《CFA三级知识点必备：Equity Portfolio Management_标准版.docx》由会员分享，可在线阅读，更多相关《CFA三级知识点必备：Equity Portfolio Management_标准版.docx（13页珍藏版）》请在课桌文档上搜索。
1、寸二6UOHq0CCl一鼠法皿肾一Rgf藤三UI IEquityInvestmentStyleClassificationsAThetwomainapproachesinstyleanalysisaretheholdings-basedapproachandthereturns-basedapproach.ATheholdings-basedapproach1undlnveslmenlsiylelooksattheattributesofeachValueBlendGrowthindividualstockinaportfolioandaggregatestheseattributesSize
2、1.argetoconcludetheoverallstyleoftheportfolio.AcommonapplicationofthisideaistheMorningstarStyleBox.MidSmallHoldings-BasedStyleAnalysisAThestyleboxapproachaimstoclassifyapproximatelythesamenumberofstocksineachofthevalue,blend,andgrowthgroups,essentiallydistributingthemarketvalueofeachrowevenlyacrosst
3、hegrid.ATheclassificationofstocksintovalue/blend/growthinvolvesassigningastylescoretoeachindividualstock.Forexample,toassignavaluescore,thedividendyieldmaybeused./StockswouldberankedaccordingtotheirdividendyieldandaSCOreallocatedtoaStcICkbasedOntheirPe(CentiIeOfthemarketvalueoftheirparticulargroup./
4、IfthestockisattheIoWer(high)endOfthedividendyieldrange,itWillreceiveaIc)W(high)SCoreCloSet。OQOO).Acomprehensivescoringmodelwouldusemanyindicatorsofvalueandcombinethemtogetherinapre-determinedweighting.Holdings-BasedStyleAnalysisATheclassificationofstocksThedifferencebetweenthestock,sgrowthandvaluesc
5、oresiscalledanetstylescore./Ifthenetstylescoreisstronglynegative,thestockisclassifiedasvalue./Ifthenetstylescoreisstronglypositivethenthestockisclassifiedasgrowth./Ifthenetstylescoreisclosetozerothenthestockwillbeclassifiedascore.Onceconstructedforastockuniverse,thegridcanbeusedasavisualaidtohelpcat
6、egorizeandtrackmanagedinvestmentportfolios.，Ataglance,aninvestorcanseewhereamanagerispositionedonthegrid,and,ifhistoricaldataexists,howthisstylehaschangedovertime.Returns-BasedStyleAnalysisAreturns-basedstyleanalysisaimstoidentifythestyleofafundthroughregressionofthefundsreturnsagainstasetofpassives
7、tyleindices.AByimposingaconstraintontheregressionthatthesumoftheslopecoefficientsshouldsumtoavalueof1,theslopecoefficientscanbeinterpretedastheanager,sallocationtothatstyleduringtheperiod.Forexample,areturn-basedstyleanalysismightconductaregressionoffundreturnsversusfourpassiveindicesasfollows:+斗存=t
8、nereturnostylendesmesamepero雪=1 and s O fors=thefundexposuretostyles(withconstraints.along-onlyportfolio)=aconstantofteninterpretedasthevalueaddedbythefundmanagert=theresidualreturnthatcannotbeexplainedbythestylesusedintheanalysisManagerSelf-IdentificationAThefud,sinvestmentstrategyisusuallyself-des
9、cribedbythemanager.AComparingthatself-descriptionwithreturns-basedandholdings-basedstyleanalysiswilleitherconfirmaconsistentidentificationorindicateaneedforfurtherinvestigationandanalysistoexplainthediscrepancy.Somestylessuchasequitylong/short,equitymarketneutralandshortbiasdonotfittraditionalstylec
10、ategoriesandthemanagersdescriptionandfundprospectusbecomesthekeysourceofinformationonstyleofsuchfunds.Equitystyleanalysis-Comparisonof2tech.AdvantagesDisadvantagesReturn-basedRequiresminimalinformationCanbeexecutedquicklyCosteffectiveMorewidelyappliedMaybeineffectiveincharacterizingcurrentstyleDiffi
11、culttodetectmoreaggressivepositionsHolding-basedMoreaccuratethanreturns-basedFacilitatescomparisonsofindividualpositionsCapturechangesinstylemorequicklyMoredataintensivethanreturns-basedanalysisLesseffectiveforfundswithsubstantialpositionsinderivatives.巨业创新憎值一2.PortfolioConstruction专业创新增值APassively-
12、managedindex-basedequityportfolioscanbeconstructedby:FullReplication:fullreplication(holdallofthesecuritiesintheindex)StratifiedSampling:holdasampleofthesecuritiesbasedonstratifiedsamplingOptimization:usemorecomplexoptimizationtomaximizedesirablecharacteristicswhileminimizingundesirablecharacteristi
13、cs.BlendedApproach:inpracticeablendoftheseapproachesmaybeused.AFullreplicationcanbecostlywhentherearelargenumbersofstockandliquidityislimited.Theportfoliomustberegularlyreconstitutedandrebalanced.Theadvantageoffullreplicationisthatitcloselymatchestheindexreturn(beforetransactioncosts).AToavoidthehig
14、hcostoffullreplication,itoftenmakesmoresenseforthemanagertousestratifiedsampling,inwhichheholdsasubsetoftheconstituentstocks,withthesampleselectedinsuchawayastoreplicatetheindexreturn/riskcharacteristics.Toimplementstratifiedsampling,themanagercreatesstrataacrosstheconstituentstocksthataremutuallyex
15、clusiveandexhaustive.Themanagermustconsidersizeofthesampleused.Asmorestocksareaddedandtheportfolioapproachesfullreplication;lessliquidstocksareadded,increasingtransactioncostandtrackingerror.Optimizationusesthetoolsofmodernportfoliotheorytoaddresstheproblemofminimizingtrackingerror.Theoptimizerseeks
16、thecombinationofstocksthatwouldhaveminimizedtrackingerrorandpossiblymaximizedreturn.Theadvantagesofoptimizationtechniquesistheytypicallyexhibitlowertrackingerror,andthattheyexplicitlyaccountforthecovarianceamongconstituentstocks.Theobviousdrawbackofoptimizationisthatitisbasedonhistoricalrelationship
17、sandthosecanchange.Maintaintheoptimizationasthedatachangecanbecostly.Anotherdrawbackisthatitcancreateportfoliosthatarenotmeanvarianceefficientrelativetothebenchmark./Thesolutionistoaddaconstraintthattotalportfoliovarianceisequaltothevolatilityofthebenchmark.ABlendedApproachFullreplicationispreferred
18、forindexeswithsmallnumbersofliquidstocks,whilestratifiedsamplingoroptimizationispreferableforindexeswithlotsofheterogeneous,thinlytradedstocks./ForlargeindexesliketheWilshire5000,theconstituentstocksrunthegamutfromlargeandliquidtosmallandthinlytraded.Inthatcaseacombinationoftwoapproaches,fullreplicationandstratifiedsamplingoroptimization.
