2010金融專題課件及相關(guān)材料ms3mkt anomalies and bf final

MSClassBehavioral 用小錢試運(yùn)會(huì)讓人絕望。樂透 樂透迷想中大獎(jiǎng)，應(yīng)簽熱門號(hào)碼簽冷門號(hào)Don’tjust yzethemarket, theinvestor.Stockpatternsemergefromcommonhumanbehavior.BoundedrationalityindecisionWeneedtoconsiderthenon-rationalbehaviorofotherswhenwemakefinancialandmanagerialdecisions.Peopleareoverconfidentaboutprecisionoftheirinformationtheirabilities.

andTheTraditionalMarketEfficiencyvsReconcilingTheoryandAnAlternativeBehavioralWhatisBehavioralCriticsofBehavioralUnifyingBehavioralandRationalI.MutualFundalReduceBiasesin/Decision明 -行為財(cái)務(wù)（反向投資策略或動(dòng)能交易基金等（如何避免過度自信或事先建立賣股的投資策略等（過度自信或過度反應(yīng)等Fuller&ThalerAssetManagement,Inc.isapioneerinbehavioralFoundedin1993byRussellFuller,Fuller&Thalercurrentlymanagesmorethan$450mmforbothpension nsandhighnet ThisFundusesabehavioralfinanceinvestmentstrategythattriestoidentifymispricedsecuritiesthatresultfrombehavioralbiasesofkeymarketparticipants.DesignandMarketingofFinancialStandardFinance:Cateringto DifferencesinpreferencesandBehavioralFinance:ProspecttheoryHedonicframingBehaviorallifecycletheoryCognitiveerrors例子、動(dòng)能策利用過去 來預(yù)測(cè)未 利用市場(chǎng)上投資人對(duì)和盈余信息react ，解釋股價(jià)動(dòng)能的現(xiàn)象理財(cái)咨al投資咨詢：知己知來之變異程度。盡管心理反應(yīng)對(duì)投資績(jī)效舉足輕重，然而將風(fēng)險(xiǎn)以情 一知半解 剛公布非常好的盈余好的利多消息，股價(jià)但公司盈余增加的原能再重投資咨詢：投資自設(shè)立P(Targetprice)SScenario)，如股價(jià)漲超過原先買入價(jià)50%超過50%（T）Cutyourloss,letyourwinners由于投資人無法當(dāng)機(jī)立斷賣出，用策略來規(guī)范整體投資投資哲凱因斯說正確投資觀、耐心與多元投MarketStage1:byStage2:byStage3:afterstartoffwiththestandardfinance:marketefficiencyandtheriskPricingpuzzlesandrationalizingobservedreturnpatternsledtoachangeoffocusfromneoclassicaltowardbehavioralThePre-History:StatisticalWhethersecuritypricesareseriallycorrelated(i.e.,aretheretrendsinstockprices)orfollowa"randomwalk,"changingtoreflectnewinformation.AnumberofstudiesconcludedthatsuccessivechangesinstockpricesaremostlynopatternthatcouldpredictthefuturedirectionofpriceL.WayneGertmenianandNikolaiTheOriginoftheEfficientMarketIn1959Robertsplottedtheresultsofaseriesofrandomly-generatednumberstoseewhetheranypatternsthatwereknowntotechnicalystswouldbeRobertsnotedthatitwasvirtuallyimpossibletolifhisplotsweregeneratedusingrandomnumbersoractualstockmarketSimulatedstockpriceTestsofMarketEfficiencyintheAnumberofdifferentapproacheswereusedtotesttheefficientmarkethypothesis.Onewastodomorestudiesonserialcorrelationofsecurityprices. Avariationofthisapproachtestedvarioustrading mendedbytechnical yststoseeiftheyhaveanyinvestmentvalue.Bothinvariablycamebackwithmostlynegativeresults."eventstudy"methodology:takeasampleofsimilareventsthatoccurredindifferentcompaniesatdifferenttimesanddeterminehow,onaverage,thiseventimpactedthestockprice.(firstappliedtostocksplit).TestsofMarketEfficiencyafterRozeffandKinneyfoundthatJanuarystockreturnswerehigherthaninanyothermonth.GibbonsandHessreported"theMonday-stockpricestendedtogodownonBothofthesefindingswereclearlyinconsistentwiththeweakformofmarketefficiency.GibbonsandHessnoticedthattheMondayeffectseemedtodecreaseovertime.(theeffectwasknowntosomemarketGrossman-StiglitzGrossmanandStiglitzarguedthatifallrelevantinformationwerereflectedinmarketprices,marketagentswouldhavenoincentivetoacquiretheinformationonwhichpricesarebased.ThislineofreasoningcametobeknownasGrossman-Stiglitzparadox.Alongwithhisothercontributions,earnedJosephStiglitzhisNobelprizein2001.)Sincetheearly1980s,therehasbeenamovementtowardincorporatingmorebehavioralscienceintofinance.ExcessShiller:priceschangetoomuchtobejustifiedbysubsequentchangesindividends.ApuzzleoftradingIfeveryoneknowsthateveryone(includinghimselforherself)isrational,theneverytradermightwonderwhatinformationthesellerhasthatthebuyerdoesn't,andviceDividendInaperfectworldaccordingtoModiglianiandMiller,investorsshouldbeindifferentbetweendividendsandcapitalgains.Intherealworld,becauseofthestructureoftheU.S.taxsystem,rationalinvestorsshouldprefercapitalgainstodividends,andcompaniesshouldprefersharerepurchasestodividends.Mostlargecompaniesdopaydividends.Inaddition,stockpricestendtorisewhendividendsareincreasedorHistorically,thisbenefithasbeenmuchgreaterthancanbeex inedbyriskalone.Finally,itseemsthatfuturereturnscan,atleastpartially,bepredictedas:price-earningsandprice-to-bookratios,earningssurprises,dividendchanges,orshareReal-worldportfoliomanagersarestillhavingahardtimetryingtobeatthemarket.Goodperformancethisyearconsistentlyfailstopredictgoodperformancenextyear.MarketTheevidencecontrarytotheCalendarpatterns-JanuarySmallfirmP/E(P/B)Cross-sectionalreturnAtparticularmomentsoftimesuchasthebeginningoftheweekortheturnofthemonth,stockreturnsreachabnormallevelsrecurrently.Forinstance,thetendencyofcommonstockreturnsforearlyJanuarytobeaboveaveragedoesnotoccurmerelybychance.Itseemsthatbothinstitutionalfactors,amongwhichtheend-of-yearcashinfusionsintothemarket(Ogden,1990),andbehavioralconsiderationsaffecttradingbyindividualinvestors(cf.Sias&Starks,1997).JanuaryIthasbeenfoundthatstocksseemtodoconsistentlybetterinJanuarythanothermonths.onlyforsmallTax-losssellingbyinvestorsattheendofDecember.Individualinvestorssellpoor-performingstockstorealizelossesfortaxpurposesBuyingpressureinearlyJanuaryresultsinadisproportionatesmall-firmpremium(see,e.g.,Roll,1983).ThisisnottosaythatinvestorsshouldwaituntiltheendoftheBut,rather,thatindividualinvestorstypicallyhaveadispositiontoholdontopoor-performingstocks(Ferris,Haugen,&Makhija,1988.Ritter(1988):individualinvestorsdonot yreinvestalltheproceedsfromearliersalesinHefindsthattheratioofstockpurchasestosalesbyindividualinvestorsdis ysaseasonalpattern,withindividualshavingabelow-normalbuy/sellratioinlateDecember,whichabruptlyswitchestoanabove-normalratioinearlyJanuary.Forthatmatter,whenindividualsreinvesttheirfunds,theytypicallybuylow-pricedandlow-marketcapitalizationstocks,justastheydothroughouttheyear.Inducedbyperiodicevaluation,professionalmanagersmostlikelyrebalancetheirportfoliospriortoyear-endtoremoveembarrassment-causinglosers,whichoftenaresmall,riskystocks(cf.Haugen&Lakonishok,1988;Ritter&Chopra,1989).Then,attheturnoftheyear,ashiftofportfolioallocationstomoreaggressivestocksgeneratespricepressure,irrespectiveofwhetherthemarketisuporSmallfirmSmallfirmsseemtohavebeenabletoconsistentlyoutperformlargefirms,evenadjustedforrisk,from1926untilrecenttimes.Widelyknowninthetherewasanimmediateproliferationofmoneymanagersandmutualfundsspecializinginsmallfirms.Result:Returnsonsmallfirmshavegenerallylaggedbigfirmsoverthelast10years-surprise!P/E&P/BLowP/EstockstendtodoslightlybetterthanthemarketinsubsequentperiodsandhighP/Estockstendtodoslightlyworse.Stockswithlowratiosofmarketpricepersharetobookvaluepershare(P/B)tendtodobetterinsubsequentperiodsthanstockswithhighmarket/bookratios.LowP/E(P/B)stocksgotthatwaybecausetheoverreactedtosomenegativeHighP/E(P/B)stocksgotthatwaybecausethemarketoverreactedtopositiveinformationand,thus,insubsequentperiodscomesbacktonormal.OverreactionandDeBondtandThaler:thestockmarkettendstooverreacttolongseriesofbadnews.PersistenceofstrongormerelyweakreturnsisindicativeofaslowtoAnunder-reactionatthemomentofitsrelease,whereasareturnreversaloracontrarymovementsuggestsanoverreactionthatissubsequentlyTheonesassociatedwithidentifiablenewsevents. Theinformationcontentofthenewscanbeinferredandalsothespeedofprocessing,thatishowmuchtimepricestaketoreflectfullyandcomple ythearrivedinformationifthereisundershootingorovershooting.#AnannouncementofacorporateAnannouncementofthefirmconcerning,e.g.,itsperformancesuchasthequarterlyearnings(e.g.,Bernard&Thomas,1990)oranintendedcourseofaction,suchasaseasonedequityoffering(e.g.,Loughran&Ritter,1997,II:11),anopenmarketsharerepurchase(e.g.,Ikenberryetal.,1995,II:12)astocksplit(e.g.,Ikenberryetal.,1996,II:13),andaninitiationoromissionofcashdividendpayments(e.g.,Michaelyetal.,1995,II:14).Apredictablepost-announcementlong-runreturnsdriftinthesamedirectionastheinitialreturnsresponseisfound,whichsuggeststhatthefulleconomicimpactofnewpubliccorporatefactsislargelyignoredorunder-weightedbyinvestorsintheshortrun.Thiseffectseemstobestrongespeciallyforout-of-favorstocks,whichtendtohavehighbook-to-marketratios.Cross-sectionalreturnParticularstocksearnpredictablehugeabnormalreturnsonaregularbasisPricesseemtobeinefficientinthesensethattheynotfullyincorporatepastinformationAprofitableinvestmentstrategycanbeAccordingtotheory,riskfactorsaloneex incross-sectionaldifferencesin(average)stockreturns,butherethestoryisthatvariablessuchaspriorreturnperformance,book-to-marketvalueandmarketcapitalizationdonot(sufficiently)pickupvariationinrisk.usibleAfactormodelimpliesa positionofrisk.Anexampleofafactoristhemarket(return),whichcaptureseconomy-wideinfluences.Unexpectedchangeshaveanimpactonallreturns,beingstrongforsomestocksandonlyweakforsomeotherones,asexpressedbythemarketbeta.Empiricalresearchshowsthatmostfactorssufferfrommeasurementerrorandlongpublicationlags,andthattheestimatedfactorbetasoftenarenon-stationaryApplication:ActiveandPassiveWhetherornotyoubelieveintheEMHinfluenceshowyouviewinvestingandactiveandpassivemanagement.CanwebeprofitfromtradingActiveandPassiveWhethertheinvestorshouldfollowabuyandholdstrategyorattempttotimethemarketbyusingsomesortofeconomic,fundamental,ortechnical Whethertheinvestorshouldattempttoearnexcessreturnsbyactivelymanagingthesecurityselectionprocess.ActiveInvolvesperiodicchangestoaportfolioinanticipationpriceAnactivelymanagedstockportfoliomightswitchitsholdingsfromgrowthstockstowardcyclicalstocksinanticipationofaneconomicexpansion.Suchaproceduregeneratesadditionalcosts,intheformofcommissions,inordertochangefromonetypeofstocktoanother.Themanagersofactivelymanagedportfoliosusuallyrelyontechnical ysis,fundamental ysis,orsomecombinationtoguidetheirportfoliochanges.Marketefficiencycastsdoubtonthepossibilitythatactivelymanagedportfolioscanconsistentlyproducesuperiorreturnsoverapassivelymanagedone.EMHholds:ActivemanagementmustprofitenoughtopaythehighertransactionscostsassociatedwithmorefrequentbuyingandIndexedAnalternativetoactiveDesignedtomimictheperformanceofsometargetindex,suchastheS&P500,Russell2,000ortheMerrillLynchCorporateMasterBondIndex.ImplicitacknowledgmentofthetruthoftheInessence,activemanagementwon'tworkoristooexpensiveorrisky,andthatthesmartthingtodoistochoosetheno-brainerapproachofAnExercise:EarningsSurprise#EarningsRendelman,Jones,andLatané(1982)studiedearningssurprisesandtheireffectonthestockTheydividedtheirsampleintotengroups(deciles)accordingtohowpositiveornegativetheearningssurprisewas.Thentheycalculatedaveragedpricepathsforstocksineachdecile.Figure2presentsasummaryoftheirfindings.VI.ASummaryof'Nobel'PrizeinEconomicswinnerDanielKahneman,WhatisbehavioralItpostulatesthatinvestorshavecognitivebiases--imperfectionsintheirperceptionsofreality.Mentalaccounting.Dividendsareperceivedasanadditiontodisposable e;capitalgainsusuallyareBiasedexpectations.Peopletendtobeoverconfidentintheirpredictionsofthefuture..Between1973and1990,earningsforecasterrorsbysecurity ystshavebeenanywherebetween25%and65%ofactualReferenceIfacertainstockwasoncetradingfor$20,thendroppedto$5andfinallyrecoveredto$10,theinvestor'spropensitytoincreaseholdingsofthisstockwilldependonwhetherthepreviouspurchasewasmadeat$20or$5.RepresentativenessInvestorsmistakegoodcompaniesforgoodstocks.Goodcompaniesarewell-knownandinmostcasesfairlyvalued.Theirstocks,therefore,maynothaveasignificantupsidepotential.Peopleoftenmakedecisionsbasedonapproximaterulesofthumb,notstrictlyrational yses.SeealsocognitivebiasesandboundedThewayaproblemordecisionispresentedtothedecisionmakerwillaffecttheiraction.MarketTheseincludemispricings,non-rationaldecisionmaking,andreturnanomalies.(RichardThaler,inparticular,haswrittenalongseriesofpapersdescribingspecificmarketanomaliesfroma BehavioralfinanceLossaversionappearstomanifestitselfininvestorbehaviorasanunwillingnesstosellsharesorotherequity,ifngsowouldthetradertorealizeanominalloss(Genesove&Mayer,2001).Itmayalsohelpexinwhyhousingmarketpricesdonotadjustdownwardstomarketclearinglevelsduringperiodsoflowdemand.Applyingaversionofprospecttheory,BenartziandThaler(1995)claimtohavesolvedtheequitypremiumWhyisBehavioralFinanceDoinvestmentshavevalue-expressiveIftheydo,weshouldnotbesurprisedthatpricingdifferencesexistbetweenotherwiseidenticalAcasuallookatadvertisingandtheextentofbrandingeffortsinthemutualfundsindustrysuggestsapositiveIsaCompromiseinArethedifferencesbetweentraditionalfinancebehavioralfinanceNotnecessarily.Onetheonehand,sensibleproponentsofbehavioralfinancerecognizeitslimitations.Ontheotherhand,advocatesoftraditionalfinanceresearcharebeginningtostudytheevidenceofvariousbehavioraleffects.(See,forexample,Chen,HongandSteinfortreatmentofopinionItisimportanttorememberthatwithoutclearindicationofanequityportfoliomanager'sabilitytobeatthemarket,theinvestormaybebetteroffwithapassivestrategy.Itisimperativeforindividualinvestorstokeepinmindthattheythemselvesmaybesubjecttovariouscognitivebiases.Criticsofbehavioralfinance:EugeneBehavioralfinanceismoreacollectionofanomaliesthanatruebranchoffinanceandthattheseanomalieswilleventuallybepricedoutofthemarketorex inedbyappealtomarketmicrostructurearguments.However,adistinctionshouldbenotedbetweenindividualbiasesandsocialbiases;theformercanbeaveragedoutbythemarket,whiletheothercancreatefeedbackloopsthatdrivethemarketfurtherandfurtherfromtheequilibriumofthe"fairprice".CriticismsfromSteveTheyareSmall$(e.g.,MCIJr.vs.Notscalable,e.g.,VolatilityFleeting(稍縱即逝E.g.,thesmallstockpremium,seeSchwertNonprofitBid/AskInformationcosts,e.g.,complexmortgageNeoclassicalvs.NoarbitrageandefficiencyproducethePsychologyproducestoomanyanswersandnoArepeopleoptimistsorpessimists–theyareNeoclassicaltheorypredictsmagnitudeaswellassignsofeffectsAesthetics;IliketheorieswithsomedistancebetweenassumptionsandconclusionsYouwantcorrelationsthenjustmakeindividualbehaviorcorrelatedPr