Swedish(Equity(Mutual(Funds( :(( Performance,(Persistence(and(Presence(of(Skill (
|
|
|
- Beatrice Fox
- 10 years ago
- Views:
Transcription
1 Swedish(Equity(Mutual(Funds( :(( Performance,(Persistence(and(Presence(of(Skill ( HarryFlam InstituteforInternationalEconomicStudies,StockholmUniversity,andCESifo RoineVestman DepartmentofEconomics,StockholmUniversity,andSIFR September,2014 Abstract( AveragegrossexcessreturnsofactivelymanagedSwedishequitymutualfundsdecreased substantiallyaround2001j2002.averagegrossandnetexcessreturnswerej0.22andj1.47per centperyearin2002j2013.thereislittleornopersistenceinreturns.whenfundsareranked onpastperformance,theirreturnsconvergetothecrossjsectionalmeaninabouttwoyears andstayclosetothatsubsequently.inaddition,actualgrossexcessreturnsdonotdiffer significantlyfrombootstrappedexcessreturnsthatarezerobyconstruction.therefore,thereis practicallynoevidenceofstockjpickingskills.selectedindexfundscanbeexpectedto outperformactivelymanagedfunds.( WearegratefultoHannaMühlradforresearchassistance,toMoneyMateandMorningstarforreturndata,to thedatacenterattheswedishhouseoffinance,andinparticularerikeklund,fortheprovisionofstockmarket andcompanydata,and,withoutimplication,toseminarparticipantsattheswedishhouseoffinance.
2 1.Introduction SwedishhouseholdsinvestheavilyinSwedishequityfunds,i.e.equityfundsthatprimarilyfocus ontheswedishstockmarket.thereareseveralchannelsthroughwhichhouseholdsare exposedtothesefunds.first,70percentofallhouseholdsholdstocksorequityfunds,anda thirdofthosehouseholdsinturnholdaswedishequityfund. 2 Second,practicallyallSwedish wageearnersownsharesinmutualfundsthroughamandatorydefinedcontributionpension planinthepublicpensionsystem.alhhoughtheplanisanopenplatformwithamenuofas manyas800fundstochoosefrom,28percentoftheinvestorsholdaswedishequityfund, whichmeansthat13percentoftheplan sinvestmentsareallocatedtoswedishmutualequity funds,aportfoliosharemuchhigherthanwarrantedbysweden sshareofglobalmarket capitalization. 3 Finally,90percentofSwedishwageearnersholdadditionalinvestmentsin occupationalpensionplanswhichareheavilyexposedtowardsswedishequityfunds.the marketvalueofswedishhouseholds totalinvestmentsinthesefundswasaboutsek360 billionin2013,theequivalentoftenpercentofgdp. 4 Consequently,thechoicebetween differentswedishequitymutualfundsmaybequiteimportantformanyswedes. Mostinvestorsinequitymutualfundspresumablyhavelittleornoknowledgeaboutthe equitiesthatfundshold;theysimplywanttohaveexposuretothestockmarket.whenthe financialindustryormediaexpertsgiveadvicetoinvestorsonthechoiceoffunds,they frequentlyrecommendactivelymanagedfundswithrelativelyhighpastperformance,basedon thebeliefthathighperformancecanbeattributedtostockpickingskillandthatsuchskillis persistent.incontrast,whenafinanceresearcherisaskedforadvice,heorsheislikelyto recommendalowjcostindexfund,basedontheefficientmarkethypothesisthatactivelyand passivelymanagedfundscanbeexpectedtoearnthesamereturnasthestockmarketbefore cost,butthelatterareexpectedtodosoatlowercost. MostoftheempiricalevidencebasedondataontheU.S.mutualfundmarketsupports theefficientmarkethypothesis.carhart(1997)findsthatmostfundsunderperformbyabout themagnitudeoftheirinvestmentexpenses,kosowskietal.(2006)thattheaveragenetriskj 2 TheaveragevalueofstocksandmutualfundsinregularsavingsaccountswasSEK223,000in2007.Ofthat,SEK 16,900(7.6percent)wasinvestedinSwedishequityfunds.Thisisasidefrominvestmentsinprivatepensionplans andlifeinsuranceproducts.weusethemicrodatasetlinda,wave2007,matchedwithdetailedassetsholdings tomakethiscomputation. 3Employersmakecontributionstotheplanequalto2.5percentofgrosswages.Basedonhistoricalratesof return,thefundedpartofthepublicpensionsystemmaycometocontributeasmuchtowageearners public pensionsasthenonjfunded,transferpartofthesystem.ourcalculationisbasedonthelinda2007samplefor whichwehaveaccesstodetailedinformationaboutthefundholdingsofthepensionplan.wehaveexcluded investmentsintotheplan sdefaultfundfromthiscalculation. 4 SwedishInvestmentFundAssociation,TheSwedishFundMarketinFigures, 1
3 adjustedexcessreturnperyearisj1.2percent,famaandfrench(2010)thatitis 1percent, Barrasetal.(2010) 0.5percent,andBerkandBinsbergen(2012) 0.7percent. Despitethefactthatinvestmentinmutualfundsismorewidespreadamonghouseholds inswedenthaninmostcountries,thereislittleevidenceontheperformanceofswedishequity mutualfunds.onlytwostudies,bothbasedonfundperformanceduringafewyearsinthe 1990 s,havebeenpublishedand,forthetimeperiodtheycover,theresultsareatoddswith thefindingsfortheu.s.market.thefirststudy,bydahlquistetal.(2000),findsthatfunds qualifyingforthepreferentialtaxtreatmentthatexistedatthetimehadnegativeaverageand mediannetexcessreturnsof 1and 0.7percent,andthatfundswithoutpreferentialtax treatmenthadpositivebutinsignificantaverageandmedianexcessnetreturnsof0.5and0.1 percentperyear.thesecondstudy,byengström(2004),findsanaveragenetexcessreturnof 1.7percent. Ourstudycoverstheperiod1993J2013.Splittingthesampleintwohalves,wefindthat excessreturnsdiffersubstantiallybeforeandafter2001j2002.in2002j2013,averagegrossand netexcessreturnswere 0.18and 1.47percentperyear.Thedeclineinreturnscouldbe explainedbyafortuitousbiastowardsitjstocksinthe1990 sandbygreatercompetition subsequently thenumberoffundsdoubled.wearguethatinferencesabouttheperformance oftoday sstockmarketshouldbebasedonthelatersubperiod,2002j2013. Thetoptenfundsplustwomore(of113)in2002J2013hadsignificantpositivegross excessreturns.thisisconventionallyinterpretedasevidenceofstockjpickingskill,butcould alsobeduetoluck.afew,fiveofthetopsevenfunds(of124),weresufficientlyskilledorlucky toearnsignificantpositivenetexcessreturn.theywereinotherwordsabletomorethan compensateinvestorsformanagementcosts. Wealsoestimateexcessreturnsofindexfundsrelativetotheirrespectivebenchmarks. Justasforactivelymanagedfunds,thereisadrop butsmaller inexcessreturnsbeforeand after2001j2002.theaveragegrossandnetexcessreturnsinthelatterperiodwere 0.22and 0.84percent. Asafirsttestofthepresenceofskillinfundmanagement,weanalyzethepresenceof persistenceinfundreturns.persistenceisofteninterpretedasanindicationofskill.however, wefindlittleevidenceofpersistenceinreturnsandconsequentlylittletoindicatethepresence ofskill.regardlessofwhetherfundsrankhighorlowbasedonpastperformance,theirreturns convergequicklytothecrossjsectionalmeanandbecomesimilaraftertwoyears.thisaccords wellwiththefindingsofcarhart(1997)foru.s.equitymutualfunds,exceptthathefindsthat thebottomdecileoffundsunderperformspersistently. 2
4 Thepresenceofpersistenthighreturnsmayindicatethepresenceofskill,butitcanalso betheoutcomeofluck.wemakeadirectinvestigationofwhethersuperiorandinferior performanceshouldbeattributedtosuperiorandinferiorskillortogoodandbadluckby employingaversionofthebootstrapmethodinkosowskietal.(2006)andfamaandfrench (2010).SimulatedcrossJsectionaldistributionsofgrossandnetreturnsarecreatedrepeatedly underthenullhypothesisthatgrossandnetexcessreturnsarezero.thecrossjsectionof(thetj statisticsof)actualgrossandnetreturnsisthencomparedtothecrossjsectionof(thetj statisticsof)simulatedgrossandnetreturns.tjstatisticsofexcessreturnsisthepreferredtest statisticratherthanexcessreturnstoaddressvariousstatisticalproblems,butweshowboth. Wefindthatonlyonefundhasactualgrossexcessreturnsthataresignificantlyhigherthanthe bootstrappedgrossexcessreturnsintheperiod2002j2013,inotherwords,thereisverylittle evidenceofskill.nofundhadasignificantpositivenetexcessreturn. 2.Datasourcesanddataconstruction Ourfunduniverseconsistsof124activelymanagedmutualfundsholdingabroadsetof equitiesofcorporationslistedonthestockholmstockexchange,plus20passivelymanaged fundstrackingavarietyofindexesrelatedtothestockexchange.specializedfunds,forexample fundsthatinvestinspecifiedindustries,areexcluded.itshouldbenotedthatsomefundsmay haveasmallshareoftheirassetsinequitieslistedabroadandthatfundsmustholdaminimum amountofliquidityfortransactionpurposes;theamountcanvaryovertime. OurperiodofinvestigationisJanuary1993toDecember2013.Theminimumnumberof consecutivemonthlyreturnsrequiredforafundtobeincludedis36. 5 Fundsthatmeetthe requirementbuthavebeenclosedandmergedwithotherfundsareincludedtoavoidsurvival bias. ThedataonmonthlynetfundreturnsweresuppliedbyMorningstarandarebasedon primarydatasuppliedbytheswedishinvestmentfundassociation.fundreturnsincluderej investeddividendsandarenetofexpenses.somefundsaresoldbothintheretailmarketand inthemarketforvariouspensionplans.insuchcases,onlythedatapertainingtotheretail marketareincluded. Grossreturnsareobtainedbyaddingbackcostsasgivenbythetotalexpenseratio (TER)tonetreturns.TERisreportedannuallytotheSwedishFinancialSupervisoryAuthorityby 5 Weexclude28activelymanagedfundsand5indexfundsthathave35orfewerreturnobservations. 3
5 mostbutnotallfunds. 6 WehaveobtainedTERfromtheannualfinancialstatementsfor114out of124funds. 7 Toobtaingrossreturns,weaddbacktheaverageTERfortheyearsforwhichwe havedata. Ouranalysisisbasedon1Jfactorexcessreturns.Datafortheconstructionofsystematic riskfactors valueversusgrowthbiasandsmallversuslargecapbias areunfortunatelynot availableafter2009.tocheckforpotentialdifferencesbetween1jfactorexcessreturnsonone handand3jand4jfactorexcessreturnsontheother,wehave,however,estimated3jand4j factorexcessreturnsfortheperiod1999j2009andcomparedthemwithestimated1jfactor excessreturns. 8 3.Performance TheCAPM,the3JfactormodelbyFamaandFrench(1993)anditsextensiontofourfactorsby Carhart(1997)areconsistentwithmodelsofmarketequilibriumwithone,threeorfour systematicriskfactors.theycanalsobeinterpretedasmodelsforperformanceattribution;we willusethemassuch. ThefollowingtimeseriesregressionsattributereturnsinexcessoftheriskJfreeinterest ratetoone(capm),three(famaandfrench)andfour(carhart)systematicriskfactors respectively: 6 WenotethattheSwedishmeasureoftotalexpenses,TKA,ismoreinclusive.Inadditiontofeesandtradingcosts, italsoincludescommissionsandresultsjbasedfeesinpercentoftheaveragevalueofassetsundermanagement duringtheyear.theterforthe114fundsforwhichwehavedatais1.3anditis1.6percentforthetka. 7 OursampleincludessomefundswhicharedomiciledabroadandthatthereforedonotreportTERandTKA.Five fundsareregisteredinluxembourg,oneinnorway,oneinireland,oneingreatbritainandtwohavenodomestic jurisdiction(isincode)reportedinthedataset. 8 ThedatawerekindlydeliveredbytheDataCenterattheSwedishHouseofFinance.Theywereoriginally constructedbysixtelekursbefore2004andbynasdaqomxfortheperiod2004j2009.weconstructthesize (SMB)andvalue(HML)factorsofFamaandFrench(1993)andofCarhart s(1997)momentum(mom)factoras follows.wefirsttakeaccountofthefactthatswedenhasamultipleshareclasssystem.foreachcorporation,we determinethelargestshareclassintermsofmarketvalueanduseonlythatclass.wethenattributetheentire marketvalueofthecorporationtothisclass.foreach12jmonthperiodstartinginaprilofyeart"1andendingwith Marchofyeartwerequire12recordedmonthlyobservations.Wethenmatchthebookandmarketvaluesfrom MarchofyeartwiththereturnseriesfromAprilofyearttoMarchofyeart"1.Wealsocontrolforwhichparticular stockmarketlistthatthestockwastradedonindecemberofyeart"1byexcludingstockmarketsthatarejudged tobetoosmallandilliquidtobesuitableasinvestmenttargetsforamutualfund. 8 TheSMBandHMLfactorswere thenconstructedasinfamaandfrench(1993).themomfactorwasconstructedbyfirstsortingallstocksontheir 12Jmonthlaggedreturns,andthenusingthebottom20percenttoconstructaportfoliotogoshortin,andthetop 20percenttoconstructanotherportfoliotogolongin.Betermieretal.(2013)alsoconstructtheirfactorsfromthis material. 4
6 "#$%&1 = + ("#$%# "#$%&1) + (1) "#$%&1 = + "#$%# "#$%&1 + "# + "# + (2) "#$%&1 = + "#$%# "#$%&1 + "# + "# +" + (3), where$ =, isthegrossornetreturnoffundi$attimet$(fundandtimesubscriptsare omitted),stibor1m$isthestockholm1jmonthinterbanklendingrate,αisthenetexcessreturn (thereturnleftunexplainedbythebenchmarkmodel),β,, andarefactorloadings, "#$%#isavaluejweightedindexforallcompanieslistedonthestockholmstockexchange, includingreinvesteddividends,wheretheweightofasinglecompanyisrestrictedtoreflect UCITregulationsthatappliestomutualfundportfolios,"#isourSwedishversionofthe HighJMinusJLowbookJtoJmarketvaluefactor,"#isourSwedishversionoftheSmallJMinusJ Bigmarketcapitalizationfactor,"isourSwedishversionofCarhart smomentumfactor, whichislongonpriorjyearwinnersandshortonpriorjyearlosers,andistheregression residual. 9 Equation(1)isestimatedforallactivelymanagedequitymutualfundswithaminimum of36monthlyreturnobservationsfortheperiodjanuary1993 December2013.Forindex funds,weestimateequation(1)usingtheirrespectivebenchmarkindexesandnotthesixprx index.requiringaminimumreturnhistoryof36monthsmeansthatwehavepracticallyno survivalbiasinourestimates,butalsothatestimatesfortheshortestjlivedfundstendtohave lowerprecision.equations(1) (3)areestimatedfortheperiodJanuary1999 December2009 foractivelymanagedfunds.grossreturnsarecomputedbyaddingtheaverageterofthefund totheestimatednetexcessreturn. 3.1Activelymanagedfunds ConsiderfirstthetimeseriesshowninFigure1.ThedashedlineshowstheSIXPRXindexand thesolidlinethethreejyearmovingaverageofmonthly1jfactornetexcessreturns(netalphas) ofallactivelymanagedswedishequitymutualfunds. 10 Itisclearthatthetimeseriesoffundnet excessreturnsexhibitsasharpbreakaround2002.beforethebreak,netexcessreturnsare 9 ThereiscontroversyaboutwhethertheaverageSMB,HMLandMOMreturnsarerewardsforriskortheresultof mispricing.regardless,itremainstruethatfundmanagerscanimplementpassivestrategiestocapturereturnsto sizeandvaluebiasandtomomentum,andthatstockjpickingability,i.e.activemanagement,shouldshowupin theintercept(alpha). 10 TakingthethreeJyearmovingaverageinsteadofashorterperiodsuchastheoneJyearaverageeliminatesmuch noise. 5
7 positiveandvaryaround2.5percentperyear,andafterthebreaktheyvaryaround 1.5per cent.theindexdoesnotshowasimilarbreak;ittakesadivein2000j2002,butreturnsto nearlythepreviouslevel. [FIGURE1] Onemayspeculateaboutthereasonsforthebreakinnetexcessreturns.Wearguethat twofactorscombinetoprovideatleastapartialexplanation:abiastowardsitjstocksinfunds portfoliosbeforethecrashin2000,andmuchgreatercompetitionamongfundsafterthe break,duetoagreatincreaseinthenumberoffunds.ourdatasetincludesatotalof60 activelymanagedfundsin1993j ,and124fundsin2002J2013.Theincreasewasprobably causedbythelaunchofthenewpensionsystemin2000,inwhichemployershavetomake contributionsequalto2.5percentofgrossincometothefundedpartofthepensionsystem andemployeescanchoosetoinvesttheircontributionsamonghundredsoffunds. Whateverthereasonsforthebreakinthetimeseriesofnetexcessreturns,itwouldbe misleadingtomakeinferencesabouttoday sperformance,persistenceandpresenceof managementskillbasedondataforthewholetimeperiod.wearguethattherearestrong reasonstobelievethatthemarketforactivelymanagedswedishequitymutualfundschanged fundamentallyaround2001j2002,andthatinferencesaboutthepresentstateofthe performanceoftheswedishequityfundindustryshouldbebasedonthelaterperiod.the choiceofyeartosplitthetimeseriesissomewhatarbitrary;wechoosetodividetheperiodinto 1993J2001and2002J2013. Beforeturningtoestimatesoffundperformance,considersomedescriptivestatisticsin Table1.TheequallyJweightedaverageabsolutenetreturn withoutriskadjustment of activelymanagedfundswas21.75percentin and9.13percentin SincetheaverageTERwas1.3percent,seeTable2,thismeansthatthegrossabsolutereturn washigherthanthestockmarketaverageincludingreinvesteddividends(sixprx)by3.2per centin ,butaboutthesameasthestockmarketaveragein2002J2013. [TABLE1][TABLE2] Table3reportsequalJweightedgrossandnetreturnsabovetheriskJfreeinterestrate andadjustedformarketrisk 1Jfactorgrossandnetexcessreturns ofactivelymanaged fundsasgivenbyestimatedalphasinequation(1). [TABLE3] 11 Dahlquistetal.(2000)include40equityfundswithpreferentialtaxtreatmentand80regularequityfundsin theirsample,whichcoversthetimeperiod1993j1997.weincludenofundswithpreferentialtaxtreatment. 6
8 RiskJadjustedexcessreturnsgivethesamepictureasabsolutereturns:riskJadjusted grossandnetexcessreturnsweresubstantiallypositivein butturnednegativein Theaveragegrossexcessreturnisclosetozerointhelaterperiodandthe underperformanceintermsofnetexcessreturnsisapproximatelyequaltotheaverageter. Thisaccordswellwiththeefficientmarkethypothesis. Table3alsoreportsexcessreturnsforcrossJsectionallyrankeddecilesoffundsandfor topandbottomfunds.thevariationinreturnsacrossdecilesislarge,particularlyintheearlier period,anditisevenlargeracrossindividualfunds.thegrossandnetexcessreturnsofthetop decilein weremorethan10percenthigherthanthatofthebottomdecileof funds,andthegrossandnetexcessreturnsofthetopfundinthesameperiodwereabout22 percenthigherthanthatofthebottomfund.wealsoplotperformanceinfigure2usinga naïveapproachtothemeasurementofoverjandunderperformance.thefigureshows whetherestimatedgrossandnetexcessreturnsaresignificantlydifferentfromzeroatthetwoj sided90percentconfidencelevel. 12 Ascanbeseen,anumberoffundsexhibitsuperior performanceintermsofbothgrossandnetexcessreturnsinthefirstsubperiod,butfewdoso inthesecond.twelveof113fundshavesignificantlypositivegrossexcessreturnsandfiveout of124havesignificantlypositivenetexcessreturnsin2002j2013. [FIGURE2] AlthoughFigure2accuratelydepictsthecrossJsectionaldistributionofnetexcess returns,itisnaïvewithrespecttotheconstructedconfidenceintervals.statisticallysignificant superiororinferiorperformancemaybeduetosuperiororinferiormanagementskillordueto goodorbadluck.infact,inanycrossjsectionoffundreturnstherewillbedispersion, suggestingthatsomefundsoverperform.ouranalysisofperformancepersistenceinthenext sectionandourbootstrapofentirecrossjsectionsoffundsinthesubsequentsectionattempts todetermineifandtowhatextentfundmanagerspossessskillinpickingstocks. Lackofstockdataforlateryearspreventsusfrombasingouranalysisonexcessreturns thatareadjustedformoresystematicriskfactorsthanmarketriskforthefulltimeperiod. However,thenecessarydata,whichneedtoaccountforSweden smultipleshareclasssystem, isavailablefrom1999to2009.tohaveanideaofwhetherfundperformanceislikelytobe substantiallydifferentwhenaccountistakenofmorethanmarketrisk,acomparisonbetween 1J,3Jand4Jfactorexcessreturnsforthe1999J2009periodismadeinTable4.Lookingat averageperformanceandtheperformanceoftopandbottomdecilesoffunds,itisclearthat performanceissomewhatbetteringeneralwhenmoreriskfactorsareaccountedfor, 12 Theconfidenceintervalwasformedunderthestandardassumptionsofindependentandhomoscedasticerrors. CorrectingforautocorrelationusingNeweyJWeststandarderrorsmatterslittle. 7
9 especiallywhenamomentumfactorisaddedtothefamajfrenchhighjversusjlowbookjtoj marketvalueandsmalljversusjbigmarketcapitalizationriskfactors.theaverage4jfactorgross andnetexcessreturnsare20j25basispointshigherthanthe1jfactorreturns,andthepositive tailofthe4jfactordeciledistributionissomewhattotherightof1jfactordistribution. [TABLE4] 3.2Indexfunds Wehavemonthlyreturnsforfiveindexfundsinthesubperiod1993J2001withaminimumof 36observations,andfor20fundsinthesubperiod2002J Thefundstracksixdifferent indicesforstockslistedonthestockholmstockexchange.byusingsuitableproxiesforsomeof thestatedbenchmarkindices,weareabletoreducethenumberofbenchmarkstothree: SIXPRX,whichtracksallcompanieslistedontheexchange,includesdividendsandcapsweights inaccordancewithucits,six30rx,whichtracksthe30largestcompaniesintermsofturnover andincludesdividends,andomxsbcapgi,whichtracksthe71largestcompaniesintermsof freefloatmarketcapitalization,includesdividendsandcapsweightsinaccordancewith UCITS Investorsshouldexpecttoreceiveareturnequaltothatofthebenchmarkincluding dividendsandminusmanagementcosts.thisisrarelythecasebecauseoftrackingerror. Trackingerrorcandependonanumberofdifferentfactors,suchasthetimingandfrequencyof adjustingindexweights,thehandlingofinflowsandoutflows,andthemethodofreplicating thebenchmark.fundsholdingstocksgenerateincomebylendingstockstemporarilyto investors.thus,thereareseveralreasonsforwhythetrackingerrorcanbebothpositiveand negative.costscanalsovarygreatlybetweenfunds.theydependonmanagementstyleand importantlyonfundsize thereareconsiderableeconomiesofscaleinfundmanagement.for largerfunds,itiscosteffectivetoobtainindexreturnsbyholdingthecomponentstocks,for smallerfundstoreplicatetheindexbyholdingindexfutures. 13 Wehaveincludedexchangetradedfunds(ETSs)ingeneral,butexcludedtwothatareleveraged. 14 ThebenchmarkindexforeachfundisreportedinTable6. 15 Onemanagementcompanyofanindexfunddoesnotchargeanyfeebutinsteadappropriatesdividendreturns. Itisevaluatedagainstitsstatedbenchmarkandthedividendreturnistreatedasafee. 8
10 Table5reportsequalJweighted1Jfactorgrossandnetexcessreturnsofeachindexfund againstitsbenchmark. 16 Apositiveornegativegrossexcessreturn apositiveornegative trackingerror meansthatthefundhasperformedbetterorworsethanitsbenchmark. [TABLE5] Performancein1993J2001shouldbedisregarded,bothbecauseofthefewnumberof fundsandbecauseoftheimprobableabnormalreturnsofthetopfund,whichdistortsthe average.lookingatthesubperiod2002j2013,wefindthatindexfundsonaveragehada negativegrossexcessreturn(equaltoanegativetrackingerror)of 0.22percentperyear. Mostfundshaveanegativetrackingerrorandthreehaveanegativetrackingerrorof61basis pointsormore.lookingatnetexcessreturns,wefindthatindexfundshaveanegativenet excessreturnof 0.84percentperyearinthesameperiod.Thetopfundhaspositivenet excessreturnwhichisequaltoitsgrossexcessreturn;itchargesnofeeanddeductsnocosts fromgrossreturns.theverynegativenetexcessreturnofthebottomfundisexplainedbythe factthatitexpresslygaveinvestorsbenchmarkreturnsexcludingdividends,butdidnotcharge anexplicitfee. 4.Persistence Financialadviceoninvestinginmutualfundsisoftenbasedonpastperformance.The presumptionisthatsuperiorpastperformanceisduetoskillandthatskillispersistent.to examinethepresenceoffundswithpersistentsuperiorperformance,wesortfundsaccording toperformanceandthenestimatetheirperformanceduringsubsequentyears,following Carhart(1997).Wenotethatthisapproachissubjecttopossiblemodelmisspecification,since thesameperformanceattributionmodelisusedtorankfundsandtomeasureperformance.if themodelhasabiasinriskadjustment,forexampleduetoanomittedvariable,thebiasin rankingwillalsoaffectthesubsequentmeasurementofperformance. WechoosetorankfundsontheirperformanceduringathreeJyearperiodratherthana shorterperiodtoreducenoise.ourfirstrankingperiodis1993j1995.basedontheirnetexcess returnalphas,wefirstrankfundsbydeciles,thefirstdecilehavingthehighestandthetenth thelowestperformance,andthenindividuallybythetopfiveandbottomfivefunds.wenext estimatetheperformanceofdecilesandtopandbottomfundsfrom1996onwardsinonejto fivejyearjlongevaluationperiods(timeperiod1996j2000).movingtherankingandevaluation periodsoneyearforwardatatime,thisisrepeated17moretimes.thelastrankingperiodis 2006J2008andthelastevaluationperiod2009J2013.Ineachevaluationperiodwerequirethat 16 Indexfundsholdallequitiesinitsbenchmark(oraderivativeofthebenchmarkitself)andthereforetakeno systematicriskintheformofvalueorgrowthbiasorexploitmomentumbyholdingonlyasubsetoftheequities includedinthebenchmark. 9
11 fundsarepresentfromthebeginningtotheendtobeincluded.wethencalculateaverage performanceforthe18evaluationperiods.theresultsarepresentedinfigure3. [FIGURE3a)andb)] Performancevariesgreatlyacrossdecilesintherankingperiod.Notethattheprocedure meansthatthecompositionofdecilesandtheidentityofafundwithaparticularrankchange overtime.thetopdecilefundhasanaveragepositivenetexcessreturnofnearlytenpercent peryearandtheaveragebottomdecilefundanaveragenegativeexcessreturnofnearlyseven percentperyear.theabsenceofpersistenceintheevaluationperiodisquitestriking.average decilereturnshavemoreorlessconvergedtothemeaninthesecondyearandremainclosein thefollowingyears.ithardlypaystoinvestinafundinthetopdecileformorethanoneyear aftertherankingperiod;theaveragenetreturnisclosetozerosubsequently.theseconddecile fundsisamorepromisingbet;excessreturnsarepositiveinthetwojtofivejyearperiods,but investorscanexpecttheworstreturnofalldecilesinthefirstyearoftheperformanceperiod. Thesamepatternofrapidconvergencetowardsthemeanisvisibleforthetopand bottomfivefunds,seepanelb)infigure3.thedispersionofreturnsacrossfundsinthe evaluationperiodismuchgreaterthanforfunddeciles,withadifferenceinyearlynetexcess returnsofnearly30percentagepointsperyearbetweentheaveragetopandbottomfund. Lookingatsubsequentperformance,itisnotclearwhattheadvicetoaninvestorshouldbe. ThetopfundendsupbeingthebottomfundinthethreeJtofiveJyearperiods,andthesecond worstfundbecomesthesecondbestinthesameperiods. Therapidreversiontothemeanofexcessreturnsandthegenerallackofpersistencein rankingsarestrongindicationsthatfundmanagerslackstockjpickingskillsandthatexcess returns positiveornegative areduetogoodandbadluck. 5.Skillorluck? Totestwhethertheperformanceofactivefundsistrulysuperiororinferior,especiallyatthe tailsofthedistribution,weemployabootstrapprocedurethatyieldsadistributionofpseudoj returnsforeachfundundertheassumption(nullhypothesis)thattruealphaofeveryfundis zero.withgrossexcessreturns,settingalphaequaltozeropresumesthatfundsobtainthe sameexcessreturnasthemarketportfolio.withnetexcessreturns,settingalphaequaltozero presumesthatfundscanobtainreturnsthatcoverthecostsofactivemanagement,butnot more.actualestimatesoffundexcessreturnsarecomparedtobootstrappedexcessreturns.if actualgrossexcessreturnsdifferfrombootstrappedreturnswithstatisticalconfidence,we concludethattheabnormalexcessreturnsareduetosuperiororinferiorskillinpickingstocks. 10
12 Ifnot,wecannotruleoutthatactualgrossreturnsareduetogoodorbadluck,andnotto superiororinferiorskill.ifactualnetexcessreturnsexceedbootstrappedreturnswith statisticalconfidence,weconcludethatthefundmanagerhassufficientskilltomorethan compensateinvestorsforthecostsofactivemanagement.ifactualnetexcessreturnsare statisticallysignificantcomparedwithbootstrappedreturnsunderthehypothesisofzeronet excessreturns,weconcludethatthefundmanagerisunabletocompensateinvestorsforthe costsofactivemanagement. Togeneratereturnsunderthenullhypothesisofzeroalphas,wesubtracteachfund s excessreturnasmeasuredbyalphafromequation(1)fromitsmonthlygrossandnetreturns forthepartof2003j2013ithasbeeninexistence.asimulationrunisarandomsamplewith replacementdrawnfromthefund s(alphajneutralized)monthlyreturns,thesamerandom sampleofmonthsforeachfund.eachtimeacertainmonthisdrawn,allfundsinexistencethat montharegivenitscorrespondingadjustedreturnandthecorrespondingfactorreturn(beta) forthatmonth.weestimateequation(1)foreachfundonthesimulationdrawofadjusted fundreturnsandfactorreturns.eachsimulationrunproducesacrossjsectionofalpha estimatesusingthesamerandomsampleofmonths.thisisdone5,000timestoproducea distributionofalphasateachpointintherankingdistribution(undertheassumptionofzero alphas).thedistributionforthetopfundisconstructedasthedistributionofthemaximum alphageneratedacrossallbootstraps,thedistributionforthesecondbestfundasthesecond bestalphaacrossallbootstraps,andsoon. Notethatinagivensimulationrun,fundswillhavethesamerandomsampleofmonths aswhentheyoverlapinexistence.thishastheadvantagethatthesimulationswillcaptureany crossjcorrelationinfundreturns.inunreportedresults,weseethatcrossjcorrelationexistsand isimportant theconfidenceintervalswiden.thejointsamplingoffundandfactorreturnshas theaddedadvantageofcapturinganyheteroscedasticityoftheexplanatoryreturnsandthe disturbancesofthebenchmarkmodel.samplingthesamemonthsforallfundshasthe disadvantagethatafundmayshowupinasimulationrunformoreorlessthanthenumberof monthsithasactuallyexistedsinceanygivenmonthmaybesampledmorethanonceornever. Presumably,overJsamplingofsomefundswillbebalancedbyunderJsamplingofothersineach simulationandoverthe5,000runsusedtomakeinferences.thereisacaveat,however.we discardfundswithareturnhistoryoflessthan36months,withtheresultthatwemayendup withabitmoreoverjthanunderjsampling Effectively,forthe2002J2013period,forwhichwehave124activelymanagedfunds,4,095bootstrapsamples containthisnumberoffunds.869bootstrapsamplescontain123activelymanagedfunds.only36bootstrap samplescontain122fundsorfewer.inthecaseswhentherearefewerthen124fundsinthebootstrapsample, weusetheconfidenceintervaloftheclosestfundpercentilejwiseintherankingdistribution,toformthe confidenceintervalbasedon5,000bootstrapsforallofthe124funds. 11
13 Therearetwomorecaveats.Usingthesamemonth sreturnforallfundsinasimulation runpreservesthecrossjcorrelationoffundreturns,buteliminatesanyeffectsof autocorrelation.thisseemstobeaminorproblem,seee.g.fama(1965)orkosowskiet.al. (2006).Also,sincewerandomlysamplemonths,weloseanyeffectsoftimevariationinthe regressionslopesof(1).fersonandschadt(1996)arguethattimevariationshouldbeallowed, butinunreportedresultsweseeverysmalleffectsonstandarderrorswhenweallowfor NeweyJWestcorrection. WearenowreadytoturntosimulationJbasedresults,andstartwith bootstrappeddistributionsofgrossexcessreturns(grossalphas).figure4shows90percent confidencebands(fivepercentoneachside)aroundthebootstrappedgross1jfactoralpha distributionandaroundthetjvaluesofthebootstrapped1jfactoralphadistributionforthesub period Itshouldbepointedoutthattheconfidencebandsarebasednotonan assumedparametricformofthedistributions,butonactualcountsofpointestimatesoverthe 5,000simulations. 18 Estimatesofactualgross1JfactoralphasandtJvaluesofactualgross1J factoralphasarealsoshown(theestimatesofalphaareofcoursethesameasinthesubplot ontherightinfigure2).fundsthatareshortjlivedhavelesspreciseestimatesofexcessreturns estimateswithgreaterstandarderrors thanfundsthatexistthewholetimeperiod.byusing thetjstatisticsoftheestimatesofexcessreturns,onecancontrolfortheeffectofdifferentlifej timesoffundsandthesubsequentprecisionoftheestimates.becauseofthis,tjstatisticsisthe preferredstatisticofbothkosowskiet.al.(2006)andfamaandfrench(2010). [FIGURE4] Ascanbeseen,thereisonlyonesignificantlypositiveactualgross1Jfactoralpha relativetobootstrappedgrossalpha(forthefundrankedtenth)andonlyonesignificant positivegross1jfactoralphawhentestingontjvaluesofactualandbootstrappedgrossalphas (forthefundrankedsecondintermsoftjstatistic).nofundhasanegativeandsignificantgross alpha. Wedrawtheconclusionthatthereisverylittleevidenceoftruestockpickingskills amongmanagersofactivelymanagedswedishequitymutualfundsin Also,there isnoevidenceofinferiorskillsattheotherendofthedistribution. Figure5showstheresultsoftheanalogousexercisefornetexcessreturns(netalphas). Asshouldbeexpectedfromthefindingofverylittleskillinthegrossalphas,nofundisfoundto besufficientlyskilledtocompensateforthecostofactivemanagementandgiveinvestorsa significantpositivenetalphain2002j2013.thisholdsforbothtestsonactualrelativeto bootstrappednetalphasaswellastestsontjvaluesofactualandbootstrappednetalphas. 18 Kosowskietal(2006)stressthattheactualdistributionmaydifferfromthenormaldistribution. 12
14 [FIGURE5] 6.Summary Mostinvestorsandfinancialadvisersseemtobelievethatsomefundmanagerspossesstrue skillinpickingstocks.financialadviserscommonlyrecommendthatinvestorslookatpast returnsoffundsasaguidetochoosewhichfundtoinvestin.thismeansthattheynotonly believeintheexistenceofsuperiorskills,butalsothatskillsarepersistentandthatreturn unadjustedforsystematicriskisagoodmeasureofskills. Weexaminewhetherthesebeliefsarewarrantedbyanalyzingperformanceand persistenceofswedishequitymutualfundsintheperiod1993j2013.performanceismeasured asthegrossandnetreturninexcessofariskjfreeinterestrateandadjustedformarketrisk (beta),i.e.grossandnet1jfactoralpha.grossexcessreturnsarerelevantfordetermining whetherfundmanagershavesuperiororinferiorskills,whileexcessreturnsnetof managementcostsarerelevantforinvestors. Wefindasharpbreakintheperformanceofactivelymanagedfundsaround In ,theaverage1Jfactorgrossalphaofactivelymanagedfundswas3.55per cent,andin itwas 0.18percent.Wearguethatthesuperiorreturnsintheearlier periodwereduetoabiastowardsitjstocksandaninefficientmarket.thenumberofactively managedfundsmorethandoubledbetweenthetwosubperiods.whateverthereasonsforthe lowerlevelofreturnsafter ,inferencesabouttoday sperformance,persistence andthepresenceofstockpickingskillshouldbebasedonthelaterperiod. Onlythetoptwelvefundsobtainedsignificantpositivegrossalphasin (with 90percentconfidence,95percentoneachside),whereasagreaternumberobtained significantnegativegrossalphas.inotherwords,somestockjpickingskillsseemtoexistamong thetopfundmanagers.italsoappearsthatalargernumberoffundmanagersatthebottom endofthedistributionhaveinferiorskillsorinefficientmanagementaccordingtoconventional tests. Welackdatatoadjustforadditionalsystematicriskfactorslaterthan2009.3Jand4J factorgrossalphasmaydifferfrom1jfactoralphasandleadtosubstantiallydifferent conclusionsaboutfundperformance.however,acomparisonof1j,3jand4jfactoralphasin 1999J2009doesnotrevealanylargedifferences. Wealsoexaminetheperformanceof20indexfundsbasedontheirownbenchmarks. Wefindthatindexfundshaveanegativetrackingerroronaveragein2002J2013,andthata 13
15 handfuloffundshaveatrackingerrorof 0.61percentperyearormore.Itisclearthat investorsshouldconsidertrackingerrorandfeeswhenchoosinganindexfund. Thefivetopdecilesofactivefundshavepositivegrossalphasin1993J2013andthefour topdecileshavepositivegrossalphasin2002j2013.ifthisisduetosuperiorstockpickingskills theyshouldexhibitpersistencyintheirreturns.persistenceisexaminedbyrankingfundson threejyearperformanceandevaluatingperformanceduringthefollowingonejtofivejyear periods.wefindthatperformanceconvergestothemeanintwoyears,andstaysclosetothe meansubsequently.thisholdsbothfordecilesandfortopandbottomfivefunds.hence,there ispracticallynoevidenceofpersistenceandthereforeofskill. Wealsoexaminethepresenceofskillbycomparingactualreturnsofactivelymanaged fundstobootstrappedreturns.thebootstrappedreturnsaregeneratedundertheassumption ofzeroalpha.thebootstrapyieldsadistributionofpseudojexcessgrossandnetreturnsfor eachfund.wetestwhethertheactualexcessreturnofeachfundissignificantlydifferentfrom thesimulatedexcessreturnin2002j2013.wefindthatonefund(rankedinsecondplace)hasa significantgrossexcessreturnatthe(twojsided)90percentlevelwhentestingonthetjvalues ofalphaestimates.similarly,onefund(rankedintenthplace)hasasignificantgrossexcess returnwhentestingonthealphaestimatesdirectly.nofundisabletogiveinvestorssignificant apositivenetexcessreturn.lookingatnegativegrossexcessreturnsin2002j2013,ninefunds closetothebottomhavesignificantlyinferiorgrossexcessreturnswith90percentconfidence whentestingonthetjvaluesofthealphaestimates,butnotwhentestingonthealpha estimatesdirectly. Inconclusion,thereisverylittleevidenceoftruestockpickingskillamongmanagersof Swedishequitymutualfundsandnoevidencethatfundscangiveinvestorspositivenetreturns. 14
16 References( Betermier,Sebastien,LaurentCalvetandPaoloSodini,2013,WhoaretheValueand GrowthInvestors?,mimeo Berk,JonathanB.andJulesH.vanBinsbergen,2013,Measuringskillinthemutualfund industry,manuscriptdatednovember4. Carhart,MarkM.,1997,Onpersistenceinmutualfundperformance,Journal$of$Finance 52,57J82. Cuthbertson,Keith,DirkNietzscheandNiallO Sullivan,2008,Journal$of$Empirical$ Finance15,613J634. Dahlquist,Magnus,StefanEngströmandPaulSöderlind,2000,Performanceand characteristicsofswedishmutualfunds,journal$of$financial$and$quantitative$analysis35,409j 423. Engström,Stefan,2004,Investmentstrategies,fundperformanceandportfolio characteristics,sse/efi Working Paper Series No Fama,EugeneF.,1965,Thebehaviorofstockmarketprices,Journal$of$Business38,34J Fama,EugeneF.andKennethR.French,1993,Commonriskfactorsinthereturnson stocksandbonds,journal$of$financial$economics33,3j56. Fama,EugeneF.andKennethR.French,2010,LuckversusskillinthecrossJsectionof mutualfundreturns,journal$of$finance65,1915j1947. Ferson,WayneE.andRudiW.Schadt,1996,Measuringfundstrategyandperformance inchangingeconomicconditions,journal$of$finance51,425j462. Kosowski,Robert,AllanTimmerman,RussWermers,andHalWhite,2006,Canmutual fund stars reallypickstocks?newevidencefromabootstrapanalysis,journal$of$finance61, 2551J
17 Table(1.((Fund(and(factor(returns( Panel(A.(Equally6weighted(portfolio(returns( ( ( 1993J J2013 No.of No.of Portfolio funds Mean Std funds Mean Std Allfunds ,72% 20,21% 153 9,11% 19,69% Activelymanaged funds 95 21,75% 20,16% 133 9,13% 19,72% Indexfunds 8 21,15% 21,69% 20 8,81% 19,88% SIXPRX J 19,85% 19,78% J 10,47% 19,50% STIBOR1M J 5,68% 0,62% J 2,41% 0,37% Panel(B.(Factor(returns( ( ( 2002J2009 Factor Mean Std SIXPRX 9,0% 21,5% STIBOR1M 2,9% 0,4% SMB 1,3% 20,3% HML 17,3% 29,2% MOM J2,0% 38,2% Note:Fundreturnsareequallyweightedmonthlyreturns.Allreturnstatisticshavebeen convertedfrommonthlytoannualfrequency. 16
18 Table(2.((Fees(and(expenses( Actively(managed(funds( Index(funds( No.offunds Mean Std No.offunds Mean Std Annualfee 115 1,3% 0,4% 17 0,5% 0,2% Entryfee 21 2,4% 2,1% 2 0,7% 0,4% Exitfee 48 1,2% 1,5% 5 0,7% 0,4% Performancefee 8 18,1% 6,9% 0 J J TER 124 1,3% 0,4% 19 0,46% 0,20% TKA 114 1,6% 0,6% 13 0,5% 0,3% Othercosts 50 0,2% 0,2% 4 0,0% 0,1% Notes:TERisdefinedasxxx.TKAisdefinedasxxx.Forfeesnotchargedorexpenseitemsnotcharged, observationshasbeencodedasmissingratherthanzero. Note:117activefundswithTERdataand19indexfundswithTERdataiffilepassive_smallisused.If returns_morningstar_sthlm_moneymateisused,thereare124activefundswithterdataand19index funds. 17
19 Table(3.((16factor(alphas(for(actively(managed(funds( ( ( ( ( Portfolio ( Net Gross Net Gross Net Gross Allfunds 2,14% 3,55% J1,47% J0,18% J0,91% 0,44% 1stdecile 9,72% 10,8% 3,83% 5,04% 4,52% 5,55% 2nddecile 5,81% 7,11% 1,29% 2,68% 1,93% 3,34% 3rddecile 4,25% 5,62% J0,14% 1,25% 0,61% 2,07% 4thdecile 3,16% 4,53% J0,83% 0,59% J0,11% 1,22% 5thdecile 2,20% 3,74% J1,38% J0,19% J0,80% 0,66% 6thdecile 1,04% 2,75% J1,91% J0,66% J1,21% 0,15% 7thdecile 0,38% 1,89% J2,29% J1,00% J1,77% J0,40% 8thdecile J0,10% 1,35% J2,96% J1,83% J2,61% J1,02% 9thdecile J1,09% 0,60% J4,23% J2,88% J3,39% J2,04% 10thdecile J3,97% J1,84% J5,71% J4,40% J5,90% J4,76% 1stfund 13,6% 14,2% 5,42% 7,14% 7,10% 8,85% 2ndfund 12,9% 14,0% 5,39% 5,94% 7,05% 6,52% 3rdfund 10,1% 8,81% 4,51% 5,65% 5,28% 6,15% 4thfund 7,39% 8,59% 4,46% 5,22% 4,65% 5,94% 5thfund 7,28% 8,29% 4,24% 4,90% 4,51% 5,35% 5thfromlastfund J2,52% J0,05% J5,47% J4,12% J4,85% J3,92% 4thfromlastfund J2,53% J0,17% J5,58% J4,21% J5,39% J4,40% 3rdfromlastfund J4,10% J1,10% J6,76% J4,62% J7,09% J4,53% 2ndfromlast fund J5,82% J4,42% J7,23% J6,40% J8,26% J6,65% Lastfund J7,01% J5,40% J8,22% J6,61% J14,7% J13,2% Notes:The1993J2001netsamplecomprisesof60fundsandthegrosssampeof56funds. The2002J2013net(gross)samplecomprisesof124(113)funds.The1993J2013net(gross) samplecomprisesof128(117)funds.thenumberoffundsineachdecileisroundedto nearestinteger(e.g.,either5or6inthe1993j2001grosssample). 18
20 Table(4.(((Benchmark6based(16factor(net(and(gross(alphas(for(index(funds( ( ( ( ( Fund ( Net Gross Net Gross Net Gross Allfunds 0,00% 0,54% J0,84% J0,22% J0,72% J0,09% 1stfund 2,40% 3,13% 0,10% 0,10% 0,49% 0,98% 2ndfund 0,18% 0,92% J0,17% 0,14% 0,10% 0,10% 3rdfund J0,69% J0,15% J0,21% 0,09% J0,12% 0,20% 4thfund J0,77% J0,46% J0,29% 0,02% J0,25% 0,05% 5thfund J1,12% J0,72% J0,36% J0,05% J0,25% 0,48% 6thfund J J J0,47% 0,03% J0,31% J0,01% 7thfund J J J0,49% J0,09% J0,36% J0,05% 8thfund J J J0,53% J0,04% J0,42% J0,02% 9thfund J J J0,57% J0,03% J0,47% 0,03% 10thfund J J J0,69% J0,29% J0,48% 0,06% 11thfund J J J0,74% J0,44% J0,59% 0,15% 12thfund J J J0,76% J0,26% J0,69% J0,29% 13thfund J J J0,79% J0,35% J0,74% J0,44% 14thfund J J J0,95% J0,22% J0,76% J0,26% 15thfund J J J1,01% J0,61% J0,79% J0,35% 16thfund J J J1,11% J0,38% J1,01% J0,61% 17thfund J J J1,17% J0,37% J1,17% J0,37% 18thfund J J J1,25% J0,73% J1,25% J0,73% 19thfund J J J1,53% J0,70% J1,53% J0,70% 20thfund J J J3,82% J J3,82% J Note:Foreachofthethreetimeperiodsthefundsarerankedonnetalpha.Sevenfundshave aversionofomxs30asbenchmarkindex,anothersevenreportaversionofomxsbas benchmark,andthreefundsreportsixrxasitsbenchmark.inaddition,onefundeach reportssix30rx,omxafgxandsix/gesethicalswedenasitsrespectivebenchmark.we allowforminoradjustmentstosuitableproxiesforeachfund'sbenchmarkindexsothatwe defactousethreeindicesintheestimation;sixprx,six30rxandomxsbcapgi.since SIX30RXdoesnotexistpriorto2000,weloseobservationsforthreeindexfundsthatexisted inthe1990sandthathaveomxs30asbenchmark. 19
21 Table&5.&Actual&and&bootstrapped&t3statistics,&13factor&model,&gross&returns&of&actively&managed&funds& & & & 1993&3&2001& & 2002&3&2013& & Actual t Bootstrappedt Actual < Bootstrapped Actual t Bootstrappedt Actual < Bootstrapped 1stdecile 3,73 1,51 0,04% 2,15 1,51 11,1% 2nddecile 2,62 0,89 0,06% 1,16 0,91 29,6% 3rddecile 2,08 0,58 0,14% 0,69 0,60 41,7% 4thdecile 1,66 0,32 0,36% 0,35 0,33 48,8% 5thdecile 1,43 0,08 0,09 27,8% 6thdecile 0,70% 17,2% 7thdecile 0,82% 15,7% 8thdecile 0,88% 10,5% 9thdecile 0,56% @1,57 4,08% 1stfund 4,73 1,95 0,00% 3,12 2,25 8,28% 2ndfund 4,22 1,78 0,12% 3,10 1,94 2,96% 3rdfund 3,38 1,53 0,28% 2,53 1,76 7,96% 4thfund 3,37 1,37 0,08% 2,30 1,64 10,4% 5thfund 3,35 1,25 0,02% 2,09 1,54 @1,57 @1,66 @1,79 @1,96 @2,28 15,7% Note:[email protected]@2013samplecovers113funds.Thethirdandsixthcolumnreportstheone@sided bootstrappedconfidenceintervalbasedon5000bootstraps.
22 Table&A1.&&1*,&3*&and&4*factor&alphas&for&actively&managed&funds&(1999*2009)& 1*factor& 3*factor& 4*factor& & Portfolio & Net Gross Net Gross Net Gross Allfunds 10,48% 0,84% 10,30% 1,03% 10,27% 1,09% 1stdecile 6,16% 7,02% 6,02% 6,93% 6,63% 7,84% 2nddecile 3,15% 4,61% 3,03% 4,51% 3,32% 4,73% 3rddecile 1,45% 2,97% 1,68% 3,10% 1,68% 3,15% 4thdecile 0,22% 1,57% 0,37% 1,77% 0,24% 1,55% 5thdecile 10,47% 0,86% 10,32% 0,99% 10,33% 0,84% 6thdecile 11,07% 0,32% 10,93% 0,43% 10,83% 0,51% 7thdecile 11,64% 10,26% 11,54% 10,14% 11,34% 0,06% 8thdecile 12,32% 10,91% 12,20% 10,80% 11,93% 10,47% 9thdecile 13,73% 12,39% 13,20% 11,73% 13,60% 11,91% 10thdecile 16,01% 14,85% 15,42% 14,25% 16,04% 14,85% 1stfund 10,3% 10,41% 9,89% 9,80% 13,6% 15,2% 2ndfund 8,67% 9,89% 8,05% 9,46% 8,13% 9,88% 3rdfund 8,39% 9,14% 7,96% 8,92% 8,00% 8,49% 4thfund 7,64% 6,96% 7,42% 6,79% 6,99% 7,88% 5thfund 5,48% 6,28% 5,65% 6,15% 6,38% 6,97% 5thfromlastfund 15,39% 14,40% 14,85% 13,73% 15,24% 14,11% 4thfromlastfund 16,03% 14,53% 15,37% 14,51% 15,48% 14,62% 3rdfromlastfund 16,20% 14,83% 15,94% 14,84% 16,43% 15,31% 2ndfromlast fund 17,09% 15,10% 17,86% 15,17% 18,00% 15,34% Lastfund 114,7% 113,17% 111,1% 19,61% 114,6% 113,1% Notes:Thenetsampleconsistsof117fundsandthegrosssampleof107funds.
23 Figure1 1
24 Figure2 2
25 Figure3a) 3
26 Figure3b) 4
27 Figure4 5
28 Figure5 6
29 FigureA1 7
! 2!!$ ,)!$- %$0. Baskı-2 ! "! #$ % #$#!&'! '! (&&)!! &!! #.! &)!$#$! /&)!!! 0! &)!$!.!! 0$! #! &)!$ &.!!#$!! 3!&!#!!3! #&!'! &! 4!!
" $ % $&' ' (&&) & )*,)$-.&&) &. &)$$ /&) 0 &)$. 0$ &)$ + 2$,)$3&) &.$ 3& 3 &' & 43 '' %$ / %$0 (%(%3 ' '& 4& 40%3 0$& (% 3 *& 0&3$ 5 %40% 4 4 4 7 8&, 40% :&&* 6 9 4-7 "& % 4 )$ 4 & &)$, %&$ ; 8&7&4 3
. ก () ก!"#ก!$%& ' ()$* ก!"#+'#$%&, '$ -+.ก NOMURA ifund!%! :;< %=)),''ก ' ;' +'#%+ >!"#+'#$%&?, %&ก 22. '#<'!$%=)=#'
ก NOMURA ifund C API T AL NOM UR A SECURITI ES. ก () ก!"#ก!$%& ' ()$* ก!"#+'#$%&, '$ -+.ก NOMURA ifund!%! :;< %=)),''ก ' ;' +'#%+ >!"#+'#$%&?, %&ก 22. '#!-"# ก!$%& ', %&ก 22.
No More Weekend Effect
No More Weekend Effect Russell P. Robins 1 and Geoffrey Peter Smith 2 1 AB Freeman School of Business, Tulane University 2 WP Carey School of Business, Arizona State University Abstract Before 1975, the
DISCUSSION PAPER PI-1404
DISCUSSION PAPER PI-1404 New Evidence on Mutual Fund Performance: A Comparison of Alternative Bootstrap Methods. David Blake, Tristan Caulfield, Christos Ioannidis & Iain Tonks. October 2015 ISSN 1367-580X
Luck versus Skill in the Cross-Section of Mutual Fund Returns
THE JOURNAL OF FINANCE VOL. LXV, NO. 5 OCTOBER 2010 Luck versus Skill in the Cross-Section of Mutual Fund Returns EUGENE F. FAMA and KENNETH R. FRENCH ABSTRACT The aggregate portfolio of actively managed
Luck versus Skill in the Cross Section of Mutual Fund Returns. Eugene F. Fama and Kenneth R. French * Forthcoming in the Journal of Finance.
First draft: October 2007 This draft: December 2009 Not for quotation: Comments welcome Luck versus Skill in the Cross Section of Mutual Fund Returns Eugene F. Fama and Kenneth R. French * Forthcoming
Chicago Booth BUSINESS STATISTICS 41000 Final Exam Fall 2011
Chicago Booth BUSINESS STATISTICS 41000 Final Exam Fall 2011 Name: Section: I pledge my honor that I have not violated the Honor Code Signature: This exam has 34 pages. You have 3 hours to complete this
The Equity Risk Premium, the Liquidity Premium, and Other Market Premiums. What is the Equity Risk Premium?
The Equity Risk, the, and Other Market s Roger G. Ibbotson Professor, Yale School of Management Canadian Investment Review Investment Innovation Conference Bermuda November 2011 1 What is the Equity Risk?
74).20;51/;),20)4)+/; %#&KFF!`!# MMMFFH=MF 9145)4+)) 674-,951! 9;- ",7)! 29) ),)+;9) 7) )*-4,95).4-0.-, # 9*1-)1 6*4951 *)5)),2624),1)2)5)-8-.2;;/04-1 +0-+;5611/)5641),17119-9160,-4)6-),4*1,*-516;),-6)*1+5;,4-,AF=HJAJB2DOIECO=>HAA@E?=7ELAHIEJOB5EAIE==JME?A2=@,AF=HJAJB2DOIECO=CEAE=7ELAHIEJOA@E?=+=CA+H=?M2=@!,AF=HJAJB>IJAJHE?=@/OA?CO4K@=5=I=A@E?=7ELAHIEJOB5EAIE==JME?A
Illiquidity frictions and asset pricing anomalies
Illiquidity frictions and asset pricing anomalies Björn Hagströmer a, Björn Hansson b, Birger Nilsson,b a Stockholm University, School of Business, S-10691 Stockholm, Sweden b Department of Economics and
Performance of UK Pension Funds. - Luck or Skill?
Performance of UK Pension Funds - Luck or Skill? Emelie Jomer Master Thesis, Department of Economics, Uppsala University June 7, 2013 Supervisor: Mikael Bask, Associate Professor of Economics, Uppsala
How to assess the risk of a large portfolio? How to estimate a large covariance matrix?
Chapter 3 Sparse Portfolio Allocation This chapter touches some practical aspects of portfolio allocation and risk assessment from a large pool of financial assets (e.g. stocks) How to assess the risk
How To Explain Momentum Anomaly In International Equity Market
Does the alternative three-factor model explain momentum anomaly better in G12 countries? Steve Fan University of Wisconsin Whitewater Linda Yu University of Wisconsin Whitewater ABSTRACT This study constructs
Financial Intermediaries and the Cross-Section of Asset Returns
Financial Intermediaries and the Cross-Section of Asset Returns Tobias Adrian - Federal Reserve Bank of New York 1 Erkko Etula - Goldman Sachs Tyler Muir - Kellogg School of Management May, 2012 1 The
When is a Total Expense Ratio not a Total Expense Ratio
Frontier Investment Management LLP Research When is a Total Expense Ratio not a Total Expense Ratio It is well known that minimising costs is one of the keys to achieving superior investment returns. Investors
Commack UFSD Remote Access for Microsoft Windows Vista, 7 and 8 Apple Macs, ipads, iphones And Android devices
Commack UFSD Remote Access for Microsoft Windows Vista, 7 and 8 Apple Macs, ipads, iphones And Android devices This allows district staff and students to access the Commack UFSD network from home or anywhere
IAPRI Quantitative Analysis Capacity Building Series. Multiple regression analysis & interpreting results
IAPRI Quantitative Analysis Capacity Building Series Multiple regression analysis & interpreting results How important is R-squared? R-squared Published in Agricultural Economics 0.45 Best article of the
Mutual Fund Performance
Mutual Fund Performance When measured before expenses passive investors who simply hold the market portfolio must earn zero abnormal returns. This means that active investors as a group must also earn
Internet Appendix for When is a Liability not a Liability? Textual Analysis, Dictionaries, and 10-Ks * Tim Loughran and Bill McDonald
Internet Appendix for When is a Liability not a Liability? Textual Analysis, Dictionaries, and 10-Ks * Tim Loughran and Bill McDonald In the Internet Appendix we provide a detailed description of the parsing
Extracting Portable Alphas From Equity Long-Short Hedge Funds. William Fung* David A. Hsieh**
Forthcoming, Journal of Investment Management Extracting Portable Alphas From Equity Long-Short Hedge Funds By William Fung* David A. Hsieh** *Centre for Hedge Fund Research and Education, London Business
Mutual Fund Performance: Luck or Skill?
INTERNATIONAL JOURNAL OF BUSINESS, 20(1), 2015 ISSN: 1083-4346 Mutual Fund Performance: Luck or Skill? Ajay Bhootra a, Zvi Drezner b, Christopher Schwarz c, Mark Hoven Stohs d a Mihaylo College of Business
Statistical Confidence Calculations
Statistical Confidence Calculations Statistical Methodology Omniture Test&Target utilizes standard statistics to calculate confidence, confidence intervals, and lift for each campaign. The student s T
Tilted Portfolios, Hedge Funds, and Portable Alpha
MAY 2006 Tilted Portfolios, Hedge Funds, and Portable Alpha EUGENE F. FAMA AND KENNETH R. FRENCH Many of Dimensional Fund Advisors clients tilt their portfolios toward small and value stocks. Relative
Why are Some Diversified U.S. Equity Funds Less Diversified Than Others? A Study on the Industry Concentration of Mutual Funds
Why are Some Diversified U.S. Equity unds Less Diversified Than Others? A Study on the Industry Concentration of Mutual unds Binying Liu Advisor: Matthew C. Harding Department of Economics Stanford University
Structured Cash Flows Buyer s Guide
Structured Cash Flows Buyer s Guide Predictable Income Fixed Returns Flexible Terms ThemarketforStructuredCashFlows Sincetheexistenceofanindividual sabilitytoreceivea incomestreamhasalsobeentheabilitytosellthatincome
Fund Performance and Top Management Turnover: Evidence from the UK Unit Trust Industry
Fund Performance and Top Management Turnover: Evidence from the UK Unit Trust Industry Zhichao Zhang a, Li Ding a, Si Zhou a a School of Economics, Finance and Business, Durham University Corresponding
Just Unlucky? A Bootstrapping Simulation to Measure Skill in Individual Investors Investment Performance *
Just Unlucky? A Bootstrapping Simulation to Measure Skill in Individual Investors Investment Performance * Steffen Meyer 1, Dennis Schmoltzi, Christian Stammschulte, Simon Kaesler, Benjamin Loos, Andreas
LAZARD GLOBAL ACTIVE FUNDS PUBLIC LIMITED COMPANY LAZARD PAN EUROPEAN EQUITY FUND. SIMPLIFIED PROSPECTUS DATE 27 August 2009
LAZARD GLOBAL ACTIVE FUNDS PUBLIC LIMITED COMPANY LAZARD PAN EUROPEAN EQUITY FUND SIMPLIFIED PROSPECTUS DATE 27 August 2009 This Simplified Prospectus contains key information in relation to the Lazard
M1 in Economics and Economics and Statistics Applied multivariate Analysis - Big data analytics Worksheet 1 - Bootstrap
Nathalie Villa-Vialanei Année 2015/2016 M1 in Economics and Economics and Statistics Applied multivariate Analsis - Big data analtics Worksheet 1 - Bootstrap This worksheet illustrates the use of nonparametric
Volatility and Premiums in US Equity Returns. Eugene F. Fama and Kenneth R. French *
Volatility and Premiums in US Equity Returns Eugene F. Fama and Kenneth R. French * Understanding volatility is crucial for informed investment decisions. This paper explores the volatility of the market,
Fractionally integrated data and the autodistributed lag model: results from a simulation study
Fractionally integrated data and the autodistributed lag model: results from a simulation study Justin Esarey July 1, 215 Abstract Two contributions in this issue, Grant and Lebo (215) and Keele, Linn
Sector Fund Performance: Analysis of Cash Flow Volatility and Returns
Sector Fund Performance: Analysis of Cash Flow Volatility and Returns Ashish TIWARI * and Anand M. VIJH ** ABSTRACT Sector funds are an important and growing segment of the mutual fund industry. This paper
Risk and return in Þxed income arbitrage: Nickels in front of a steamroller?
Risk and return in Þxed income arbitrage Université d Evry June 2005 1 Risk and return in Þxed income arbitrage: Nickels in front of a steamroller? Jefferson Duarte University of Washington Francis Longstaff
Gressett Termite Control, Inc.
1164 WOOD DESTROYING PESTS AND ORGANISMS INSPECTION REPORT Building No. Street City Zip Ordered by: MIRCHA PANDURU [email protected] DOVER LANE FOSTER CITY 94404 5/16/2011 5 14393 East 14th Street Suite
ASSESSING FINANCIAL EDUCATION: EVIDENCE FROM BOOTCAMP. William Skimmyhorn. Online Appendix
ASSESSING FINANCIAL EDUCATION: EVIDENCE FROM BOOTCAMP William Skimmyhorn Online Appendix Appendix Table 1. Treatment Variable Imputation Procedure Step Description Percent 1 Using administrative data,
Private Information in the Chinese Stock Market: Evidence from Mutual Funds and Corporate Insiders *
Private Information in the Chinese Stock Market: Evidence from Mutual Funds and Corporate Insiders * Yeguang Chi October 1, 2014 ABSTRACT I find evidence of valuable private information in the Chinese
Discussion of "The Cross Section and Time Series of Stock and Bond Returns" by Koijen, Lustig & Van Nieuwerburgh
Discussion of "The Cross Section and Time Series of Stock and Bond Returns" by Koijen, Lustig & Van Nieuwerburgh Monika Piazzesi Stanford University & NBER AFA Atlanta 2010 Summary A ne model in which:
L13: cross-validation
Resampling methods Cross validation Bootstrap L13: cross-validation Bias and variance estimation with the Bootstrap Three-way data partitioning CSCE 666 Pattern Analysis Ricardo Gutierrez-Osuna CSE@TAMU
Smart ESG Integration: Factoring in Sustainability
Smart ESG Integration: Factoring in Sustainability Abstract Smart ESG integration is an advanced ESG integration method developed by RobecoSAM s Quantitative Research team. In a first step, an improved
Cyber-Security Analysis of State Estimators in Power Systems
Cyber-Security Analysis of State Estimators in Electric Power Systems André Teixeira 1, Saurabh Amin 2, Henrik Sandberg 1, Karl H. Johansson 1, and Shankar Sastry 2 ACCESS Linnaeus Centre, KTH-Royal Institute
Clustered Standard Errors
Clustered Standard Errors 1. The Attraction of Differences in Differences 2. Grouped Errors Across Individuals 3. Serially Correlated Errors 1. The Attraction of Differences in Differences Estimates Typically
New Zealand mutual funds: measuring performance and persistence in performance
Accounting and Finance 46 (2006) 347 363 New Zealand mutual funds: measuring performance and persistence in performance Rob Bauer a,rogér Otten b, Alireza Tourani Rad c a ABP Investments and Limburg Institute
SEI s Approach to Asset Allocation
SEI s Approach to Asset Allocation Presented by: Jim Smigiel Managing Director and Portfolio Manager Portfolio Strategies Group What is diversification? Sharpe ratio? Peak Sharpe Ratio Loss of efficiency:
A Multivariate Test For Similarity of Two Dissolution Profiles
A Multivariate Test For Similarity of Two Dissolution Profiles * H. Saranadasa and K. Krishnamoorthy Ortho McNeil Pharmaceutical, Inc., Raritan, New Jersey, USA University of Louisiana at Lafayette, Lafayette,
Web Application Security
Web Application Security John Zaharopoulos ITS - Security 10/9/2012 1 Web App Security Trends Web 2.0 Dynamic Webpages Growth of Ajax / Client side Javascript Hardening of OSes Secure by default Auto-patching
APPENDIX 7.4. An updated estimate of the required return on equity
APPENDIX 7.4 An updated estimate of the required return on equity Energex revised regulatory proposal July 2015 An updated estimate of the required return on equity REPORT PREPARED FOR ENERGEX June 2015
Sales forecasting # 2
Sales forecasting # 2 Arthur Charpentier [email protected] 1 Agenda Qualitative and quantitative methods, a very general introduction Series decomposition Short versus long term forecasting
Mutual fund attributes and their relationship to riskadjusted return: A study on the performance and characteristics on the Swedish fund market
STOCKHOLM SCHOOL OF ECONOMICS Master Thesis in Finance Mutual fund attributes and their relationship to riskadjusted return: A study on the performance and characteristics on the Swedish fund market Johan
SPECIAL RECRUITMENT DRIVE FOR PERSONS WITH DISABILITIES (PWD) Category of Disabilit y. Reservati on Category 1 OH-OL UR 6370-17620
HINDUSTAN INSECTICIDES LIMITED (A Govt. of India Enterprise) P.O UDYOGAMANDAL, Dist. Ernakulam, Kochi, (Kerala) Ph : 08 255121 23, Web: www.hil.gov.in Website HIL, Central Public Sector Company engaged
Low-Volatility Investing: Expect the Unexpected
WHITE PAPER October 2014 For professional investors Low-Volatility Investing: Expect the Unexpected David Blitz, PhD Pim van Vliet, PhD Low-Volatility Investing: Expect the Unexpected 1 Expect the unexpected
!"#$%&'()*+,-$..)/)0+1'$+ 23*'4/)%"+2'$#/-+2).($5&4&
!"#$%&'()*+,-$..)/)0+1'$+ 23*'4/)%"+2'$#/-+2).($5&4& 67.$7/.8!"#$"%&'()*#+)#$"#$,"-'*$.#,&./%#-'#0,$..--"+"/1#+"#23%#4$..#-5#$"#$'2%'(6#7-)2#$"%&'()*)#-88&'#+"#23%# $-'2$9#23%#*$+"#$'2%'(#23$2#8$''+%)#,.--:#5'-*#23%#3%$'2#2-#23%#'%)2#-5#23%#,-:(6#!"#$"%&'()*#23$2#-88&')#
AGENCY CONFLICTS IN DELEGATED PORTFOLIO MANAGEMENT: EVIDENCE FROM NAMESAKE MUTUAL FUNDS. Abstract
The Journal of Financial Research Vol. XXX, No. 4 Pages 473 494 Winter 2007 AGENCY CONFLICTS IN DELEGATED PORTFOLIO MANAGEMENT: EVIDENCE FROM NAMESAKE MUTUAL FUNDS Stephen P. Ferris and Xuemin (Sterling)
The Stock Market s Reaction to Accounting Information: The Case of the Latin American Integrated Market. Abstract
The Stock Market s Reaction to Accounting Information: The Case of the Latin American Integrated Market Abstract The purpose of this paper is to explore the stock market s reaction to quarterly financial
Stock Market -Trading and market participants
Stock Market -Trading and market participants Ruichang LU ( 卢 瑞 昌 ) Department of Finance Guanghua School of Management Peking University Overview Trading Stock Understand trading order Trading cost Margin
Active bond-fund excess returns: Is it alpha... or beta?
Active bond-fund excess returns: Is it alpha... or beta? Vanguard research September 213 Executive summary. Active U.S. bond funds have, on average, performed exceptionally well over the past four years.
Do Implied Volatilities Predict Stock Returns?
Do Implied Volatilities Predict Stock Returns? Manuel Ammann, Michael Verhofen and Stephan Süss University of St. Gallen Abstract Using a complete sample of US equity options, we find a positive, highly
FADE THE GAP: ODDS FAVOR MEAN REVERSION
FADE THE GAP: ODDS FAVOR MEAN REVERSION First Draft: July 2014 This Draft: July 2014 Jia-Yuh Chen and Timothy L. Palmer Abstract When a stock opens a day s trading at a lower price than its previous day
Fundamental Analysis and the Cross-Section of Stock Returns: A Data-Mining Approach
Fundamental Analysis and the Cross-Section of Stock Returns: A Data-Mining Approach Xuemin (Sterling) Yan and Lingling Zheng * Abstract A key challenge to evaluate data-mining bias in stock return anomalies
Handling missing data in large data sets. Agostino Di Ciaccio Dept. of Statistics University of Rome La Sapienza
Handling missing data in large data sets Agostino Di Ciaccio Dept. of Statistics University of Rome La Sapienza The problem Often in official statistics we have large data sets with many variables and
Finding outperforming managers. Randolph B. Cohen Harvard Business School
Finding outperforming managers Randolph B. Cohen Harvard Business School 1 Conventional wisdom holds that: Managers can t pick stocks and therefore don t beat the market It s impossible to pick winning
VI. Real Business Cycles Models
VI. Real Business Cycles Models Introduction Business cycle research studies the causes and consequences of the recurrent expansions and contractions in aggregate economic activity that occur in most industrialized
