisrestrictedalsoleadstohigherlevelsofcompensation.Athirdpossibilityisthatinsidertradinglawsarethemselvesaresponsetocross-countrydifferencesinpaypractices.Specifically,incountriesinwhichfirmsusegreaterequityincentives,strongerinsidertradinglawsarerequiredtomitigatetrading-relatedagencyconflictswithexecutives.

Toprovideevidenceonthesehypotheses,analyzebothlevelsoftopexecutivecompensationandtheuseofequity-basedincentivesforabroadsetofexecutivesin41differentcountries.Ourprimarysampleconsistsof468non-U.S.firmswithAmericanDepositoryReceipts(ADRs)and1852U.S.firmsin2006.TheprimaryvirtueofanalyzingcompensationinforeignfirmswithADRsisthatsuchfirmsarerequiredtofileForm20-FwiththeSEC.Thus,weareabletoobtaincomplete,standardizedcompensationdataatthefirmlevelforallofoursamplefirms.Bycontrast,mostpriorcross-countrycompensationstudieshavebeenforcedtorelyuponsurvey-basedandcountry-aggregatecompensationdata.1WerecognizethatapossiblelimitationofourdataisthatfirmswithADRsarenotrepresentativeofthepopulationoffirmsinthatcountryandlateraddressthispotentiallimitation.

Wemeasureinsidertradingrestrictionsinways.First,followingDuandWei(2004),weuseaninsidertrading(ITR)indexthatisbasedonglobalexecutiveopinionsurveysabouttheextentofinsidertradingrestrictionsinindividualcountries.Second,weuseaninsidertradinglaw(ITL)indexfromBeny(2006)thatcapturesdifferencesinthestrengthofinsidertradinglaws.Importantly,forourpurposes,bothITRandITLexhibitsubstantialcross-countryvariation.

Ourbaselineanalysisindicatesthatequityincentivesarepositivelyrelatedtoinsidertradingrestrictions.Thesefindingsarerobusttotheinclusionofavarietyoffirm-levelandcountry-levelcontrolvariables,suchasfirmsize,leverage,R&D,growthopportunities,boardstructure,shareholderprotection,andcountryGDP.Moreover,theimpliedimpactofinsidertradingrestrictionsonequityincentivesisalsoeconomicallyimportant.Aone-unitincreaseintheITRindex(approximatelyonestandarddeviation)isassociatedwithanincreaseinoverallequityincentivesofover200%,andanincreaseinthepercentageofequity-basedpay(i.e.,theincrementalflowofincentives)ofabout22percentagepoints.Wealsofindthattheleveloftopexecutivetotalpayispositivelyassociatedwithinsidertradingrestrictions.However,wecannotrejectthatthisfindingisdrivenbythegreateruseofequityincentives(and,therefore,higherriskpremium)incountrieswithstrongerinsidertradingrestrictions.

Thesebaselinefindingsareconsistentwithallthreehypothesizedchannelsfortheassociationbetweeninsidertradingrestrictionsandcompensation.Inaddition,afourthpossibilityisthatthereisnocausalconnectionbetweeninsidertradingrestrictionsandtopexecutivepay/incentives.Underthisexplanation,theassociationbetweenthetwoisaspuriousbyproductofthefactthatourregressionsomitpotentiallyimportantfactorsthatarecorrelatedwithbothinsidertradingrestrictionsandexecutivepay.

Tofurtherdiscriminateamongthesealternativeexplanations,therefore,weconductseveraladditionaltests.First,weexploittime-seriesvariationininsidertradingrestrictionstoestimatepanelregressionswithcountryandyearfixedeffects.Theresultsfromthesetestsindicatethatgreaterrestrictionsoninsidertradingareassociatedwithsignificantincreasesintheuseofincentivecompensation.

Second,weanalyzechangesincompensationaroundthedatesofinitialenforcementofinsidertradinglaws.Aftercontrollingfortimetrendsandfixedcountryeffects,wefindthatboththeleveloftotalexecutivecompensationandtheuseofequity-basedincentivesincreasesignificantlyfollowingtheinitialenforcementofinsidertradinglaws.

Third,weanalyzewhethertheobservedlinkbetweenexecutivecompensationandinsidertradingrestrictionsisassociatedwiththelevelofinsiderownership.Becausehigherinsiderownershipdiminishestheneedforadditionalincentivesatthemargin,thisshouldweakenthelinkbetweeninsidertradingrestrictionsandtheuseofequityincentivesifthedirectionofcausationrunsfromtradingrestrictionstoexecutivepay.Consistentwiththisargument,theresultsfromthesetestsindicatethathigherinsideownershipweakensthelinkbetweeninsidertradingrestrictionsandbothoverallequityincentivesandthelevelofpay.

Basedontheresultsoftheseadditionaltests,weconcludethattheevidenceappearsmostconsistentwithacausallinkthatrunsfrominsidertradingrestrictionstocompensationincentives.Suchalinkisconsistentwithboth(i)insidertradingservingasanimplicitformofcompensation,and(ii)firmsoptimallychoosingtousegreater(fewer)equityincentiveswheninsidertradingrestrictionsarestrong(weak).Because(i)dependsoninsidertradingrestrictionsleadingtoreducedtradingprofitsforexecutiveswhereas(ii)doesnot,distinguishingbetweenthesetwoexplanationsrequiresobservationofwhetherstrongerinsidertradingrestrictionsactuallydoreduceexecutivetradingprofits.

Unfortunately,weareunawareofsystematicdataontheprofitabilityofinsidertradingoutsideoftheU.S.Therefore,weconductanindirecttestbyanalyzingtherun-upinstockpricespriortoacquisitionannouncementsineachofoursample

ExamplesincludeAbowdandBognanno(1995),AbowdandKaplan(1999),andMurphy(1999)whorelyonTowersPerrin'sWorldwideTotalRemunerationreports.

1

D.J.Denis,J.Xu/JournalofAccountingandEconomics56(2013)91–11293

countries.Priorstudiesdocumentboththeexistenceofabnormalinsidertradingbyexecutivespriortotakeoverannouncements(AgrawalandNasser,2012)andanassociationbetweenpre-announcementrun-upsandinformedtrading(ArshadiandEyssell,1991).Wethustestwhethercross-countrydifferencesinstockpricerun-upsareassociatedwiththestrengthofinsidertradinglaws.Consistentwithreducedinformedtradingincountrieswithtighterrestrictionsoninsidertrading,wefindasignificantnegativeassociationbetweenpre-acquisitionstockpricerun-upsandthestrengthofinsidertradingrestrictions.However,wenoteanimportantcaveatwithinterpretingthisfindingasevidenceofreducedinsiderprofitsforexecutivesincountrieswithstrongerinsidertradingrestrictions.AlthoughthestudybyArshadiandEyssell(1991)findsthatthereisareductioninnetinsiderpurchasespriortotenderofferannouncementsafterthepassageofanewinsidertradinglawintheU.S.,theyreportthatthelargesteffectisontradingbybeneficialownersofthetarget'sshares.Thus,ourfindingoflowerrun-upsincountrieswithstrongerinsidertradingrestrictionscanbeviewedasevidenceofreducedtradingprofitsforexecutivesonlyifthetradesofthoseexecutivesarecorrelatedwiththoseofotherbeneficialowners.Unfortunately,welackdirectevidenceonthatissue.

Byidentifyinginsidertradingrestrictionsasanimportantchannelthroughwhichdifferencesincross-countrycompensationandincentivescanbeexplained,ourfindingscontributetothegrowinginternationalexecutivecompensationliterature.2OurfindingsalsocomplementandextendthoseofRoulstone(2003),whofindsthatself-imposedinsidertradingrestrictionsinU.S.firmsarerelatedtohigherexecutivecompensationandagreaterlevelofincentivecompensation.

Theremainderofthepaperisorganizedasfollows.InSection2,weprovidebackgroundontheliteraturethathypothesizesalinkbetweeninsidertradingandcompensation.Section3describesoursampleselectionprocessanddescribesourprimarydata.Section4reportstheresultsofourcross-sectionalregressions.InSection5,wereporttheresultsfromaseriesofadditionalteststhatexplorealternativeexplanationsforourfindings.Section6reportstheresultsfromouranalysisofstockpricerun-upspriortoacquisitionannouncementsandSection7concludes.

2.Insidertrading,equityincentivesandtopexecutivecompensation

Inthissection,wediscussseveralchannelsthroughwhichinsidertradingrestrictionsmightbeassociatedwithcompensationlevelsandequityincentives.Thesechannelsdifferintheextenttowhichtheyrelyonasubstitutionbetweeninsidertradingprofitsandcompensation,andthedirectionofcausationinthehypothesizedassociationbetweeninsidertradingrestrictionsandcompensation.

http://wendang.chazidian.compensationandinsidertradingprofitsassubstitutes

Alargebodyofacademicliteraturereportsthatinsidertradingallowsinsiderstoprofitablyexploittheirprivateinformationandrealizesignificanttradingprofits.3Severalstudiespointtosuchtradingprofitsasprovidingapotentiallinkbetweeninsidertradingrestrictionsandcompensationpolicies.InBaimanandVerrecchia(1995,1996),managerscantradeprofitablybasedontheirprivateinformation,butshareholderscomputetheexpectedamountofthisredistributionfromtradersanddeductitfromexplicitmanagerialpay.Ifequilibriumwagesaredeterminedcompetitivelyandthelevelofcompensationismeasuredasthesumofexplicitcompensationandinsidertradingprofits,thesemodelspredictapositiveassociationbetweeninsidertradingrestrictionsandequilibriumcompensationlevels.4

Otherstudiespointtowardsinsidertradingprofitsasasubstitutesourceofequityincentives.Forexample,bothCarltonandFischel(1983)andManne(1966)hypothesizethatbecauseinsidertradingallowsinsiderstoprofitfromtheirinnovationandeffort,itrepresentsameansforprovidingincentivestotopexecutives.Moreover,CarltonandFischel(1983)arguethatallowinginsidertradingismoreefficientthanexpostsalaryrenegotiationsofthetypeinFama(1980)becauseitavoidsfrequentandcostlyrenegotiations.Thus,ifinsidertradingisrestricted,firmsneedtomakegreateruseofotherformsofincentivecompensationinordertomaintainoptimalincentivelevels.Thegreateruseofincentivesinturnleadstohigherlevelsofcompensationasexecutivesarecompensatedforbearinggreaterrisk.

Althoughtheaboveargumentssuggestthatrestrictionsoninsidertradingwillleadfirmstosubstituteintootherformsofequityincentives,criticsarguethatinsidertradingcancreateperverseincentivesaswell.Forexample,severalstudiespointoutthatbecauseinsidertradingallowsexecutivestoprofitfrombadnewsaswellasgoodnews,managersmaybelesswillingtoexertefforttoincreasefirmvalue,andmayeventakeactionsthatcreateunfavorablenews.5Thisargumentsuggeststhatwheninsidertradingisallowed,firmsshouldmakegreateruseofincentivecompensationinordertocounteracttheperverseincentiveeffectsofinsidertrading.

SeealsoAbowdandBognanno(1995),ConyonandSchwalbach(1997),andFernandesetal.(2013).

See,forexample,DamodaranandLiu(1993),Fidrmucetal.(2006),Meulbroek(1992),andSeyhun(1986).4

AsimilarargumentisalsomadebyKhannaetal.(1994).5

See,forexample,BagnoliandKhanna(1992),Levmore(1982),andSchotland(1967).CarltonandFischel(1983)contendthattheseadverseincentiveeffectsofinsidertradingareofsecond-orderimportancebecauseoflimitsonshort-sellingaswellasreputationandlitigationconcerns.

32

94D.J.Denis,J.Xu/JournalofAccountingandEconomics56(2013)91–112

2.2.Insidertradingrestrictionsandtheoptimaluseofincentives

Thepredictionsoutlinedaboverelyonasubstitutionbetweenexecutivetradingprofitsandtopexecutivecompensation.Similarpredictionscanbeobtained,however,evenifthereisnosuchsubstitution.Forexample,BaimanandVerrecchia(1995)modeltwotypesofagencyproblemsrelatedtoinformedtradingbymanagersintheirownshares.Inthefirst,theabilityofmanagerstotradeintheirownsharesweakensthelinkbetweencompensationpaidbyshareholdersandtheconsumptionofmanagers.Thisreducesshareholdercontroloverthemanager'seffortdecisiontherebydiminishingtheusefulnessofequityincentives.Inthesecond,managerstradingintheirownsharescanaffectthemarketpriceofthosesharesand,consequently,theirlevelofcompensation.Shareholderswilloptimallyaccountforthispotentialformanipulationbyloweringtheweightonstockpriceinthemanager'scompensationcontract.6

Thus,evenifthereisnosubstitutionbetweeninsidertradingprofitsandexplicitcompensation,theseargumentspredictapositiveassociationbetweencountry-levelinsidertradingrestrictionsandequityincentives.Again,executivesoffirmsusinggreaterequityincentiveswillrequirehigherlevelsoftotalpayascompensationforbearinggreaterincentiverisk.Theseargumentsthereforealsopredictapositiveassociationbetweencountry-levelinsidertradingrestrictionsandthelevelofcompensation.

2.3.Insidertradinglawsasanendogenousresponsetocountrypaypractices

Althoughtheabovehypothesespredictacausalassociationthatrunsfrominsidertradingrestrictionstoequilibriumcompensationincentives,itispossiblethatthedirectionofcausationisreversed.Thatis,country-levelinsidertradinglawsthemselvesmightbeanendogenousresponsetotypicalpaypracticeswithinthatcountry.Forexample,incountriesthatmakegreateruseofequityincentivesincompensationplans,regulatorsmightsettighterinsidertradingrestrictionstomitigatetrading-relatedagencyconflictsofthetypedescribedinBaimanandVerrecchia(1995,1996).Again,compensationlevelswillnecessarilybehigherinthesecountriestocompensateexecutivesforbearinggreaterincentiverisk.2.4.Relatedevidence

Therearethusseveralplausiblereasonstoexpectapositiveassociationbetweencountry-levelinsidertradingrestrictionsandbothcompensationlevelsandtheuseofincentives.Theexistenceofsuchanassociationanditsunderlyingreasonsareempiricalissuesthatourstudyaddresses.Todate,theliteratureprovidesonlylimited,indirectevidence.Roulstone(2003)studiesthelinkbetweencompensationandself-imposedinsidertradingrestrictions.Hefindsthatfirmswithsuchrestrictionsexhibithigherlevelsofcompensationandmakegreateruseofincentivecompensation.AlthoughRoulstoneinterpretsthesefindingsasconsistentwiththeviewthatinsidertradingplaysaroleinrewardingandmotivatingemployees,itisdifficulttoruleoutthepossibilityofacorrelatedomittedvariable.Moreover,becauseRoulstone(2003)analyzesinsidertradingrestrictionsandcompensationinjustonecountry,theU.S.,hisstudydoesnotaddresswhethercross-countrydifferencesincompensationpracticesareassociatedwithcountry-levelrestrictionsoninsidertrading.Ourstudyaimstofillthisgapbyprovidingdirectevidenceontheassociationbetweeninsidertradingrestrictionsandexecutivecompensationinacross-countrysetting.73.Sampleselectionanddatadescription

Inthissection,wedescribetheexecutivecompensationdatathatweuseinourempiricalanalysisaswellasourprimarymeasuresofinsidertradingrestrictions.Wethenreportsummarystatisticsforthesample.3.1.Executivecompensationdata

ForeigncompaniesissuingADRsintheU.S.marketsarerequiredtofileForm20-FreportswiththeSecuritiesandExchangeCommission(SEC).Thisformcontainsinformationontheboardofdirectors,thecompensationofexecutivesanddirectors,thelocationofthebusiness,thecompany'sindustrysector,andothermiscellaneousitems.8Fromthe20-Freports,http://wendang.chazidian.compensationdataforexecutivesofU.S.firmsisobtainedfromExecuComp.

Notethatalternatively,shareholderscouldimposetheirownrestrictionsonexecutivetradingthroughmorestringentvestingrequirements,aswellasrestrictionsonoptionexercisesandstocksales.Thus,thishypothesisimplicitlyassumesthatthecostsofimposingandenforcingsuchfirm-specificrestrictionsexceedthebenefitsofdoingso.

7

BaimanandVerrecchia(1996)alsosuggestthattheanalysisofinternationaldatacouldbeusefulforaddressingwhetherhigherlevelsofexecutivecompensationintheU.S.isrelatedtodifferencesininsidertrading:“ThegreaterdiffusenessofU.S.capitalmarketsandtheconsequentlessprofitableopportunitiesforinsidertradingbymanagersmayprovideapartialexplanationfortheobservedhigherlevelofdirectcompensationreceivedbyU.S.CEOs.”(p.2–3).

8

SeeBryanetal.(2006)foradetaileddescriptionofthissourceofcompensationdata.

6

D.J.Denis,J.Xu/JournalofAccountingandEconomics56(2013)91–11295

WebeginwithalistofallLevel2andLevel3ADRsasofMay2008fromtheJPMorganADRGroupwebsite.WethensupplementthissetoffirmsbyexaminingthelistofallADRsfrom1961to2007downloadedfromCRSP.WeexcludeLevel1andRule144AADRsbecausetheyareeithertradedoverthecounterorareprivateplacements,makingthemexemptfromSECreportingrequirements.FortheresultingsetofADRs,wesearchSEC'sEDGARdatabasefor20-Ffilingsin2006.Ifasamplefirmdoesnothavea20-Ffilingin2006,weobtaintherelevantdatafromeitherits2005or2007filing,whereavailable.

Thereissomevariationinthelevelofdetailwithwhichexecutivecompensationdataisavailable.Insomefilings,informationoncompensationofindividualexecutives(typicallythemosthighly-paidexecutives)canbeobtainedfromsummaryexecutivecompensationtables.WelaterrefertothesefirmsasU.S.stylereportingfirms.Inothercases,thecompensationtablesreportonlytheaggregatecompensationforallexecutives(anddirectors).Foreachfirm,weconstructfirm-levelcompensationvariablesthatcapturethelevelandstructureofcompensationfortheaverageexecutiveinthatfirm.Wheneveravailable,weusetheindividualexecutive-leveldatatoconstructourfirm-levelmeasures.Otherwise,weusetheaggregatefirm-leveldatatoconstructaveragevaluespertopexecutive.

Thecompensationdatacontainvariouscomponents:salary,bonus,equity-basedcompensationsuchasrestrictedstockawardsandoptiongrants,andothercompensationsuchaspensionsandperquisites.Ouranalysisfocusesprimarilyonthedollaramountoftotalcompensation(TotalPay)andameasureofoverallequityincentives(EquityIncentives).TotalPayisdefinedasthesumofsalary,bonus,equityincentivecompensation(includingrestrictedstockawardsandoptiongrants)andothercompensation.9Whenstatedinlocalcurrencies,TotalPayisconvertedintoU.S.dollarsusingyear-averageexchangeratesincorrespondingdatayearsastheconversionrate.Inaddition,allTotalPayfiguresareconvertedinto2006realdollarsbyadjustingfortheinflationratebetween2006andthegivenyear.EquityIncentivesaremeasuredastheoverallsensitivityofexecutivewealthtoa1%changeinthefirm'sstockprice.Becausethismeasurerequiresdataattheindividualexecutivelevel,testsusingthismeasureareconductedonthesubsetoffirmswithU.S.stylereporting.Nonetheless,becausewedonothavecompletedataonthedetailsofoutstandingoptions(e.g.,maturity,exerciseprice,etc.),thesesensitivitiesareestimatedbasedonfullydilutedsharesoutstanding;i.e.,assumingthatalloutstandingoptionsareexercised.Finally,insometests,weuseameasureoftheincrementalflowofincentives(EquityPayRatio)—definedasthefractionoftotalcompensationthatiscomprisedofequity-basedincentivepay,whereequity-basedincentivepayismeasuredasthetotalgrant-datevalueofrestrictedstockawardsandoptiongrants.ThedefinitionsfortheseandothervariablescanbefoundinAppendixA.

Apossibleconcernwithoursampleisthat,becausetheADRfirmsarelistedintheU.S.,theinsidertradingrestrictionsthatapplytothemaredifferentfromthosethatapplytothegeneralpopulationoffirmsintheirhomecountry.Forexample,perhapsbylistingintheU.S.,ADRfirmsaresubjecttoU.S.insidertradingregulation.However,becauseoftheSEC'slong-standingpolicygoaloffacilitatingaccessofforeignissuerstoU.S.capitalmarkets,ADRfirmsareprovidedwithavarietyofexemptionstoU.S.insidertradingrules.Forexample,theExchangeActRule3a12-3(b)10exemptsforeignprivateissuersfromtheCommission'sproxyrules11,andfromtheinsiderstocktradingreportsandshort-swingprofitrecoveryprovisionsunderSection1612oftheExchangeAct.13RegulationFD,whichlimitsprivatecommunicationsofmaterialinformation,alsoexemptsforeignprivateissuersfromitscoverage.14Foreignprivateissuerswouldbesubjecttothepensionblackouttradingrestriction(RegulationBTR),aclarificationtoSection306(a)oftheSarbanes–OxleyActof2002,butonlyunderacertaincondition.15Thus,ADRfirmsarenotsubjecttothesamelevelofinsidertradingregulationasU.S.firms.Moreover,becausethemajorityofADRfirmsarealsolistedontheirdomesticexchange,theyarerequiredtofollowdomesticinsidertradinglaws.Finally,evenifADRfirmsareaffectedprimarilybyU.S.insiderlawandenforcement,thiswillbiasourtestsagainstfindinganyassociationbetweencompensationandourmeasuresofinsidertradingrestrictions.

ArelatedconcernisthatbecauseADRfirmstendtobelargerandmoreprofitablethanthetypicalfirmintheirhomecountry,ourfindingsaresubjecttoasampleselectionbias.Itisimportanttopointout,however,thatanysuchselectionbiasappliestoallofthecountriesthatmakeupoursample.Thus,itisunlikelytobiasourestimatesofthecross-sectionalassociationbetweencompensationandinsidertradingrestrictions.

Nonetheless,toprovidefurtherevidenceonthisissue,wecompareourdatawiththesummarycompensationdatafromFernandesetal.(2013).Forthe23countriesthatarecommontobothstudies,ourcompensationmeasures,TotalPayandEquityPayRatio,bothhavecorrelationcoefficientsofaround0.55(significantatthe1%level)withsimilarmeasuresinFernandesetal.(2013).Thesehighcorrelationsimplythat,despitepotentialdifferencesbetweenADRfirmsandtheirlocal

Forbothrestrictedstockawardsandoptiongrants,weusethegrant'smarketvalueatthetimeoftheaward.Thegrantdatefairvalueofoptionsgrantedisusedinthecalculation,whichislargelymadeavailableinthe20-F,followingtheInternationalFinancialReportingStandards2(IFRS2).

10

Alsoreferredtoas17CFR240.3a12-3(b).11

17CFR240.14a-1etseq.12

15U.S.C.78p.13

See,forexample,OwnershipReportsandTradingbyOfficers,DirectorsandPrincipalSecurityHolders,RIN3235-AI62,FinalRule,Securities

andExchangeCommission,footnote12.Alternatively,seeForeignIssuerReportingEnhancements,RIN3235-AK03,FinalRule,SecuritiesandExchangeCommission,footnote37.

14

SelectiveDisclosureandInsiderTrading,RIN3235-AH82,FinalRule,SecuritiesandExchangeCommission.15

InsiderTradesDuringPensionFundBlackoutPeriods,RIN3235-AI71,FinalRule,SecuritiesandExchangeCommission.Accordingtotherule,foreign

privateissuersareonlysubjecttoRegBTRiftheblackoutweretoaffectatleast50%ofthepensionplanbeneficiarieslocatedwithintheU.S.andsuchpersonsrepresentedmorethan15%ofallparticipantsandbeneficiariesunderallindividualaccountplansoftheissuer.

9