CFA三级基础：资本市场预期_打印版.docx

CFA：级培训项HTopic in CFA Level IllETHICS & PROFESSIONAL STANDARDS &：2Study Session 3BEHAVIORAL FINANCEStudySession1-2StudySession4CAPITALMARKETEXPECTATIONSStudy Session 5Study Session 6Study Session 78Study Session 910Study Session UStudy Session 12-13Study Session 14Study Session 15Study Session 16ASSETALL<XATIONANDRELATEDDECISIONSINPoRTFgL1。MANAGEMENTDER1VA11VESANDCURRENCYMANAGEMENTFIXED-INCOMEPORTFOUOMANAGEMENTI&2)EQUITYPORTFOUOMANAGEMENT1：&2:ALTERNATIVEINVESTMENTSFORPORTFOUOMANAGEMENTPRIVATEWEALTHMANAGEMENTU&2；PORTFOLIOMANAGEMENTFQR1NST11TK)NALINVESTORSTRADINGPERFORMANCEEVALUATIONANDMANAGERSELECTIONCASES!NPORTFOUOMANAGEMENTANORISKMANAGEMENTFrameworkApplicationsofCapitalMarketExpectationsSS4:CapitalMarketExpectationsRlOCapitalMarketEx>e<tations1Part1:FrameworkandMacroConsiderationsRllCapitalMarketExpectations.Part4-1152:ForecastingAssetClassReturnsSdBM.EtaCapitalMarketExpectations,Part1:FrameworkandMacroConsiderations11«ZiflFramework1.ChallengeswhenformulatingCME2.ThetrendIateofgrowth3.Approachestoeconomicforecasting4.Thebusinesscycle5.Monetarypolicyandfis<alpolicyFrameworkandChallengesCapitalmarketexpectationsareskandreturnexpectationsregardingclassesofassets.i【-t：long-termassetclassspecifiedtheinvestmentpolicystatement(IPS)*Althoughprojectingassetclassreturnsmaybesubjecttoforecastingeos.IiwestaisshouldensureIhEPoNoliQSareinternallyconsistent/Cross-sectionalconsistencyreferstoconsistent,acrossassetclassesregardingportfolioriskJndreturncharacteristics.Intertemporalconsistencyreferstoconsistencyovervariousinvestmenthorizonsreqardinportfoliodecisionsovertime.Theycara>createshort-termex>÷<tat*>sf，:，makingactiveinvestmentdecis>:ns&15.M.3taAFrameworkforDevelopingCME，ThefollowingisJframeworkoradisciplinedapproachtosettingCME.1.Specif,teset:fexpectationsneededInClUdingthEtimehorizon：s:t>whichtheyapply.2.Researchrhehistoricalrecord3.Specifythemethod(s)and/orm父同，$>IQbeusedandthenInfsmationrequirements.4.Determinethebe“sourcesforinformation'eed>.5.Interpretthecurrentinvestmentenvironmentusingtheselecteddataandmethods,applyingexperiefreandjudgment.6.Povdethesetofexpectationsneededdocum?Eingconclusions7.Monitor.providingfeedbacktoimproveThee×pectatns-sengprocess.Bd2.M2LChallengesinForecasting1.LimitationstousingeconomicdataTl-timelagTheInternationalMonetaiyFund,forexample,reportsdatawithalagofJSmuchJStwoyears.一revisedasthepubhcaton.Additionally,datadefinitionsandmethodologychangeovertime.Dataindexesdieofterebased÷,timeIel!basep<MKhtheyarecalculatedischanged:.Althoughaebasingi$notasubstantialchangeinthedaaitself,theunawareanalystcouldcalculatechangesnthevalueoftheindexesincorrectlyfshedoesnotmakeanappropriateadjustment.ChallengesinForecasting2.DatameasurementerrorsandbiasesTranscriptionerrorsarethemisreportingorincorrectrecordingofinfomotionandaemostSeiaUSiftheyaebiasedinonedirection.SurvivorshipbiasseriesisdeletedfronIhehistoricalpeormaeecodofmanagersorfirmsCH?etion$areoftentiedtopoorperformanceandbiastleIrstoncalreturnUPWJldAppraisaldatafc，illiquidanainfrequent;p,cedassetsmakesthepathofetunsappea,smoothervitactuallys.ThisbiasesdownwardthecalculatedStandaiddeviationandnukesIhereturnsseemlesscorrelatedKlQSetQO)withmoreliquidpricedassetsThis田aparticularproblemfosometypesofalternativeassetssuchasrealestate.Sd5MEgChallengesinForecasting，3.LimitationsofhistoricalestimatesValuesfromhistoricaldatamustoftenbeadjustedgoingforv.ardasecoonicpolitical,recuhtory.3'Idtechnolog<alenvonnentschangeThAispartculaIvtrueforvolatileassetsSUChasequityTliesechangesareknowna.regimechangesj11dresultinnonstationarydata/Fsexample,theglobalfinancialcrisisin20072009?SUlIedinreturnsdatathatweremarkedlydifferentthanthosefromthepreviousfiveyears.11-115/Nonstationarityouldmeandiffaentperiodsinthetimeserieshavedffeentstatisticalpropertiesandcreateproblemswithstandardstatisticaltesngmethods12.wZiQChallengesinForecasting，3.Limitationsofhistoricalestimates(con,t)AIQHgtimePeMQdTpefeablef<>See?alreasons：It,vj.bestatisticallyrequiredTQCWLHaIkIlDWJlco,.aance(andcorrelation),thenumberofdatapointsmustexceedthenunb0ofcovariancetobecalculated/AlargerdatasetTimeperiod)providesmore,e<isestatisticalestimateswithsmallervaancetotheestimates.Usingashorttimeperiodcreatesatemptationtousemefrequentdatasuchasv/eeklydata,ratherthanmonthlydatapointsInordertohavealargersamplesizeUnfortunately,morefrequentdatapointsareoftenmorelikelytohavemissi%joroutdatedvaluesIthHisUJledasynchronousdataandc,esultrlower,distortedcorrelationcalculationsChallengesinForecasting4.ExPostriskasabiasedriskmeasureofExAnteriskUsir'÷>r<stdata,atrerrhefactodetermineeanv?beforethefactriskandreturnSnbepiOblematic.Theanalyst/.ouldunderestimatetherisksthatequitvresorsfa<eanoverestimatetheirpotentialreturnsHQ，ciatasenes<deseven<,)ecbs-.,substantiallyoverstatethelikelihoodofSUCheventshappeninginthefuture./As3simpleexampletherevere21tradingdaystnJuly2018.On26July,thepriceQfRKebookstockcloseddown19：.Basedonthissamplethe<interpolated)d<y5c：VJRsFa<ebookstockis173%.&15.M.3taChallengesinForecasting，5.BiasesinAnalysts'MethodsUsinqhistoricaldata,ara,>tsa】alsouncoverpatternsinsecritetrnsthatareunlikelytooccurinthefutureandcanproducebiasesinthedata.3Data>mining:Justbyalom<hace.somevariableswilloppeartohave3relationshipwithSeCLrftyreturns,when,infacttheserelationshipsareunlikelytopersist.14-115Timeperiodbiaselatestoresultsthataretimeperiodspecific.Researchfindingsareoftenfoundtobesensitivetotheselectionofstartingardorendingdates.12.WZiQChallengesinForecasting5.BiasesinAnalysts'MethodsH<>wtoavoidthesebiasesFirstananalyst$lK)uldfirstWSk'HTV->an.ec<n*:：?as$thevariablesfoundtoberelatedc-stockerurrs/SecondheShQUldscrutinizethemrgprocesstosus<÷>tbMtobias.Third,theanalystshouldtestUYd2：SmIm:;÷la*ms*.*叶:-it-of-sampedatatcdeemineifH*r.W5N's$÷r<srertThis.c>uldbeonebyr<tint<r÷l.j11:'>lr,<repon»onofthehistorical3X3andverr÷rirItv.ranotierportionChallengesinForecasting6.FailuretoaccountforconditioninginformationTherelationshipber.÷÷nsecu,n.÷rnsandeconomicariablessnotCOiistantOVWltime.卜-.，_-一_-1 .1:andeconomicconditions.Thusanalystsshouldaccountforcurentconditionsintheirforecasts.$我aChallengesinForecasting7.MisinterpretationOfCorrelationst；thesesituations*r><p-e,.:,÷htjv÷jtior$r'n；:<UlJMdFKbeusedira：.÷<<.(i'.eJT/MNitinvestigatingtheunderlyinglinkagesAlthoughapparentlysqf<art<oelatic>nscanbespurious,itisalsoTrueThatackofastrongConWJtioncant>emisleading/Anegligiblemeasured<relatonmayeflectastrongbutnonlinearrelationshipAnalystsshouldexplorethispossibilityftheyhaveasolidrea$onforbelieving3relationshipexists.口r"ChallengesinForecasting，8.PsychologicalbiasesAnchoringbias:Iwfust111.fat<,e<edscewegltedStatusquobias:”改hcm”sarehighl,l用Iu-JdhJrEr÷certpas.Confirmationbias:onlyinformation<upxtngtheexistrgbeliefKconsidered,andSlKhevidencenaybeactivelysoughtwhileQtherevidenceisignored.Overconfidencebias:pastmistakesarer>redelackofBmmTntSfromothersistaker*asagreement,andtheaccuracyofforecastsisoverestimated.Prudencebias:forecastsareovelycnsevat.etoavoidtheregretfrommakingextremeforecaststhatcouldendupbeingincorrect.Availabilitybias:.at$easiesttorem÷mb÷!OrtRnarextremea、misoverweighted.ChallengesinForecasting9.ModeluncertaintyModeluncertainty:referstoselectingtheColaTtmatmParameteruncertainty:etes1<:esUmjlinnTnOrSinmodelparameters.Inputuncertainty:'i÷-.lmodel.；“>aAnalysisofEconomicGrowthr÷÷r:<"trend：t>llong-term，；F-f三FE<r.r-:r÷r<b;.-.>rU：-；-,_一Businesscycles'a,*djst<5o.,t:彳>?门:Zdngrec÷s''.<3T÷:-ou<lzccms.MshockscannotbepredictedFoiexample,turnolintheMiddleEastmaychangethelong-termtrendforoilpnces.nflaron.andeconomicgrowthinThedevelopedwork!.Shocksma,alsoarisethroughthbankingsystem/Ane>lremeexampleistheU.S.bankingCsisofthe1930$,.whenasevereslowdowninbanklendingparalyzedtheeconomy.ExogenousShockstoGrowthaExogenousShOCkSareunanticipatedeventsthatoccurousdethenormals:'JnKUorlBe<auserleeventsyunanticipatedthe：.JlemNalreadybuiltintocurrentmarketprices,whereasnormaltrendsinanecc>nomy1whichwouldbeconsideredendogenous,arebuhintomarketprices.1statisticalregimechanges.ExogenousShockstoGrowthExogenousShoCkScanbecausedbyseveralfactorsChangesingovernmentpoliciesGovernnientpoliciesthatSnencowagelong-termgrowthincludesoundscalpolicyminima!governmentinterferencewithfreemarkets,facilitatingcompetitionintheprivatesector,developmentofinfrstructureandhumancapital,andsoundtaxpolicies.Politicalevents22-ll/Geopoliticaltensionsdialdvetesou<estolesspodu<tveu$e$mayleadtodecreasesingrovh.Conversely,CUtSindefensespendingduetohigherlevelsofworldpeacemayleadtoincease$incpovh.Sd5.wWgExogenousShockstoGrowthTechnologicalprogressThecreationofnevandinnovativemarkets,products,andtechnologieshasthepotentialtoimprovegrowth.Naturaldisasters/NJtuiaIdisasterslikelyreduceshot4emgrov/th.Ixitmaarguably；encouragelong-termgrowthIfmoreefficientcapa<ityrepla<espreviouscapacity.DiscoveryofnaturalresourcesPnjductionofnewnaturalre>ou<÷s>,r÷r>djchc'orR-、.“可：Storecoverexistingresources<an÷narc÷g,j.thFinancialcrises/ShOCkStothfinancialSySteMiIl÷adtoJ<r.><ofconficlr'<÷amongmarketparticipants.Finan<lLKWSv.educethle/elo；economicoutputintheshortremar?ra3$.：.decreaseete*drateofgrowth.叁.-MApplicationofGrowthAnalysistoCMETherendrateEgrowthsanimp0113rtnjr÷11>÷"i<aijlmarkerexpe<tatons.SomeofthekeyconsiderationsQfecono11n<goMhtrendanalysisareasfollows:ForecastingetunseitDCFmodelsincorporaterhetrendrateOfgrowth.Highertendg,vtr7tesna,÷adtohigherstockretrnsassumingthegrowth4notalreadyreflectedinstockpricesWhenwespeakofhighetef<0v.1h13【eswemeantheeconomycangrowatafasterpacebeforeinflationbecomesamajorconcern.ThisconsiderationinfluencesmonwyPOliCyandtelevelofbondyields.HlTg!'dUl15；yields.DecompositionofGDPGrowth“ThetrendgrowthinGDPisthesumofthefollowing:GlaVthfromlaborinputsgrowthinpotentiallaborforcesize/growthinactuallaborforceparticipationGlgwthfromlaborproductivity/growthfromcapitalinputs：M1:/totalfactorproductivity(TFP)growth(i.e.growthfromincreaseintheproductivityinusingcapitalinputs)、2轩HBEconomicGrowthTrends，ForecastingGDPTrendGrowthIfwehavethef?lowinassumptions/TleSiZTQftheCanadiarllaborforceKillgcaat1percentPEyearbasedCnpopulationqiectios«LaborforcepartipatiowillgrowJt0.25percentperyear.Growthfomcapitalinputswillbe1.5percentperyea./Totalfactorproductivitygrowthwillbe0.758.Forecastthetrendgrov/thinCanadianGDRCorrectAnswer:26-115ThetrendgrowthinGDP=growthinpotentiallaborforcesize+growthinactualIaboiforceparticipation*growthfromcapitalinputs*totalfactorproductivitygrowth=1%+0.25%+L5%+O75%=35%.11«二IQ2.AnchoringAssetReturnstoTrendGrowthThetrendgrowthratealsoprovidesananchorforlong-runequityappreciation.7rofthreefactors:thelevelofnominalGDPtheshareofprofitstheeconomy,S(eaningsGDP;anc；Ih=PEiatKTPE：17=GDPfXSfXPEt3Asjresultinthelongrun,thegtovhYaOftI记tocaIaiUJofequityinaneconomyislinked10theg<vvthIamfGDPOverfinitehorizons,thewayinwhichtheshareofcapitalandtheP/EmultipleareexpectedtoCharqewillalsoJ讦metrheforecastOfthetotalvalueOfequity,JSwell3$tscorespondinggroMliateoverthatperiod.Example”Long*RunEquityReturnsandEconomicGrowthInJanuary2000.Alena¾ornsdottirCFA,wasupdatingherfirmsprojectionsforUSequityreturns.Thefirmhadalwaysusedthehistoricalaveragereturnwithlittleadjustment.Bjornsdottirwasawarethathistoricalaveragesaresubjecttolargesamplingerrorsandwasespeciallyconcernedaboutthisfactbecauseofthesequenceofveryhighreturnsinthelate1990$.ShedecidedtoexaminewhetherUSequityreturnssinceWorldWar11hadbeenconsistentwitheconomicgrowtl.Fortheperiod1946-1999,thecontinuouslycompounded(i.e.,logarithmic)returnwas12.18%perannm1whicheflectedthefollowingcomponents:RealGDPGrowthInflationEPS/GDP(Chg)PZE(Chg)DividendYield3.14%4.12%0.00%0.95%3.97%ExampleQUeStiOnS1.WhatconclusionwasBjomsdottirlikelyhavedrawnfromthisanalysis?，CorrectAnswers:Bjomsdottirislikelyhaveconcludedthatthepost-warstockreturnexceededwhatwouldhavebeenconsistentwithgrowthoftheeconomy.29山5Inpa11kular.therisingP/Eadded95%ofextrareturnperyearfor54years,adding51.3%(=540.95%)tothecumulative,continuouslycompoundedreturnandleavingthemarket67%(e"冷=1.67)above"fairvalue：11«二IQExample"Questions2.IfShebelievedthatinthelongrunthattheUSIaboiinputwouldgrowby0.9%percnumandlaborproductivityby1.5%thatinflationwouldbe2.1%thatthedividendyieldwouldbe2.25%,andthattherewouldbenofurthergrowthinP/E,whatislikelytohavebeenherbaselineprojectionforcontinuouslycompoundedIong-IHTnUSequityreturns?CorrectAnswer:HerbaselineprojectionisIlkelytohavebeen6.75c?=0.9°：+1.5:,：+2IGc+2.25%.Example”Questions3.Inlightofheranalysis,howmightshehaveadjustedherbaselineprOjection?"CorrectAnswer:Bl-U5SheislikelytohaveadjustedheprojectiondownwardtosomedegreetoreflectthelikelihoodthattheeffectoftheP/Ewoulddeclinetowardzeroovertime.Assuming,forexample,thatthiswouldoccurove30yearswouldimplyreducingthebaselineprojectionby171%=(51.3%3O)peryear.jf2.wEta3.ApproachestoEconomicForecastingThreeapproachestoeconomicforecastingareeconometricmodeling,useofeconomicindicators,andachecklistapproach.«1.EconometricanalysisusesstatisticalmethodstoexplaineconomicIelJtionshipsandformulateforecastingmodels.Stiuctuialmodelsarebasedoneconomictheorywhilereduced-formmodelsarecompactversionsofstructuralapproaches.Structuralmodelsspecifyfunctionalrelationshipsamongvariablesbasedoneconomictheory.Thefunctionalformandparametersofthesemodelsarederivedfromtheunderlyingtheory.32山5VReduced-formmodelsIiaveJlooserconnectiontotheory.11«二IQApproachestoEconomicForecastingAdvantagesofEconometricAnalysis3Modelingcanincorporatemanyvariables.OnCethemodelisspecified,itcanbereused.OutputisquantifiedandbasedonaconsistentsetofrelationshipsDisadvantagesofEconometricAnalysisModelsarecomplexandtime-consumingtoconstructThedatamaybedifficulttoforecastandtherelationshipscanchange.Outputmayrequireinterpretationorbeunrealistic.ITdoesnotworkwelltofore<:astturningpointsApproachestoEconomicForecasting2.Economicindicatorsaeavailablefromgovernments,internationalorganizations,andprivateorganizations.Tl-mostusefulindicator