上财会计与财务研究方法论李增泉教授-PPT课件.pptx

引子：独立董事的价值2010年度优秀独立董事证监会是否应该规定公司必证监会是否应该规定公司必须聘请独立董事？须聘请独立董事？能够监督内部人能够为公司提供咨询能够降低监管部门的责任能够为会计教授提供赚钱机会不能够监督内部人不能够为公司提供咨询不能够降低监管部门的责任不能够为会计教授提供赚钱机会他们真的很厉害？他们真的很厉害？可以抑制掏空、提供咨询、可以抑制掏空、提供咨询、扶贫教授、保护官员？扶贫教授、保护官员？答案好像是唯一的！答案好像是唯一的！推理可信吗？事实可靠吗？以独立董事为例有人发现，独立董事越多（比例或人数，或有/没有）的公司，价值（会计指标或股价指标）越高。疑问1：有其他解释吗？疑问2：穷尽了任何公司的任何时点吗？疑问3：事实可信吗？（何谓公司价值）该发现可该发现可以说明独以说明独立董事的立董事的价值吗？价值吗？以上例子中的方法论问题研究问题：实证性命题（是什么？为什么）与规范性命题（怎么办？）检验方法：推理与感知、归纳与演绎计量方法：内生性6PartI:Methodology7Chap1:ThePhilosophyoffinanceandaccountingresearch8ResearchProgressTheResearchProgressThestatementoftheresearchproblemorissuetobeinvestigated,orthetheoreticalexplanationtobetestedThetranslationoftheabstractideasinthistheoreticalexplanationintoconcrete,explicitlyidentifiableideasThedevelopmentofmeasuresoftheimportantvariablesinthetheoreticalexplanation9ThedevelopmentofaresearchdesigntoguidetheinquiryintotheresearchproblemThecollectionoftheresearchdataTheanalysisofthisinformationinthecontextoftheproposedexplanationTheinterpretationoftheinformation,orintegrationofthefindingsoftheresearchwiththeexistingknowledgebase10SomeissuesonphilosophyWhatisresearch?Researchisaprocessofintellectualdiscovery,whichhasthepotentialtotransformourknowledgeandunderstandingoftheworldaroundus.ResearchmethodologicalvariedwithsuchassumptionsaboutThenatureofrealityTheroleoftheoryThesignificanceofempiricalexperimentationWhyisresearchmethodologyimportant?Whenapieceofresearchischaracterizedbypoortechnique,acriticmayarguethattheresearchisdefective,weakormisapplied,however,whenamethodologicaldisputeisinvolvedtheresearchissimplylabeledasnonsensical”.11Researchinaccountingandfinanceisgenerallyacceptedasbeingsocialscientific,asappropriatestandardsofscientificenquiryareappliedtosocialissuesratherthannaturalphenomena,whichistakentobethedomainofthenaturalsciencesandofphysicsinparticularThesocialscientificisseriousaffectedbyDualitywithinwesternthoughtRealitycouldbecharacterizedbyoppositesThereisanessentialdualityinallthings12Individual subjectknowledgebeliefsperceptionsappearancereasonExternal object13EpistmologyorwhatisknowledgeEmpiricismandrationalism:thesourceofknowledgeSocrateandPlatoVs.Aristotle|Descartes,HegelandMarxVs.NewtonTraditionally,classicalempiricistsacceptedthatCertaintyofbeliefinwhatweknowcanonlybeapproachedthroughperceptionUltimatelyallknowledgeisderivedfromperceptionthroughoursensesIntherealmofdiscoursestatementsareeithertrueorfalsebecauseofthewaytheworldisorbecauseofsomeformalpropertiesofthelanguageweused14Realismversusidealism:ontologyofwhatweknowRealismsholdthatrealitysubsistswithinobjects;idealistsholdthatitexistswithinthemindofthesubjectHumeVs.BerkeleyKant:TranscendentalidealismThereisanobjectiveworldorexperiencewhichwerelateandultimatelytestourclaimstoknowledgeofwhatistrueorfalsewecanknowthembytheapplicationofcertaintranscendentprinciplesrelativism15LogicalpositivismandinstrumentalismLinguisticderivativeofempiricismwheretheworldofmeaningfuldiscoursediscontrolledbyaparticularvarietyofthecorrespondencetheoryoftruthHoldthat:Truebeliefisgroundedinwhatweperceiveandthatwhatweperceiveisderivedfromavalue-free,independentrealityMeaningfulstatementsareonlythosewhichcan,inprincipleatleast,beverifiedbyappealtoobservation16TwodifficultiesThevalidityoftheverificationprincipleitselfGenerallawsandtheoreticaltermsInstrumentalismTherealismofgiventheoreticaltermsisquiteirrelevantindeterminingthevalidityofanytheoreticalconstructionsderivedfromthemorinwhichtheyareembedded.E.g.,Thefactthatatheoryisunrealisquiteirrelevantprovidedthatitworksinpractice.Friedman(1953)Falsificationism(Popper,1959)Anystatementwhichcannotbedemonstratedtobefalseismeaninglessandanytheorywhichcantbefalsifiedisdevoidofempiricalcontent17PositiveeconomicsFriedman(1953)Alchian(1950)张五常（2001）HolthausenaandWatts(2001)Uncertainty,EvolutionandEconomicTheoryArmenA.AlchianInJournalofPoliticalEconomy58,no.3(June1950):211-221OutlineInthispaper,AlchiansuggestsanapproachThisapproachembodiestheprinciplesofbiologicalevolutionandnaturalselectionbyinterpretingtheeconomicsystemasanadoptivemechanismwhichchoosesamongexploratoryactionsgeneratebytheadaptivepursuitofsuccessorprofitThismodificationincorporateincompleteinformationanduncertainforesightasaxioms,butdispensewithprofitmaximization,anddoesnotrelyonthepredictableindividualbehaviorThemethodofuse,ratherthantheusefulness,ofeconomictoolsandconceptsisaffectedbythismodification“ProfitMaximization”IsNotAGuideToActionCurrenteconomicanalysisofeconomicbehaviorreliesheavilyondecisionsmadebyrationalunitscustomarilyassumedtobeseekingperfectlyoptimalsituations(eg.Profitmaximizationorutilitymaximization).Attacksonthismethodologyarewidespread,butonlyTintnersattackhasbeenreallydamaging.Hedeniesthatprofitmaximizationevenmakesnosenseunderuncertaintybecauseunderuncertaintyeachactionthatmaybechosenisidentifiedwithadistributionofpotentialoutcomes,notwithauniqueoutcome.UsingoptimumdistributionasasubstitutionstillpostulateamodecontainingcertaintySuccessIsBasedOnResults,NotMotivationTheapproachsuggestedtreatsthedecisionsandcriteriadictatedbyeconomicsystemasmoreimportantthanthosemadebytheindividualsinit.ThisapproachdirectsattentiontotheinterrelationshipsoftheenvironmentandtheprevailingtypesofeconomicbehaviorwhichappearthroughaprocessofeconomicnaturalselectionSuccess(survival)accompaniesrelativesuperiorityItdoesntrequirepropermotivationbutmayratherbetheresultoffortuitouscircumstancesChanceorLuckIsOneMethodofAchievingSuccessTwosubstantialelementsindeterminingthesituationselectedanditsappropriatenessorviabilityChanceorluckConsciousadaptingThesurvivorsmayappeartobethosehavingadaptedthemselvestotheenvironment,whereasthetruthmaywellbethattheenvironmenthasadoptedthemEg.ThetravelersandgasolinestationChanceDoesNotImplyNondirected,RandomAllocationofResourcesTheimplicationofsheerchancemodelIndividualrandombehaviordoesnoteliminatethelikelihoodofobservingappropriatedecisionsIndividualbehavioraccordingtosomeforesightandmotivationdoesntnecessarilyimplyacollectivepatternofbehaviorthatisdifferentfromthecollectivevarietyofactionsassociatedwitharandomselectionofactionsChanceDoesNotImplyNondirected,RandomAllocationofResourcesAchance-dominatedmodeldoesnotmeanthataneconomistcannotpredictorexplainordiagnoseThepredictionwillnotassertthateveryfirmnecessarilychangesitscharacteristicsEmpiricalinvestigationsviaquestionnairesmethodsareincapableofevaluatingthevalidityofmarginalproductivityanalysisIndividualAdaptingViaImitationandTrialandErrorTwoformsofconsciousadaptivebehaviorImitationTrialanderrorImitationandinnovationsImitationaffordsrelieffromthenecessityofreallymakingdecisionsandconsciousinnovations.ThewillingnesstoabandonitattherighttimeandcircumstancesaccountsforsuccesswhenconditionshavechangedInnovationisalsoaccountedforbyimitationIndividualAdaptingViaImitationandTrialandErrorTrialandprofitmaximizationTwoconditionsforconvergencefromtrialtoprofitmaximizationAtrialmustbeclassifiableasasuccessorfailureContinualrisingtowardsomeoptimumoptimorumwithoutinterveningdescentsTheaboveconvergenceconditionsdonotapplytoachangingenvironment,fortherecanbenoobservablecomparisonoftheresultofanactionwithanyotherSuccessisdiscoveredbytheeconomicsystemthroughablanketingshotgunprocess,notbytheindividualthroughaconvergingsearchConclusionsObservablepatternsofbehaviorandorganizationarepredictableintermsoftheirrelativeprobabilitiesofsuccessorviabilityiftheyaretriedUsingthepresentanalyticaltoolsdevelopedintheanalysisofthefirmundercertainty,canpredictthemoreadoptableorviabletypesofeconomicinterrelationshipsthatwillbeinducedbyenvironmentalchangeevenifindividualsthemselvesareunabletoascertainthemDoesthispaperindicatethatthepresenttoolsforthediagnosisoftheoperationofeconomicsystemarealsofortheinternalbusinessbehaviorofeachfirm?TheMethodologyofPositiveEconomicsMiltonFriedmanIn,1953,ChicagoU.PressOutlineInthispaper,FriedmanwilldiscusstherelationshipbetweenpositiveandnormativeeconomicsThetargetandelementsofpositivetheoryTheroleofassumptionsofpositivetheoryTheRelationshipBetweenPositiveandNormativeEconomicsJohnNevilleKeynes(1891)ApositivescienceisabodyofsystematizedknowledgeconcerningwhatisAnormativeorregulativescienceisabodyofsystematizedknowledgediscussingcriteriaofwhatoughttobeAnartissystemofrulesfortheattainmentofgivenendNormativeeconomicscannotbeindependentofpositiveeconomicsthecurrentlyinUSdifferencesabouteconomicpolicyamongdisinterestedcitizensderivepredominantlyfromthepredictionconsequenceofsuchpolicy,whichisthequestionofpositiveeconomics,ratherthanfromfundamentaldifferencesinbasicvalues,whichisthebasesofnormativeeconomicsEg.Minimum-wagelegislationPositiveEconomicsTheultimategoalofapositivescienceisthedevelopmentofatheoryorhypothesisthatyieldsvalidandmeaningfulpredictionsaboutphenomenanotyetobservedSuchatheoryisacomplexintermixtureoftwoelementsAlanguagedesignedtopromotesystematicandorganizedmethodsofreasoning.AbodyofsubstantivehypothesesdesignedtoabstractessentialfeaturesofcomplexrealityPositiveEconomicsViewedasalanguage,theoryhasnosubstantivecontent,itisasetoftautologies.Thecriteriabywhichitistobejudgedarethoseappropriatetoafilingsystem.thevalidofthiselementcanbeshownbylogicalandfactualconsideration.Eg.ThetermofSupplyandDemandViewedasabodyofsubstantivehypotheses,theoryistobejudgedbyitspredictivepowerfortheclassofphenomenawhichitisintendedtoexplain.OnlyfactualevidencecanshowwhetheritisvalidPositiveEconomicsThechoiceamongalternativehypothesesmusttosomeextentbearbitrary,thoughsimplicityandfruitfulnessisthegeneralagreedcriteriaAtheoryissimplerthelesstheinitialknowledgeneededtomakeapredictionwithinagivenfieldofphenomenaItismorefruitfulthemoreprecisetheresultingprediction,thewidertheareawithwhichthetheoryyieldspredictions,andthemoreadditionallinesforfurtherresearchitsuggestsPositiveEconomicsThecharactersofempiricalevidenceThenecessityofrelyingonuncontrolledexperienceratherthanoncontrolledevidencemakesitdifficulttoproducedramaticandclearcutevidencetojustifytheacceptanceoftentativehypotheses.Empiricalevidenceisvitalattwodifferentstages:inconstructinghypothesisandintestingtheirvalidityTheroleofassumptionsAtheorycantbetestedbytherealismofitsassumptions,butbythepredictionofthetheoryTherelationshipbetweenthesignificanceofatheoryandtherealismofitsassumptionsisoppositeEg.S=1/2gt2ThedensityofleavesaroundatreeThemaximizationofreturnshypothesisTheroleofassumptionsTheassumptionsofatheoryplaythreedifferentpositiverolesAneconomicmodelofdescribingorpresentingatheoryTheassumptionsistheconceptualworldofabstractmodelsimplerthantherealworld,andcontainingonlytheforcesthattheforcesthatthehypothesisassertstobeimportant.Thereisnoplaceinthemodelfor,andnofunctiontobeservedby,vagueness,maybes,orapproximations.TherulesforusingthemodelcantbeabstractandcompleteFacilitateandindirecttestofthehypothesisbyitsimplicationsAconvenientmeansofspecifyingtheconditionsunderwhichthetheoryisexpectedtobevalidConclusionsApositiveeconomicsisconcerningwhatis,andcanbeaninputofnormativeeconomicsApositivetheoryisacomplexintermixturewithalanguageandabodyofsubstantivehypothesis.Thehypothesisismadeoftwoparts:theassumptionsandtherules.Thevalidofpositivetheorycantbejudgedbytherealismofassumptions,butthepredictionofthistheoryCouldatheorybeimprovedbymodifyingtheassumptions?科学的方法张五常摘自经济解释卷一科学说需求，2001，香港花千树出版社提纲在本文中，张五常将要讨论主观感情对经济研究的影响科学方法的几个原则科学理论的现实性核心论点科学不是求对，也不是求错；科学所求的是“可能被事实推翻”。可能被事实推翻而没有被推翻，就算是被解释了人与科学人的感情往往与理智混淆，这样科学上的推断可能被感情左右，搞得拖泥带水，但也可以精彩绝伦，使人觉得妙不可言。科学可以有艺术的美。但是，科学的本质不是艺术，而是以解释现象为主旨。科学的著作可以将客观分析与主观感情相结合、并用，但二者要分得清楚主观感情与客观分析的清楚划分比较困难，但并非说不能做到，而是比自然科学困难。现象必有规律不管是大自然还是人的行为，都有规律可循（寻）科学的形成是基于三个重要的信念，是任何对科学感兴趣的人都要遵守的主观的现象要被客观的认同，共信所有被众所认同的现象都是有迹可循，有规律的假若现象没有规律，那么就不能被系统的解释任何规律的发生都是有原因的事实不能解释事实解释现象需要非事实的抽象理论这是因为事实的规律不能不言自明，自我解释若用事实解释事实，会存在以下两个困难世界上的现象规律何其多也，从而理论就会汗牛充栋，各不相同，从而毫无一般性的解释能力有规律的现象，在不同的情况下，其规律可能会改变事实不能解释事实几位名人观点事实不能以事实作解释（K.Brunner)事实的规律是要被解释的（M.Friedman)这些争议的经验告诉我们，除非经过理智的考究与解释，我们不可能从事实中学到些什么。这也教训了我们，使我们知道最鲁莽而又虚伪的，是那些公开声言让事实自作解释的理论家，而或者无意识地，自己在幕后操纵事实的选择与组合，然后提出如下类的推论：在这之后所以这就是原因。（A.Marshall)特殊理论与套套逻辑任何科学理论，若被事实推翻，我们总可以多加条件来挽救的，但挽救理论是须付代价的，所谓特殊理论（ad hoc theory)就是特殊到只能解释一个现象的理论例：温度重量/风/斜坡/海拔所谓套套逻辑（tautology），是指一些言论，在任何情况下都不可能是错的例如，四足动物有四只脚，争取个人最大化利益与跳楼套套逻辑并非不可能是一个重要的概念。假若我们能把范围加以约束、收窄，有时可以促成一个有内容的理论，例如，科斯定理大有可取的，足以解释世事的理论是处于两者之间的可能被事实推翻的重要性在整个科学方法的结构中最重要的一点是，理论的推测一定要可能 被事实推翻不可能被事实推翻的理论之所以没有解释力，是因为这样的理论不能被事实验证假若一个理论被事实推翻了，我们有两个选择放弃设法附加验证条件验证一个理论含意的唯一办法，是以事实反证，A-B,等价于，NOT B-NOT A.而非NOT A-NOT B例如Lester通过调查雇主如何确定司机工资的方法来推翻边际理论是错误的白痴建加油站，说明NOT A 并非推出NOT B.天狗吃月亮来推断月食，被现代时刻理论取代的原因并非天狗不会吃月亮，而是后者具有更强的一般性几种不能被事实推翻的情况套套逻辑模糊不清概念或分析模糊不清，不可能清楚地错，所以也不能清楚地被事实推翻例子马克思的“剩余价值”Pigou的福利经济学Richardo的资本与成本概念凯恩斯的通论Bentham的功用Knight的风险与不确定性的区分几种不能被事实推翻的情况没有意义没有意义并非空洞，而是因为言论互相矛盾，逻辑上前后不一致，例如，以自私人为前提，但推论却否定之非事实验证含意的事实性并不否认理论构建中的抽象无限制无论验证时所牵涉的现象怎样多，一个限制是必需的不均衡可解作推断的现象没有限制理论的真实性因为事实不能解释事实，以理论解释现象，在某种程度上一定是抽象的。抽象的思想当然不是事实，要说理论非真实也是可以的但有解释力的理论的非真实性，起码有四种含意理论本身必定有抽象的成分所有事实或观察的描述，一定要简化为了便于处理而简化事实验证条件的非真实性但无论怎样简化，也一定要与真实世界的情况大致吻合结论科学方法论牵涉到哲学上的逻辑学和知识理论，是人类文化历史上最湛深的学问但是科学的成就往往与此学问无关不懂科学方法论的科学高手屈指难算科学方法论的高手很少是有成就的科学家由于中国人的文化传统，往往大谈仁义道德，缺乏科学精神，对科学辩证的本质有根深蒂固的误解，我认为，科学方法论对中国人比对其他民族更为重要Therelevanceofthevalue-relevanceliteratureforfinancialaccountingstandardsettingRobertW.Holthausena,RossL.WattsInJournal of Accounting and Economics 31(2001)375OutlineThispaperwilldiscussWhatstheunderlyingtheoryofvalue-relevanceliteratures?Whytheyarenotdescriptive?Whatshouldwedo?theunderlyingtheoryWhatisthevalue-relevanceliterature?papersinvestigatetheempiricalrelationbetweenstockmarketvalues(orchangesinvalues)andparticularaccountingnumbersforthepurposeofassessingorprovidingabasisofassessingthosenumbersuseorproposeduseinanaccountingstandard.theunderlyingtheoryThreetypes(24-53-7/62)Relativeassociationstudiescomparetheassociationbetweenstockmarketvalues(orchangesinvalues)andalternativebottom-linemeasures.Incrementalassociationstudiesinvestigatewhethertheaccountingnumberofinterestishelpfulinexplainingvalueorreturns(overlongwindows)givenotherspecifiedvariables.MarginalinformationcontentstudiesinvestigatewhetheraparticularaccountingnumberaddstotheinformationsetavailabletoinvestorstheunderlyingtheoryTheoriesunderlyingsuchstudiesIndirectvaluationtheoryaccountingearningsisintendedtoeithermeasure,orbehighlyassociatedwith,equitymarketvaluechangesorlevels(viapermanentincome).Givendirectvaluationtheory,standardsetterswouldbeinterestedintheresultsofastudyoftherelativestockpriceassociationsofalternativeaccountingearningsorbookvalueofequitymeasures.theunderlyingtheoryIninputs-to-equityvaluationtheory,accountingsroleistoprovideinformationoninputstovaluationmodelsthatinvestorsuseinvaluingfirmsequity.Underaninputs-to-equityvaluationtheorystandardsettersaremorelikelytobeinterestedinastudythatsuggestsinvestorscoulduseanaccountingnumberorapotentialaccountingnumberintheirvaluationmodels.theunderlyingtheoryInbothcases,earningswouldmeasure,orbehighlyassociatedwithequitymarketvaluesorchangesinequitymarketvalues,andequitybookvalueswouldmeasure,orbehighlyassociatedwith,equitymarketvalues.Essentiallyvalue-relevancestudiesimplyaccountingsroleistoprovideestimatesofequitymarketvaluesorlineartransformationsofequitymarketva