CFA三级培训项目：Behavioral Finance 标准版.docx

U LDoI>二更UqSSdCUdWaaAe koll4尸。圈O皿邑暮 FK111V:Ducraczso>QxmTopicinCFALevelIIIStudySession1-2StudySession3StudySession4StudySession5StudySession6StudySession7-8StudySession9-10StudySession11StudySession12-13StudySession14StudySession15StudySession16ContentETHICS&PROFESSIONALSTANDARDS(1)&(2)BEHAVIORALFINANCECAPITALMARKETEXPECTATIONSNEWASSETALLOCATIONANDRELATEDDECISIONSINPORTFOLIOMANAGEMENTDERIVATIVESANDCURRENCYMANAGEMENTNEWFIXED-INCOMEPORTFOLIOMANAGEMENT(1)&(2)EQUITYPORTFOLIOMANAGEMENT&(2)ALTERNATIVEINVESTMENTSFORPORTFOLIOMANAGEMENTNEWPRIVATEWEALTHMANAGEMENT(l)8i(2)NEWPORTFOLIOMANAGEMENTFORINSTITUTIONALINVESTORSNEWTRADING,PERFORMANCEEVALUATION,ANDMANAGERSELECTIONNEWCASESINPORTFOLIOMANAGEMENTANDRISKMANAGEMENTNEWBehavioralFinancekSuccessKeys:RememberKeyWordsStudySomeCasesofTheseKeyWordsExplaintheKeyWordsinOneSentencekRecommendationsMuchoftheterminologyisredundantinthatmorethanonetermcanmeanthesamething.Manyoftheconceptsareoverlapping,andmostofthequestionsdependheavilyoncomprehendingtheterminology.Yourfocusshouldbeonunderstandingthebasicmeaningofeachtermasgiveninthematerial.s° y<CJ iSSSQUOJdIubeisauiPUP UUPLL-BJo->8 6工S-BnP>PUIOS S-8 -20一 >-s8 =18x>巧dsdUUBULz-ejo>cqCQCl NH yueuiz-e-o>eqCQ-ESS大O?Lriuueuiz-B-O5eq>oMEei®三三:>uQuueuiz-e,lo>eqQqI6e三pexTraditionalFinancevs.BehavioralFinanceWithintraditionalfinance,individualsareassumedtoberisk-averse,selfinterestedutilitymaximizers,(rational).Traditionalfinancefurtherhypothesizesthat,atthemarketlevel,pricesincorporateandreflectallavailableandrelevantinformation.kBehavioralfinanceincludesbehavioraleconomics,investorpsychology,behavioralscience,experimentaleconomics,andCognitiVePSyChC)IOgyThevarietyofapproachestakentoexamineinvestorbehavioraddstotheconfusionaboutwhatismeantbybehavioralfinance,(normal)behavioralfinancemicro(BFMI)examinesbehaviorsorbiasesthatdistinguishindividualinvestorsfromtherationalactorsenvisionedinneoclassicaleconomictheory./BFMIquestionstheperfectrationalityanddecision-makingprocessofindividualinvestors.orbehavioralfinancemacro(BFMA)considersmarketanomaliesthatdistinguishmarketsfromtheefficientmarketsoftraditionalfinance./BFMAquestionstheefficiencyofmarkets.Decision-MakingProcessToprovideaframeworkforunderstandingtheimplicationsofthedecisionmakingprocessforfinancialmarketpractitioners,wewilluseanapproach,developedbydecisionTheorist,HowardRaiffa.才Heusesthetermsnormativeanalysis,descriptiveanalysis,prescriptiveanalysis.Normativeanalysisisconcernedwiththerationalsolutiontotheproblemathand.Itdefinesanidealthatactualdecisionsshouldstrivetoapproximate.Descriptiveanalysisisconcernedwiththemannerinwhichrealpeopleactuallymakedecisions.Prescriptiveanalysisisconcernedwithpracticaladviceandtoolsthatmighthelppeopleachieveresultsmorecloselyapproximatingthoseofnormativeanalysis.Traditionalfinanceassumptionsaboutbehaviorasnormative.Behavioralfinanceexplanationsofbehaviorsasdescriptive.kEffortstousebehavioralfinanceinpracticeasprescriptive.CognitiveErrorsorEmotionalBiaseskBehavioralbiasescanbecategorizedascognitiveerrorsoremotionalbiases.Cognitiveerrorsstemfrombasicstatistical,information-processing,ormemoryerrors;cognitiveerrorsmaybeconsideredtoresultfromreasoningbasedonfaultythinking.Emotionalbiasesstemfromimpulseorintuition;emotionalbiasesmaybeconsideredtoresultfromreasoninginfluencedbyfeelings.kBehavioralbiases,cognitiveoremotional,maycausedecisionstodeviatefromtherationaldecisionsoftraditionalfinance.g三-3wO-To-£88UnCBqOnX一sPUtilityTheorykTomaximizeutility,arationalinvestorwillmakedecisionsconformingtothefouraxiomsofutility:completeness,transitivity,independence,andcontinuity.Completeness:Choicesandpreferencesareknown./Theindividualisawareofallavailablechoicesandcanvalueandassignpreferencestoeach,suchthatbetweenanytwochoices,theindividualprefersoneovertheotherorisindifferentbetweenthem.Transitivity:Rankingsareappliedconsistently./IftheinvestorpreferschoiceXtochoiceYandpreferschoiceYtochoiceZ,theinvestorwillpreferchoiceXtochoiceZ.UtilityTheoryIndependence:Utilitiesareadditiveanddivisible.AddingchoiceZtobothXandYwillnotaffectthepreferencerankingofXandY.If,forexample,theinvestorprefersXtoYandweaddZtobothchoices,theinvestorwillprefer(X+Z)over(Y+Z).AlSo,assumingtheinvestorprefersXtoY,ifanyportion,p,ofZisaddedtoXandY,theinvestor'spreferencerankingdoesnotchange.Theinvestorwillprefer(X+pZ)to(Y+pZ).Iftherankingisbasedonthesizeoftheproportion,thechoicesarenotindependent.kContinuity:Indifferencecurvesaresmoothandunbroken.Assumetherearethreechoices,L,M,andN,suchthattheinvestorprefersLtoMandMtoN.TheremustbeacombinationofLandN(portionsaandb)thatmakestheinvestorindifferentbetween(aL+bN)andM.Thisensuresthatindifferencecurvesareunbroken(i.e.,continuous).Bayes,FormulakTheFormulais:p(ab' P(B)*(/)where=ConditionaIprobabilityofeventAgivenB.Itistheupdated6<俯biityofAgiventhenewinformationB.=ConditionaIprobabilityofBgivenA.ItistheprobabilityofthenewiH(囿HbtionBgivenA.=prior(unconditional)probabilityofinformationB.（"）ateofbaseprobabilityofeventA.MakingDecisionsinAPerfectWord彳Inaperfectworld,whenpeoplemakedecisionsunderuncertainty,theyareassumedtodothefollowing:Adheretotheaxiomsofutilitytheory;Behaveinsuchawayastoassignaprobabilitymeasuretopossibleevents;IncorporatenewinformationbyconditioningprobabilitymeasuresaccordingtoBayes'formula;Chooseanactionthatmaximizestheutilityfunctionsubjecttobudgetconstraints(consistentlyacrossdifferentdecisionproblems)withrespecttothisconditionalprobabilitymeasure.kHomoeconomicsorrationaleconomicman(REM):Principlesofperfectrationality,perfectself-interest,andperfectinformationgovernREMseconomicdecisions.RiskAversekExpectedutilitytheorygenerallyassumesthatindividualsarerisk-averse.Thismeansthatanindividualmayrefuseafairwager,(awagerwithanexpectedvalueofzero),andalsoimpliesthathisutilityfunctionsareconcaveandshowdiminishingmarginalutilityofwealth;Giventwochoicesinvestingtoreceiveanexpectedvaluewithcertaintyorinvestinginanuncertainalternativethatgeneratesthesameexpectedvaluesomeonewhopreferstoinvesttoreceiveanexpectedvaluewithcertaintyratherthaninvestintheuncertainalternativethatgeneratesthesameexpectedvalueiscalledrisk-averse;kSomeonewhoisindifferentbetweenthetwoinvestmentsiscalledriskneutral;"Someonewhopreferstoinvestintheuncertainalternativeiscalledriskseeking;Intraditionalfinance,individualsareassumedtoberisk-averse.UtilityFunctionofWealthkAconcaveutilityfunctionmeansthatutilityincreasesatadecreasingratewithincreasesinwealth;therisk-averseindividualhasadiminishingmarginalutilityofwealth.Aconvexutilityfunctionmeansthatutilityincreasesatanincreasingratewithincreasesinwealth;therisk-seekingindividualhasanincreasingmarginalutilityofwealth.Thedegreeofriskaversioncanbemeasuredbythecurvatureoftheutilityfunction.ChallengestoTraditionalFinanceandtheREMkThosewhochallengeREMdosobyattackingthebasicassumptionsofperfectinformation,perfectrationality,andperfectself-interest.SomeofthebehavioralchallengestoREMinclude:Boundedrationalityassumesthatindividuals,choicesarerationalbutaresubjecttolimitationsofknowledgeandcognitivecapacity,anddonotbehavewithperfectrationality;Peoplehavedifficultyprioritizingshort-term(spending)goalsoverlong-term(spendingversussaving)goalsanddonotbehavewithperfectself-interest;Itisintuitivelyobviousthatmanyeconomicdecisionsaremadeintheabsenceofperfectinformationanddonotbehavewithperfectinformation(monetarypolicy);Wealthutilityfunctionsmaynotalwaysbeconcaveasassumedbyutilitytheory,andindividualscansometimesexhibitriskseekingbehavior.ChallengestoUtilityMaximizationkIndifferencecurveanalysismayincorporatebudgetlinesorconstraints,whichrepresentrestrictionsonconsumptionthatstemfromresourcescarcity;kIfthetwoitemsareperfectsubstitutes,thentheindividualiswillingtotradeonefortheotherinafixedratio;then,theindifferencecurveisalinewithaconstantslopereflectingthemarginalrateofsubstitution;kIfthetwoitemsareperfectcomplements,thenthecurvewouldbeL-shaped.Anadditionalamountofeithergoodaddsnoextrautilitybecausethegoodsareonlyusedincombination;kRiskplaysanimportantpartinmakingutility-maximizingdecisions.ChallengestoAttitudesTowardRiskkAssumingthatindividualsarerisk-averseandthatutilitycurvesareconcaveandexhibitdiminishingmarginalutilityseemsreasonable,butobservedbehaviorsarenotalwaysconsistentwiththeassumptionofanindividualwhoisconstantlyrisk-averse.Suchasbuyinglotteryticketsandbuyinginsurance,inwhichexpectedutilityislowbutpeople(evenwithlowincomes)participateinthepurchase;Ifaninvestmentoffersafewextremelylargeprizes,itsattractivenessisincreasedfarbeyondtheaggregatevalueoftheprizes./Thosewithlessincomeprefereithercertaintyorariskthatoffersasmallchanceofalargegaintoariskthatismoderate;/Middle-incomepeoplearemorelikelytobeattractedbysmallfairgambles.kRiskevaluationisreference-dependent,meaningriskevaluationdependsinpartonthewealthlevelandcircumstancesofthedecisionmaker.Itisnotnecessarilytruethatanindividual'sutilityfunctionhasthesamecurvatureconsistently:Ud0bi_iConvexRiskSeekingConcaveRiskAverseZ芸业创新增值Neuro-EconomicsbNeuroeconomicscombinesneuroscience,psychology,andeconomicsinattemptingtoexplainhowhumansmakeeconomicdecisions;"Neuro-economicsattemptstobridgethegapbetweenresearchondecisionbehaviorandeconomictheorybyunderstandingthebrainactivityOfjudgmentandmakingchoices;才Itistheamygdalathatcreatesafightorflightresponseduringasuddeneventortrauma.Forinvestors,theamygdalamayberesponsibleforapanickedresponseratherthanananalyticalresponsetoadroppingmarket;“Althoughneuro-economicsresearchisinterestingandmayprovidefurtherinsightsintoindividualeconomicdecisionmaking;itseffectoneconomictheoryremainstobeseen.DevelopmentoftheTheoryBernoulli:FranckKnightPascal:"expected"ValueOpriceRiskOmeasurableUtilityestimateUncertaintyOnotNeumann&MorgensternBoundedMaximizeexpectedutilityRationalitySavage:SubjectiveexpectedutilityDecisionTheory-TraditionalFinance-Decisiontheoryisconcernedwithidentifyingvalues,probabilities,andotheruncertaintiesrelevanttoagivendecisionandusingthatinformationtoarriveatatheoreticallyoptimaldecision.Thedevelopmentofthedecisiontheoryisasfollowing.Theinitialfocusofdecisiontheorywasonexpectedvalue;Expectedvalueofanitemisbasedonitsprice,whichisthesameforeveryonebecausethepricedependsonlyontheitemitself.Expectedutilityofanitemisbasedontheworthassignedtoitbythepersonmakingtheestimate;FrankKnightdefinesriskasrandomnesswithknowableprobabilitiesanduncertaintyasrandomnesswithunknowableprobabilities;ThetheoriesofvonNeumannandMorgensternandSavage(SEU)extendthescopeofexpectedutilitytheorytosituationsinwhichonlysubjectiveprobabilitiesareavailable.BoundedRationality才Boundedrationality:recognizingthatpeoplearenotfullyrationalwhenmakingdecisionsanddonotnecessarilyoptimizebutrathersatisfice(definedbelow)whenarrivingattheirdecisions;Satisficing("satisfy"&"suffice")isfindinganacceptablesolutionasopposedtooptimizing,whichisfindingthebest(optimal)solution;Theoptimalsolutionistheonethatmaximizestheutilityrealizablefromthesituation.”Decisionmakersmaychoosetosatisficeratherthanoptimizebecausethecostandtimeoffindingtheoptimalsolutioncanbeveryhigh;kWhenaspirationsarereached,peopletendtoadjusttheaspirationsupward;whenaspirationsarenotreached,peopletendtoadjustdownward;kDecisionsaremadeprogressivelyuntilthegoalstateisachieved:Thefirstdecisionismadetogetonestepclosertothegoalstate,thenextdecisionresultsingettingstillclosertothegoal,anddecisionscontinuetobemadeuntilthegoalstateismet.Anotherexampleisthedivide-and-conquerprocedure.ProspectTheory-BehavioralFinancekProspecttheoryhasbeenproposedasanalternativetoexpectedutilitytheory;kProspecttheoryassignsvaluetogainsandlosses(changesinwealth),ratherthantofinalwealth,andprobabilitiesarereplacedbydecisionweights;Thevaluefunctionisdefinedbythedeviationsfromareferencepointandisnormallyconcaveforgains(implyingriskaversion),convexforlosses(risk-seeking),andsteeperforlossesthanforgains(lossaversion).InvestorsareassumedtoPlaCeagreaterVdlUeOnaIossjtharLorLagainOftheSameamount;Givenapotentiallossandgainofequalsizes,theincreaseinIltilityassociatedWiththePOtentialgainisSmnHerthanthedecreaseinUtility(i.e.,disutility)associatedWiththePotentiallons;InvestorstendtofearIoSSeSandcanbecomeriskseeking(assumeriskierpositions)inanattempttoavoidthem.ValueFunctionReferencePointGainsValueProspectTheory-EditingProcessInprospecttheory,basedondescriptiveanalysisofhowchoicesaremade,therearetwophasestomakingachoice:anearlyphaseinwhichprospectsareframed(oredited)andasubsequentphaseinwhichprospectsareevaluatedandchosen.Dependingonthenumberofprospects,theremaybeuptosixoperationsintheeditingprocess:Codification:Peopleperceiveoutcomesasgainsandlossesratherthanfinalstatesofwealthorwelfare.Againorlossis,ofcourse,definedwithrespecttosomereferencepoint.Combination:Prospectsaresimplifiedbycombiningtheprobabilitiesassociatedwithidenticalgainsorlosses.Segregation:Therisklesscomponentofanyprospectisseparatedfromitsriskycomponent./Forexample,aprospectinitiallycodedas(300,0.8;200,0.2)isdecomposedintoasuregainof(200,1.0)andariskyprospectof(100,0.8;0,0.20).Thesameprocessisappliedforlosses.ProspectTheory-EditingProcessCancellation:Cancellationinvolvesdiscardingcommonoutcomeprobabilitypairsbetweenchoices./Forexample,thepairs(200,0.2;100,0.5;20,0.3)and(200,0.2;300,0.4;-50,0.4)arereducedto(100,0.5;20,0.3)and(300,0.4;-50,0.4).Simplification:Prospectsarelikelytoberoundedoff./Aprospectof(0.51,0.49)islikelytobeseenasanevenchancetowin50.Also,extremelyunlikelyoutcomesarelikelytobediscardedorassignedaprobabilityofzero.Detectionofdominance:Outcomesthatarestrictlydominatedarescannedandrejectedwithoutfurtherevaluation.¢=30smqpqo>口°dsalIp-SC一dW8ORPU2OdLp-.:ZX-XJLJM+(ZX)A(Nd)M+(IX)A(Id)MHn.ShortCommentkUtilitytheoryassumesriskaversion,prospecttheoryassumeslossaversion.Boundedrationalityrelaxestheassumptionsofperfectinformationandmaximizingexpectedutility.Prospecttheoryfurtherrelaxestheassumptionofriskaversionandinsteadproposeslossaversion.Prospecttheoryissuitedtoanalyzinginvestmentdecisionsandrisk.Itfocusesontheframingofdecisionsaseithergainsorlossesandweightinguncertai