透明度與公司治理

TransparencyandCorporateGovernanceMAuthor:BenjaminHamelinAbstractAnobjectiveofmanyproposedcorporategovernancereformsisincreasedtransparency.Thisgoalhasbeenrelativelyuncontroversial,asmostobserversbelieveincreasedtransparencytobeunambiguouslygood.Wearguethat,fromacorporategovernanceperspective,therearelikelytobebothcostsandbenefitstoincreasedtransparency,leadingtoanoptimumlevelbeyondwhichincreasingtransparencylowersprofits.Thisresultholdsevenwhenthereisnodirectcostofincreasingtransparencyandnoissueofrevealinginformationtoregulatorsorproduct-marketrivals.Weshowthatreformsthatseektoincreasetransparencycanreducefirmprofits,raiseexecutivecompensation,andinefficientlyincreasetherateofCEOturnover.Wefurtherconsiderthepossibilitythatexecutiveswilltakeactionstodistortinformation.Weshowthatexecutivescouldhaveincentives,duetocareerconcerns,toincreasetransparencyandthatincreasesinpenaltiesfordistortinginformationcanbeprofitreducing.IntroductionsInresponsetorecentcorporategovernancescandals,governmentshaverespondedbyadoptedanumberofregulatorychanges.Onecomponentofthesechangeshasbeenincreaseddisclosurerequirements.Forexample,Sarbanes-Oxley(sox),adoptedinresponsetoEnron,WorldCom,andotherpublicgovernancefailures,requireddetailedreportingofoff-balancesheetfinancingandspecialpurposeentities.Additionally,soxincreasedthepenaltiestoexecutivesformisreporting.Thelinkbetweengovernanceandtransparencyisclearinthepublic's(andregulators')perceptions;transparencywasincreasedfbrthepurposeofimprovinggovernance.Yet,mostacademicdiscussionsabouttransparencyhavenothingtodowithcorporategovernance.Themostcommonlydiscussedbenefitoftransparencyisthatitreducesasymmetricinformation,andhencelowersthecostoftradingthefirm'ssecuritiesandthefirm'scostofcapital.Tooffsetthisbenefit,commentatorstypicallyfocusonthedirectcostsofdisclosure,aswellasthecompetitivecostsarisingbecausethedisclosureprovidespotentiallyusefulinformationtoproduct-marketrivals.Whilebothofthesefactorsareundoubtedlyimportantconsiderationsinfirms'disclosuredecisions,theyarenotparticularlyrelatedtocorporategovernance.Inthispaper,weprovideaframeworkforunderstandingtheroleoftransparencyincorporategovernance.WeanalyzetheeffectthatdisclosurehasonthecontractualandmonitoringrelationshipbetweentheboardandtheCEO.Weviewthequalityofinformationthefirmdisclosesasachoicevariablethataffectsthecontractsthefirmanditsmanagers.Throughitsimpactoncorporategovernance,higherqualitydisclosurebothprovidesbenefitsandimposescosts.Thebenefitsreflectthefactthatmoreaccurateinformationaboutperformanceallowsboardstomakebetterpersonneldecisionsabouttheirexecutives.Thecostsarisebecauseexecutiveshavetobecompensatedfortheincreasedrisktotheircareersimplicitinhigherdisclosurelevels,aswellasfortheincrementalcoststheyincurtryingtodistortinformationinequilibrium.Thesecostsandbenefitscomplementexistingexplanationsfordisclosure.Moreover,becausetheyaredirectlyaboutcorporategovernance,theyareinlinewithcommonperceptionsofwhyfirmsdiscloseinformation.WeformalizethisideathroughanextensionofHamelinandWasatch(1998)andHamelin's(2005)adaptationofHailstorm's(1999)career-concernsmodeltoconsiderthequestionofoptimaltransparency.Section2laysoutthebasicsofthismodel,inwhichthecompanychoosesthe“quality"oftheperformancemeasurethatdirectorsusetoassesstheCEO'sability.Inthismodel,theoptimalqualityofinformationforthefirmtorevealcanbezero,infinite,orafinitepositivevaluedependingontheparameters.Whenwecalibratethemodeltoreflectactualpubliclytradedlargeuscorporations,wefindthattheparametersimpliedbythecalibrationleadtoafinitevaluefbroptimaldisclosurequality.Thus,ouranalysissuggeststhatdisclosurerequirementsgoingbeyondthisoptimallevelarelikelytohaveunintendedconsequencesandtoreducevalue.WeevaluatetheimplicationsofpenaltiesandincentivesthatpotentiallyaffectthemotivesofCEOstodistorttheinformationcomingfromtheirfirms.Measuresthatpunishexaggeratingeffortcanbeeffectiveiftheyaresufficientlyseveretocurtailthiseffort;however,relativelyminorpenaltiescanbecounterproductive.Inaddition,incentivesfbrCEOstoimprovetheaccuracyofinformationcanharmshareholdersbecausesuchincentivespushaCEOtodisclosemorethanthevalue-maximizingquantityofinformation.ConcealingInformationInlightofsomerecentcorporatescandals,oneconcernisnotthatexecutivesdistortinformation,butratherthattheyconcealit.Inthissubsection,webrieflyaddresswhatouranalysiscansaywithrespecttoconcealinginformation.OnequestioniswhethertheotherplayersknowiftheCEOhasconcealedinformation?Ifso,thenpresumablytheycanpunishtheCEOfornon-disclosureMoreover,ifitiscommonknowledgethattheCEOknowsthevalueofsignalsthatheconceals,thenanunravelingargument(Grossman,1981)applies:Whatevertheinferredexpectedvalueofunrevealedsignalsis,theCEOwillhaveanincentivetorevealthoseabovethatexpectedvalue.Hence,theonlyequilibriumisoneinwhichunrevealedsignalsareinferredtohaveIhelowestpossiblevalueandtheCEOiscorrespondinglyinducedtorevealallsignals.Wepredictthereforethatconcealmentisunlikelytobeanissueiftheotherplayersknowwhatthesetofsignalsis.Suppose,incontrast,thattheotherplayersdidnotknowwhatthecompletesetofsignalswas(e.g.,thesetvariesovertime).IftheCEOdidnotknowtherealizedvaluewhenhedecidingtorevealorconcealasignal,thenhewouldwishtoconcealallsignalsthathecould:moresignalsmeansamorepreciseposteriorestimateofhisability,whichmeansgreatercareerriskforhim.Ourmodel,thus,predictsthatwhen(i)theCEOhasdiscretionoverwhatsignalsarerevealedand(ii)mustcommittorevealorconcealpriortolearningthevalueofthesignals,hewillchoosetocommittoconcealallsignalsoverwhichhehasdiscretion.If,instead,theCEOisnotcommittedtoadisclosuredecisionpriortolearningthevalueofthesignals,thenhewillbetemptedtorevealthosethatarcfavorabletohim.Theotherplayerswillinferthattheyaregettingabiasedsampleand,thus,makeadownwardadjustment.Inthissense,thesituationissimilartothatof^exaggeratingeffort."Thedetailsoftheanalysisare,tobesure,differentandawaitfutureanalysis,butourgeneralconclusionswillgenerallyhold.DiscussionsandConclusionMostcorporategovernancereformsinvolveincreasedtransparency.Yet,discussionsofdisclosuregenerallyfocusonissuesotherthangovernance,suchasthecostofcapitalandproduct-marketcompetition.Thelogicofhowtransparencypotentiallyaffectsgovernanceisabsentfromtheacademicliterature.Wcprovidesuchanalysisinthispaper.WeshowthattheleveloftransparencycanbeunderstoodasderivingfromthegovernancerelationbetweentheCEOandtheboardofdirectors.Thedirectorssettheleveloftransparency(e.g.,amountandqualityofdisclosure)anditis,thus,partofanendogenouslychosengovernancearrangement.Increasingtransparencyprovidesbenefitstothefirm,butentailscostsaswell.Bettertransparencyimprovestheboard'smonitoringoftheCEObyprovidingitwithanimprovedsignalabouthisquality.Butbettertransparencyisnotfree:ThebetterablethemarketistoleamabouttheCEO'sability,thegreatertherisktowhichtheCEOisexposed.Inoursetting,theprofit-maximizingleveloftransparencyrequiresbalancingthesetwofactors.Ourmodelimpliesthattherecanbeanoptimalleveloftransparency.Consequently,attemptstomandatelevelsbeyondthisoptimumdecreaseprofits.Profitsdecreasebothbecausemanagerswillhavetobepaidhighersalariestocompensatethemfortheincreasedcareerrisktheyface,andbecausegreatertransparencyincreasesmanagerialincentivestoengageincostlyandcounterproductiveeffortstodistortinformation.Weemphasizethattheseeffectsoccurinamodelinwhichallotherthingsequal,betterinformationdisclosureincreasesfirmvalue.Onekeyassumptionwemakethroughoutthepaperisthattheboardreliesonthesameinformationthatisreleasedtothepublicinmakingitsmonitoringdecisions.Undoubtedly,thisassumptionisliterallyfalseinmostfirms,astheboardhasaccesstobetterinformationthanthepublic.Nonetheless,CEOsdohaveincentivestomanipulateinformationtransferstoimprovetheboard'sperceptionofthem,andthisideahasbeenanimportantfactorinanumberofrecentstudies(see,c.g.,AdamsandFerreira,inpress).Inaddition,inanumberofpublicizedcases,boardshavebeenkeptinthedarkexceptthroughtheirabilitytoaccesspubliclydiscloseddocuments;thecircumstancesinwhichboardsmustrelyonpubliclyavailableinformationarelikelythecasesinwhichtheboard-CEOrelationshipismostadversarial,andhencearethecasesinwhichboardmonitoringislikelymostimportant.Certainly,ourbasicassumptionthatthequalityofpublicdisclosurehasalargeimpactontheboard'sabilitytomonitormanagementisplausible.Ourmodelissetinthecontextofaboardthatisperfectlyalignedwithshareholders?interests.Onecouldequallywellapplyittodirectmonitoringbyshareholders.Iftherewereanincreaseinthequalityofavailableinformationeitherduetomorestringentreportingrequirementsorbecauseofbetteranalysis(e.g.,ofthesortperformedbyinstitutionalinvestorsoramoreattentivepress),thenourmodelcontainsclearempiricalpredictions.Inparticular,itsuggeststhatconsequencesofimprovedinformationwouldbeincreasesinCEOsalariesandtherateofCEOturnover.Infact,bothCEOsalariesandCEOturnoverhaveincreasedsubstantiallystartinginthe1990s,withatleastsomescholars5attributingtheincreasetothehigherlevelofpressscrutinyandinvestoractivism(seeKaplanandMinton,2006).ThispatternofCEOsalariesandturnoverisconsistentwithourmodel;moreover,itisconsistentwiththeideathatbetterinformationhasbothcostsandbenefitsthroughitsimpactoncorporategovernance.Someissuesremain.Asdiscussedabove,wehaveonlyscratchedthesurfacewithrespecttoissuesofmanagerialconcealmentofinformation.Wehaveabstractedawayfromanyoftheconcernsaboutrevealinginformationtorivalsorregulatorsthatearlierworkhasraised.Wehavealsoignoredothercompetingdemandsforbetterinformation,suchastohelpbetterresolvetheprincipal-agentproblemthroughincentivecontracts(seee.g.,GrossmanandHart,1983,Singh,2004).Finally,wehaveignoredthemechanicsofhowthefirmactuallymakesinformationmoreorlessinformative(e.g.,whataccountingrulesareused,whatorganizationalstructures,suchasreportinglinesandofficeorganization,areemployed,etc.).Whilefutureattentiontosuchdetailswill,webelieve,shedadditionallightonthesubject,weremainconfidentthatourgeneralresultswillcontinuetohold.譯文透亮度與公司治理資料來(lái)源:哈佛商學(xué)院教育網(wǎng)作者：本杰明?埃爾馬蘭，邁克爾?魏斯巴赫1摘要很多提出的關(guān)于公司治理改革的目標(biāo)是提高透亮度。這個(gè)目標(biāo)已比較不會(huì)引起爭(zhēng)議，由于大多數(shù)觀看家認(rèn)為提高透亮度還是清晰明確的比較好。我們認(rèn)為，從公司治理的角度來(lái)看，可能存在著成本與效益均提高了透亮度，導(dǎo)致超過(guò)其最佳水平即提高了透亮度卻降低了利潤(rùn)。這一結(jié)果認(rèn)為，即使沒(méi)有提高透亮度的直接成本，也沒(méi)有向監(jiān)管機(jī)構(gòu)或產(chǎn)品市場(chǎng)競(jìng)爭(zhēng)對(duì)手信息披露的問(wèn)題。我們的分析表明，改革旨在提高透亮度可以削減公司利潤(rùn)，提高管理層薪資水平并提高缺乏力量的首席執(zhí)行官的離職率。我們應(yīng)進(jìn)一步考慮存在著行政管理人員將實(shí)行措施來(lái)扭曲信息的可能性。我們的分析表明，由于就業(yè)問(wèn)題管理人員就可以有動(dòng)力，提高透亮度和增加信息的扭曲，從而增加罰金降低利潤(rùn)。2論文簡(jiǎn)介對(duì)于最近的公司治理丑聞，各國(guó)政府已作出回應(yīng)，并實(shí)行了一系列的監(jiān)管調(diào)整措施。這些調(diào)整措施的一個(gè)組成部分是提高信息披露要求。例如，薩班斯-奧克斯利法案說(shuō)明，應(yīng)對(duì)安穩(wěn)、世通和其他公共治理的失敗，要求對(duì)資產(chǎn)負(fù)債表表外融資和特殊目的實(shí)體進(jìn)行具體報(bào)告。此外，薩班斯-奧克斯利法案中提出增加了管理人員誤報(bào)的懲罰。關(guān)于治理與透亮度之間的關(guān)系，公眾和監(jiān)管機(jī)構(gòu)的看法是明確的，即增加透亮度是為了提高治理。然而，大多數(shù)關(guān)于透亮度的學(xué)術(shù)爭(zhēng)論認(rèn)為透亮度與公司治理沒(méi)有關(guān)系。最常見(jiàn)的關(guān)于信息透亮度的好處的爭(zhēng)論認(rèn)為它削減了信息不對(duì)稱，從而降低了公司證券的交易成本和公司的資本成本。為了彌補(bǔ)這個(gè)優(yōu)勢(shì)，評(píng)論家通常集中在信息披露的直接費(fèi)用以及產(chǎn)生的競(jìng)爭(zhēng)力成本上，由于披露可能為產(chǎn)品市場(chǎng)競(jìng)爭(zhēng)者供應(yīng)有用的信息。而這兩個(gè)因素在公司信息披露的打算中無(wú)疑是重要的考慮因素，它們并不特殊與公司治理相關(guān)。在本文中，我們供應(yīng)了一個(gè)了解透亮度在公司治理中的作用的框架。我們分析了信息披露在合同方面以及董事會(huì)與首席執(zhí)行官之間的監(jiān)測(cè)方面的作用。我們認(rèn)為公司信息披露質(zhì)量作為選擇變量影響了合同的企業(yè)及其管理者。通過(guò)其對(duì)公司治理的影響，更高質(zhì)量的信息披露能供應(yīng)效益和強(qiáng)加的成本。這些效益反映的事實(shí)是更精確的業(yè)績(jī)信息能促使董事會(huì)做出更好的關(guān)于他們的管理人員的人事打算。這些成本的產(chǎn)生是由于管理人員在較高的隱性披露水平常必需取得增加他們事業(yè)風(fēng)險(xiǎn)的賠償，以及他們?cè)噲D歪曲平衡的信息引發(fā)的增量成本。這些成本和效益補(bǔ)充了披露的現(xiàn)有解釋。此外，由于他們直接對(duì)企業(yè)進(jìn)行治理，在與公司為什么要披露信息的公共看法全都。我們正式通過(guò)埃爾馬蘭，魏茲巴赫(1998),埃爾馬蘭(2005)擴(kuò)展了霍姆斯特羅姆(1999)的職業(yè)生涯關(guān)注模型的適應(yīng)性的想法是考慮了最優(yōu)透亮度的問(wèn)題。其次章展現(xiàn)了最基本的模型，在該公司選擇績(jī)效的“質(zhì)量”指標(biāo)來(lái)衡量，從而使董事用來(lái)評(píng)估首席執(zhí)行官的力量。在這個(gè)模型中，公司的信息披露質(zhì)量最優(yōu)可以是零，無(wú)限的，或一個(gè)有限價(jià)值取決于參數(shù)。當(dāng)我們校準(zhǔn)模型來(lái)反映我們的實(shí)際大型上市公司，我們發(fā)覺(jué)所示意的參數(shù)校準(zhǔn)導(dǎo)致最優(yōu)信息披露質(zhì)量的有限價(jià)值。因此，我們的分析表明，信息披露要求超出了這個(gè)最佳水平很可能產(chǎn)生意想不到的后果，并削減價(jià)值。我們?cè)u(píng)估的懲處和嘉獎(jiǎng)可能影響首席執(zhí)行官的動(dòng)機(jī)去扭曲他們企業(yè)的信息。懲處措施夸大了工作的有效性，假如他們能足夠遏制這一努力的話；然而,相對(duì)較輕的懲罰可能會(huì)適得其反。此外，為了提高信息的精確性，首席執(zhí)行官們的嘉獎(jiǎng)可能會(huì)損害股東利益，由于這種激勵(lì)措施推動(dòng)首席執(zhí)行官透露的信息價(jià)值最大化的數(shù)量更多。3隱蔽信息在最近的一些公司丑聞的曝光中，關(guān)懷的不是管理人員歪曲信息，而是隱蔽信息。在本小節(jié)中，我們簡(jiǎn)潔談?wù)勎覀兊姆治隹梢哉f(shuō)是關(guān)于隱瞞信息。一個(gè)問(wèn)題是，假如首席執(zhí)行官隱蔽了信息是否其他參加者都知道？假如是這樣，那么他們也許可以懲處就非公開(kāi)的首席執(zhí)行官。此外，假如眾所周知，首席執(zhí)行官知道他隱蔽的信號(hào)值，進(jìn)而闡明參數(shù)（格羅斯曼，1981）適用于：凡是未顯露信號(hào)推斷預(yù)期值時(shí)，首席執(zhí)行官將有動(dòng)機(jī)透露上述那些預(yù)期的價(jià)值。因此，唯一的均衡是指未透露的信號(hào)推斷出的可能最低價(jià)值和首席執(zhí)行官相應(yīng)的誘導(dǎo)顯示全部的信號(hào)。因此，我們猜測(cè)隱瞞不太可能是一個(gè)問(wèn)題，假如其他參加者知道什么是信號(hào)集合。假設(shè)，與此相反，其他參加者不知道什么是一套完整的信號(hào)（例如，隨時(shí)間的變化而異）。假如首席執(zhí)行官在打算顯示或隱蔽一個(gè)信號(hào)時(shí)，不知道實(shí)現(xiàn)價(jià)值，那么他就會(huì)想要隱蔽他的全部信號(hào)：更多的信號(hào)是指以更精確的方法估量出的他的力量，這意味著他的職業(yè)風(fēng)險(xiǎn)更大。因此，我們的模型猜測(cè)，（一）當(dāng)首席執(zhí)行官打算揭示什么信號(hào),（二）必需承諾事先學(xué)習(xí)顯示或隱蔽的信號(hào)價(jià)值,他會(huì)選擇隱瞞承諾過(guò)的全部信號(hào)。相