心理學風險和投資

Psychology,Risk

andInvestment

DanielKahneman,PhD

PrincetonUniversity

Thestandardtheory

Therationalagentofeconomictheory:hasconsistentopinionsandbeliefs

usesallavailableinformationunbiasedbyemotion,‘herd’effectshascoherentpreferences

tangiblemotives(wealth,security)unaffectedby‘framing’ofproblems

BehavioralFinanceInputsfrompsychologyCognitiveerrorsEmotionalfactorsOuragendatodayBoldforecasts--errorsinjudgingtheoddsNarrowframingofdecisionsLossaversion--errorsinvaluingrisksFocusonindividualinvestors….ButexpertsarenotimmuneBoldForecasts—OptimismBiasRosyviewoflikelyoutcomes Becomingrichandfamous Becominganalcoholic HavingcancerExaggerationofskills 80-90%areabovemedian

Drivingskill SenseofhumorBoldForecasts—OptimismBias(continued)IllusionofcontrolExaggerateelementofskillDenytheroleofchanceOptimismaboutspecificchoicesWhydopeopleopenarestaurantwheremanyrestaurantshavefailed?OverconfidenceMakeaHIGHestimateoftheexchangerateon1/1/04 Youshouldbe99%sure

thatyourestimateistoohighMakeaLOWestimateoftheexchangerateon1/1/04 Youshouldbe99%sure

thatyourestimateistoolowOverconfidence(continued)Youshouldbe98%surethatthetruevaluewillfallinsideyourconfidenceintervalYoushouldhaveaprobabilityof2%thattheoutcomewillbeasurpriseatthe2%level

Overconfidence(continued)

FACT:massiveoverconfidence

Often20%surprisesatthe2%level 10-15%surpriseswhen“absolutelysure”

CAUSE:Limitedimagination Surprisesoccurinmanyunlikelyways

RESULT:underestimationofuncertaintyOptimisticOverconfidenceintheMarket

WhydoyouthinkYOUcanbeatthemarket?Thecostofhavingideas

(TerryOdean’sresearch) Whenaninvestorsellsastockand immediatelybuysanother:

thestockthatissolddoesbetterby3.5% inthefollowingyearFraming:differentwaystothinkaboutadecisionDifferentwaystothinkaboutcoldcuts:‘10%fat’or‘90%fat-free’Peopleframetheirowndecisions:somewaystothinkaboutadecisionproblemaremorenaturalthanothers

Framesvaryinbreadth:peopletendtoadoptframesthatareoverlynarrowFramingafinancialdecision:

Gains/lossesvs.wealthwouldyouacceptthisgamble?50%chancetowin$15,00050%chancetolose$10,000Nowmakeacrudeestimateofyourwealthwouldyouacceptthisgamble?50%chance:yourwealth+15K50%chance:yourwealth-10Kgain/lossframevswealthframeComparingtheframesWhichframeismorenatural?wenormallythinkintermsofgain/lossWhichframeismorereasonable?thebroaderviewEffectsoftheframesgain/lossframe---extremeriskaversionwealthframe---closertoriskneutralityTheaversiontolossesConsiderthisgamble50%tolose$10050%towin$XWhatXmakesthegambleacceptable?Acommonanswer:$200--$250Coefficientoflossaversionisabout2.5AnotherpairofchoicesWouldyouacceptthisgamble?50%tolose$100050%towin$1500mostpeoplerefuseWouldyouaccept10suchgambles?mostpeopleacceptIsthisyourlastriskydecision?Whichframeisbroader?morereasonable?Lossaversion&narrowframing:the‘‘dispositioneffect’Sellingstocks:Peopletendtohangontolosers,stocksthatarenowworthlessthantheirpurchaseprice.TheytendtosellwinnersBut…taxesAndwinnersdobetterintheshortrunThemoral:Havingdifferentattitudestowinnersandloserscostsmoney“Near-proportional”riskattitudesWhatisyourcashequivalentfor(100,.5)?Whatisyourcashequivalentfor(1,000,.5)?Howabout(10,000,.50)?TheCErisesalmostasfastasthestakesCoefficientoflossaversionalsostableWhydoesthisnotmakesense?lifeoffersmoresmallgambles,andthelawoflargenumbersreducesrelativeriskApplicationstounderstandingofmarketphenomenaTwomajorpuzzles:Theequity-premiumhasbeenabout7%Highvolumeoftrade--littleinformationTheanswers:“Myopiclossaversion”(BenartziandThaler)Overconfidenceoftraders(Odean)Exceptionstoriskaversion:LongshotsPeoplebuylotteryticketsChoosebetween50%towin5Kand50%towin15K95%towin9Kand5%towin29KExceptionstoriskaversion:RiskseekinginlossesChoosebetween90%tolose$2,000lose$1,800forsureThisisachoicebetweenasurelossandahighprobabilityofanevenlargerloss,withasmallchancetostayevenPeoplechoosetogambleThecertaintyofalossmakesitmoreaversiveRisk-seekingdecisionsAcceptingdefeat,orfightingonSettlingabadcase,orlitigatingabadcase?EscalatingcommitmentThedifficultyof“cuttinglosses”VerycommoninbadinvestmentsThe““agencyproblem”incommitmentPerseveranceinthefaceofadversityDecisionsofexecutiveswithnothingtoloseBehavioralTheory:TheRealInvestorNarrowframingexacerbatestwobiasesOptimisticbias–boldforecastsRiskaversion–timiddecisions(KahnemanandLovallo,1993)MostofwhathasbeensaidappliestoexpertsHighlyinaccurateandoverconfidentLossaverseundersupervisionRisk-seekingtoavoidsurelossesIndividualsvs.OrganizationsOrganizationsareliabletooverconfidenceNotimmunetoriskerrorsButOrganizationshavebroaderframesOrganizationshavebroaderattentionspanNotafaircontest…Regret:Thenextresearchfrontier?DeterminantsofregretHindsightComingcloseCommission>omissionEffectsofregretAnticipatedregret--conservatismUnanticipatedregret--notstayingthecourseHindsightTheinevitabilityofthepastMondaymorningquarterbacksEvening-afterDow-JonesgeniusesDistortingpastodds"Itwasalwaysobvious"Distortingyourownpastbeliefs"Ialwaysknewit"isnobetterDenyingtheuncertaintyofthepastImplicationsforuncertaintyoffutureThePainsofaCloseMissMissingaflightMr.TeesandMr.Cranewerescheduledtoleavetheairportondifferentflights,atthesametime.Theytraveledtogetherfromtowninthesamelimousine,werecaughtinatrafficjam,andarrivedattheairport30minutesafterthescheduleddeparturetimeoftheirflights.ThePainsofaCloseMiss(continued)Mr.CraneistoldthathisflightleftontimeMr.Teesistoldthathisflightwasdelayed,andjustleftfiveminutesagoWhoofthemismoreupset?------------------------