人工智能財(cái)務(wù)金融應(yīng)用

AIinFinancialApplication1財(cái)務(wù)金融事件研究法

(EventStudiesinFinance)EventStudies

inFinance2DoronKligerandGregoryGurevich(2014),

EventStudiesforFinancialResearch:

AComprehensiveGuide,

PalgraveMacmillan3Source:/Event-Studies-Financial-Research-Comprehensive/dp/1137435380/事件研究法：

財(cái)務(wù)與會(huì)計(jì)實(shí)證研究必備

沈中華、李建然(2000)4Source:沈中華、李建然(2000)，事件研究法：財(cái)務(wù)與會(huì)計(jì)實(shí)證研究必備，華泰文化EventStudies

for

FinancialResearch56/EventStudies

inEconomicsandFinance7Source:MacKinlay,A.C.(1997).Eventstudiesineconomicsandfinance.Journalofeconomicliterature,35(1),13-39.GoodNewsNoNewsBadNewsEventStudy8Source:RajeshMudholkar(2014),"Eventstudies:ConfirmsMarketEfficiencyorBehavioralAnomalies?",/watch?v=VErwDaQNB74EventStudy

Timelineforaneventstudy9Source:/index.php/event-study-methodology/EventEventStudyMethodology10Source:/index.php/event-study-methodology/0T2T3T1T0event

windowpostevent

windowestimation

windowL1L2EventStudyMethodology11Source:/index.php/event-study-methodology/0T2T3T1T0event

windowpostevent

windowestimation

windowL1L2-5-4-3-2-10+1+2+3+4+5-30+30EfficientMarkets12BehavioralEconomics13BehavioralFinance14HershShefrin(2007),

BeyondGreedandFear:

UnderstandingBehavioralFinanceandthePsychologyofInvesting,OxfordUniversityPress15Source:/Beyond-Greed-Fear-Understanding-Association/dp/0195304217AndreiShleifer(2000),

InefficientMarkets:AnIntroductiontoBehavioralFinance,

OxfordUniversityPress16Source:/Inefficient-Markets-Introduction-Behavioral-Clarendon/dp/0198292279LucyAckertandRichardDeaves(2009),

BehavioralFinance:Psychology,Decision-Making,andMarkets,

South-WesternCollegePub17Source:/Behavioral-Finance-Psychology-Decision-Making-Markets/dp/0324661177EdwinBurtonandSunitN.Shah(2013)

BehavioralFinance:UnderstandingtheSocial,Cognitive,andEconomicDebates,Wiley18Source:/Behavioral-Finance-Understanding-Cognitive-Economic/dp/111830019XRationalBehavior

IrrationalBehavior19Emotion

Sentiment20ModernFinancialResearchTheoreticalFinancestudyoflogicalrelationshipsamongassets.EmpiricalFinancestudyofdatainordertoinferrelationships.BehavioralFinanceintegratespsychologyintotheinvestmentprocess.21Source:RobertA.Strong(2004),PracticalInvestmentManagement,South-WesternBehavioralFinanceThemesHeuristic-DrivenBiasFramingDependenceInefficientMarkets22Source:HershShefrin(2007),“BeyondGreedandFear:UnderstandingBehavioralFinanceandthePsychologyofInvesting”,OxfordUniversityPress.EfficientMarketHypothesis

(EMH)23Source:DoronKligerandGregoryGurevich(2014),EventStudiesforFinancialResearch:AComprehensiveGuide,PalgraveMacmillanEfficientMarketHypothesis(EMH)(Fama,1970)24Efficientcapitalmarkets:AreviewoftheoryandempiricalworkBGMalkiel,EFFama

-TheJournalofFinance,1970-WileyOnlineLibraryThispaperreviewsthetheoreticalandempiricaliteratureontheefficientmarketsmodel.

Afteradiscussionofthetheory,empiricalworkconcernedwiththeadjustmentofsecurity

pricestothreerelevantinformationsubsetsisconsidered.First,weakformtests,inwhichthe

informationsetisjusthistoricalprices,arediscussed.Thensemi-strongformtests,inwhich

theconcerniswhetherpricesefficientlyadjusttootherinformationthatisobviously...

Citedby20928

Relatedarticles

All28versionsMalkiel,B.G.,&Fama,E.F.(1970).Efficientcapitalmarkets:Areviewoftheoryandempiricalwork.TheJournalofFinance,25(2),383-417.25EfficientMarketHypothesis(EMH)(Fama,1970)Malkiel,B.G.,&Fama,E.F.(1970).Efficientcapitalmarkets:Areviewoftheoryandempiricalwork.TheJournalofFinance,25(2),383-417.EfficientMarketHypothesis(EMH)(Fama,1970)26Source:Malkiel,B.G.,&Fama,E.F.(1970).Efficientcapitalmarkets:Areviewoftheoryandempiricalwork.TheJournalofFinance,25(2),383-417.27Source:Malkiel,B.G.,&Fama,E.F.(1970).Efficientcapitalmarkets:Areviewoftheoryandempiricalwork.TheJournalofFinance,25(2),383-417.CumulativeAverageResiduals28Source:Malkiel,B.G.,&Fama,E.F.(1970).Efficientcapitalmarkets:Areviewoftheoryandempiricalwork.TheJournalofFinance,25(2),383-417.29Source:Malkiel,B.G.,&Fama,E.F.(1970).Efficientcapitalmarkets:Areviewoftheoryandempiricalwork.TheJournalofFinance,25(2),383-417.CumulativeAverageResidualsMarketEfficiency30Source:Malkiel,B.G.,&Fama,E.F.(1970).Efficientcapitalmarkets:Areviewoftheoryandempiricalwork.TheJournalofFinance,25(2),383-417.TypesofEfficiencyMarketWeakFormSecuritypricesreflectallinformationfoundin

pastpricesandvolume.Semi-StrongFormSecuritypricesreflectallpubliclyavailable

information.StrongFormSecuritypricesreflectallinformation—publicandprivate.31Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillCanFinancingDecisions

CreateValue?32Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillWhatSortofFinancingDecisions?Typicalfinancingdecisionsinclude:HowmuchdebtandequitytosellWhen(orif)topaydividendsWhentoselldebtandequityJustaswecanuseNPVcriteriatoevaluateinvestmentdecisions,wecanuseNPVtoevaluatefinancingdecisions.33Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillHowtoCreateValuethroughFinancingFoolInvestorsEmpiricalevidencesuggeststhatitishardtofoolinvestorsconsistently.ReduceCostsorIncreaseSubsidiesCertainformsoffinancinghavetaxadvantagesorcarryothersubsidies.CreateaNewSecuritySometimesafirmcanfindapreviously-unsatisfiedclienteleandissuenewsecuritiesatfavorableprices.Inthelong-run,thisvaluecreationisrelativelysmall,however.34Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillEfficientCapitalMarketsAnefficientcapitalmarketisoneinwhichstockpricesfullyreflectavailableinformation.TheEMHhasimplicationsforinvestorsandfirms.Sinceinformationisreflectedinsecuritypricesquickly,knowinginformationwhenitisreleaseddoesaninvestornogood.Firmsshouldexpecttoreceivethefairvalueforsecuritiesthattheysell.Firmscannotprofitfromfoolinginvestorsinanefficientmarket.35Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillReactionofStockPricetoNewInformationinEfficientandInefficientMarkets36StockPrice-30 -20 -10 0 +10 +20 +30Daysbefore(-)andafter(+)announcementEfficientmarketresponseto“goodnews”O(jiān)verreactionto“goodnews”withreversionDelayedresponseto“goodnews”Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillReactionofStockPricetoNewInformationinEfficientandInefficientMarkets37StockPrice-30 -20 -10 0 +10 +20 +30Daysbefore(-)andafter(+)announcementEfficientmarketresponseto“badnews”O(jiān)verreactionto“badnews”withreversionDelayedresponseto“badnews”Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillVersionsofEMH/Info-EfficiencyWeak-formefficiency:PricesreflectallinformationcontainedinpastpricesSemi-strong-formefficiency:PricesreflectallpubliclyavailableinformationStrong-formefficiency:Pricesreflectallrelevantinformation,includeprivate(insider)information38allpublic&privateinfoallpublicinfopastmarketinfoSource:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityRelationshipamong

ThreeDifferentInformationSets39Allinformation

relevanttoastockInformationset

ofpubliclyavailable

informationInformation

setof

pastpricesSource:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillEfficientMarketAnefficientmarketincorporatesinformationinsecurityprices.TherearethreeformsoftheEMH:Weak-FormEMHSecuritypricesreflectpastpricedata.Semistrong-FormEMHSecuritypricesreflectpubliclyavailableinformation.Strong-FormEMHSecuritypricesreflectallinformation.ThereisabundantevidenceforthefirsttwoformsoftheEMH.40Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillWhyTechnicalAnalysisFails41StockPriceTimeInvestorbehaviortendstoeliminateanyprofitopportunityassociatedwithstockpricepatterns.Ifitwerepossibletomakebigmoneysimplybyfinding“thepattern”inthestockpricemovements,everyonewoulddoitandtheprofitswouldbecompetedaway.SellSellBuyBuySource:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillEvidenceonMarketEfficiencyReturnPredictabilityStudiesEventStudiesPerformanceStudies42Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudiesObjectiveExamineifnew(companyspecific)informationisincorporatedintothe

stockpriceinonesinglepricejumpuponpublicrelease?43Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudiesMethodologyDefineasday“zero”thedaytheinformationisreleasedCalculatethedailyreturnsRitthe30daysaroundday“zero”:

t=-30,-29,…-1,0,1,…,29,30CalculatethedailyreturnsRmtforthesamedaysonthemarket(oracomparisongroupoffirmsofsimilarindustryandrisk)DefineAbnormalReturns(AR)asthedifferenceCalculateAverageAbnormalReturns(AAR)overallNeventsinthesampleforall60referencedaysCumulatethereturnsonthefirstTdaystoCAAR44ARit=Rit

–Rmt

Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudiesMethodologyStep1.

Defineasday“zero”thedaytheinformationisreleased45Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudiesMethodologyStep2.

CalculatethedailyreturnsRit

the30daysaroundday“zero”:

t=-30,-29,…-1,0,1,…,29,3046Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudiesMethodologyStep3.Calculatethedailyreturns

Rmtforthesamedaysonthemarket(oracomparisongroupoffirmsofsimilarindustryandrisk)47Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudiesMethodologyStep4.

Define

AbnormalReturns(AR)

asthedifference48ARit=Rit

–Rmt

Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudiesMethodologyStep5.

Calculate

AverageAbnormalReturns(AAR)

overallNeventsinthesamplefor

all60referencedays49Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudiesMethodologyStep6.

Cumulatethereturnsonthe

firstTdaysto

CumulativeAverageAbnormalReturns(CAAR)50Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudiesMethodologyDefineasday“zero”thedaytheinformationisreleasedCalculatethedailyreturnsRitthe30daysaroundday“zero”:

t=-30,-29,…-1,0,1,…,29,30CalculatethedailyreturnsRmtforthesamedaysonthemarket(oracomparisongroupoffirmsofsimilarindustryandrisk)DefineAbnormalReturns(AR)asthedifferenceCalculateAverageAbnormalReturns(AAR)overallNeventsinthesampleforall60referencedaysCumulatethereturnsonthefirstTdaystoCAAR51ARit=Rit

–Rmt

Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityMarketEfficiencyinEventStudies52-30-25-20-15-10-5051015202530Over-reactionEfficientReactionUnder-reactionTSource:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudy:

EarningAnnouncements53EventStudybyBallandBrown(1968)Pre-announcementdriftpriorto

earningsduetoinsidertrading

!againststrong-formPost-announcementdrift

!againstsemi-strongformSource:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudy:

EarningAnnouncement54Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudy:StockSplits55SelectionbiasorInsidertradingEventStudyonStockSplitsbyFama-French-Fischer-Jensen-Roll(1969)SplitisasignalofgoodprofitPre-announcementdriftcanbedue

toselectionbias(onlygoodfirms

split)orinsidertrading.

!inconclusiveNopost-announcementdrift

!forweakformSource:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudy:Take-over56Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEventStudy:DeathofCEO57Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityEvidenceI:PredictabilitiesStudiesStatisticalvariableshaveonlylowforecastingpower,butButsomeforecastingpowerforP/EorB/MShort-runmomentumandlong-runreversalsCalendarspecificabnormalreturnsduetoMonayeffect,Januaryeffectetc.CAVEAT:Datamining:Findvariableswithspuriousforecastingpowerifwesearchenough58Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityLong-RunReversalsLong-runReversalsReturnstoprevious5year’swinner-loserstocks(marketadjustedreturns)59Source:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityShort-runMomentum60MomentumMonthlyDifferenceBetweenWinnerandLoserPortfoliosatAnnouncementDates1357911131517192123252729313335MonthsFollowing6MonthPerformancePeriodMonthlyDifferenceBetweenWinnerand-1.5%-1.0%-0.5%0.0%0.5%1.0%LoserPortfoliosSource:MarkusK.Brunnermeier(2015),“Lecture10:MarketEfficiency”,Finance501:AssetPricing,PrincetonUniversityGettingTechnical

Barron’sMarch5,200361Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillGettingTechnical

BacktoBuyLow,SellHigh Barron’sMarch12,200362Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillWhatPatternDoYouSee?63Withdifferentpatterns,youmaybelievethatyoucanpredictthenextvalueintheseries—eventhoughyouknowitisrandom.Source:Rossetal.(2005),"CorporateFinance",7thEdition,McGraw?HillEventStudies:DividendOmissions64Efficientmarketresponseto“badnews”S.H.Szewczyk,G.P.Tsetsekos,andZ.Santout“DoDividendOmissionsSignalFuture