andmomentumstrategiesinChinastockmarket19

ContrarianandmomentumstrategiesinChinastockmarket:1993-2000JosephKang*,Ming-HuaLiu,XiaoyanNiNanyangBusinessSchool,NanyangTechnologicalUniversity,SingaporeAbstractUsing“A”sharesaccessibleonlytolocalinvestors(whoaccountfor99%ofstockinvestorsinChina),thispapertestsifshort-horizoncontrarianandintermediate-horizoninvestmentstrategiesgenerateabnormalprofits.Wefindstatisticallysignificantabnormalprofitsforboththearbitrageportfolio-investmentstrategies.Detailedanalysisindicatesthat:(1)anabsolutedominanceofnon-institutionalinvestorsleadstoanenvironmentofexcessivespeculationandhenceexcessiveoverreactiontofirm-specificinformation;(2)theoverreactiontofirm-specificinformationisthesinglemostimportantsourceoftheshort-termcontrarianprofit;(3)thestockreturnsintheintermediatehorizonexhibitlaggedoverreactiontocommonfactors;and(4)thelead-lagoverreactiontocommonfactoristhemajorreasonbehindtheintermediate-termmomentumprofit.Thesefindingsarerobust,amongotherthings,tobid-askspreadandnonsynchronoustrading.JELclassification:G14;G15Keywords:Momentumandcontrarianstrategies;Chinastockmarket;overreactionandunderreaction;firm-specificinformation;commonfactor;lead-lagstructure.1.IntroductionAnextensivebodyoffinanceliteraturedocumentsthatpaststockreturnscanpredictthefuturestockreturnsinshort-term,intermediate,andlong-termhorizons,althoughthepredictabilityweakensoverlongerhorizons.Forexample,Jegadeesh(1990)andLehmann(1990)findreturnreversalsinrelativelyshort-termhorizons(onemonthandsixmonths,respectively).JegadeeshandTitman(1993)documentreturncontinuationsinintermediatehorizons(threetotwelvemonths)wherepastwinnerscontinuetooutperform,onaverage,pastlosers.DeBondtandThaler(1985and1987)reportlong-term(e.g.,threetofiveyears)pricereversalswherepastlong-term*Correspondingauthor.Email:[e-mail];Tel[phone];Fax[phone].WewouldliketothankfortheirvaluablecommentsLillianNg,AndrewChen,JosephWilliams,QianSunandJackChenaswellasparticipantsoftheCREFSseminarattheNanyangBusinessSchoolinSingaporeandthe13thPACAP/FMAAnnualFinanceConferenceinSeoul,Korea.Wealsoacknowledgethatanearlierversionofthispaperreceivedthebestpaperawardatthe13thPACAP/FMAAnnualFinanceConference.2losersoutperformpastlong-termwinners.1Givensuchtime-seriespatternsincross-sectionalstockreturns,onecanformulatetwoarbitrageportfolio-investmentstrategies:contrarianandmomentumstrategies.Underthecontrarianstrategy,pastlosersareboughtandpastwinnersareshorted.Underthemomentumstrategy,pastwinnersareboughtandpastlosersareshorted.Abnormalreturnsofthesestrategiesaredocumentedintheliteraturecitedabove.Abnormalprofitsofthemomentumandcontrarianstrategiesarealsodocumentedinnon-USequitymarkets.Forexample,AhmetandNusret(1999)findabnormalreturnsoflong-termcontrarianstrategiesinthestockmarketsofsevennon-USindustrializedcountries.Chang,McLeaveyandRhee(1995)findabnormalreturnsofshort-termcontrarianstrategiesinJapanstockmarket,whereasHameedandTing(2000)findthesameinMalaysiastockmarket.Rouwenhorst(1998)findsmomentumprofitsintwelveEuropeanequitymarkets,whereasRouwenhorst(1999)findsabnormalreturnsofmomentumstrategiesinsix(outoftwenty)emergingequitymarkets.HameedandYuanto(2000)findthata(diversifiedcountry-neutral)momentumstrategygeneratessmallbutstatisticallysignificantreturnsinsixAsianstockmarkets.Schiereck,DeBondtandWeber(1999)findabnormalreturnsforbothintermediate-termmomentumstrategyandshort-andlong-termcontrarianstrategiesinGermanyequitymarket.Fama(1991)notesthatthepredictabilityofstockreturnsovertimeisamongthemostcontroversialissuesonstockmarketefficiency.Thecontroversyledtovariousexplanationsonthepossibilityandthesourcesofabnormalprofitsofcontrarianandmomentumstrategies.Theexplanationsinclude:onebasedonbehavioralirrationalityofinvestorsandtheotherbasedonstockmarketefficiency.1Inthesestudies,losersarethosestockswhosereturnsaresmallerthanmarketindexreturnswhereaswinnersare3Themostfrequently-citedexplanationoftheabnormalprofitsofcontrarianstrategiesis:market’sover-reactiontofirm-specificinformationandthesubsequentcorrection.Forexample,Mun,VasconcellosandKish(1999),andBacmannandDubois(1998)positthatanoverreactiontofirm-specificinformationistheprimaryreasonbehindtheabnormalprofitsofshort-termcontrarianstrategies.DeBondtandThaler(1985)arguethatinvestors’overreactiontorecentpasteventsalsocanleadtolong-termcontrarianprofits.LoandMacKinlay(1990)demonstratethatreturnreversalisnottheonlysourceofcontrarianprofitsandidentifysecondpotentialsourceofcontrarianprofitsthatariseswhensomestocksreactmorequicklytoinformationthanothers.Thealternativesourceofcontrarianprofitsisreferredtoaslead-lagstructurebecausethereturnsofsomestocksleadthereturnsofothers.JegadeeshandTitman(1995)andBoudoukh,etal.(1994)showthatthelead-lagstructureratherarisesfrominvestors’overreactionordelayedreactiontocommonfactorsandalsoshowthatboththeoverreactionandunderreaction(orequivalentlydelayedreaction)ofpricestoinformationcanintheorycontributetocontrarianprofits.Anotherexplanationisthatshort-term(andlong-term)contrarianprofitscanresultfromtime