在交易型开放式基金中的卖空【外文翻译】.doc

1、毕业论文外文翻译外文题目SHORTINTERESTINEXCHANGETRADEDFUNDS出处SWISSSOCIETYFORFINANCIALMARKETRESEARCH作者JEFFMADURATHANHNGO原文SHORTINTERESTINEXCHANGETRADEDFUNDSABSTRACTSHORTSELLINGEXCHANGETRADEDFUNDSETFSHASBECOMEACOMMONMEANSOFSPECULATINGORHEDGINGINRESPONSETOPESSIMISTICEXPECTATIONSABOUTASPECIFICMARKETORSECTOR,ASTHESHO

2、RTINTERESTOFETFSISMORETHAN10TIMESTHATOFINDIVIDUALSTOCKS,ONAVERAGEWEDETERMINETHATSECTORBASEDETFSHAVEANABNORMALLYLARGESHORTINTERESTLEVEL,WHEREASINTERNATIONALETFSHAVEANUNUSUALLYSMALLSHORTINTERESTLEVELTHELEVELOFSHORTINTERESTISLARGERFORETFSTHATHAVEAHIGHERTRADINGVOLUMEANDALOWERMARKETCAPITALIZATION,REGARDL

3、ESSOFTHETYPEOFETFASSESSEDTHELEVELOFSHORTINTERESTISLOWERFORETFSREPRESENTINGINDEXESTHATHAVETRADABLEDERIVATIVES,BUTHIGHERFORINTERNATIONALETFSREPRESENTINGINDEXESTHATHAVETRADABLEDERIVATIVESWEALSODETERMINETHATTHELEVELOFSHORTINTERESTINANETFSERVESASANEFFECTIVESIGNALOFBEARISHSENTIMENTWHENCONSIDERINGALLETFS,B

4、UTISNOTANEFFECTIVESIGNALWHENISOLATINGANYPARTICULARTYPEOFETFKEYWORDSEXCHANGETRADEDFUNDSSHORTSALESSHORTSELLINGIINTRODUCTIONANETFISASPECIALIZEDINVESTMENTTRUSTTHATISCREATEDTOMIRRORASPECIFICEDPORTFOLIOOFSECURITIESMOSTETFSREPRESENTPORTFOLIOSOFSTOCKSANDCANBECLASSIFIEDASBROADBASED,SECTORBASED,ORINTERNATIONA

5、L,DEPENDINGONTHETYPEOFINDEXTHEYMIMICTHEYDIFFERFROMOPENENDINDEXMUTUALFUNDSINTHATTHEYARETRADEDCONTINUOUSLYONANEXCHANGEANDCANBEPURCHASEDORSOLDANYTIMETHEMARKETISOPENTHESEETFSARECREATEDWHENANAUTHORIZEDPARTICIPANTSUCHASASPECIALISTOBTAINSTHEPORTFOLIOOFSTOCKSANDSTORESTHESTOCKSATACUSTODIALBANKINRETURN,THECUS

6、TODIALBANKPROVIDESSHARESOFETFSTOTHEAUTHORIZEDPARTICIPANTTHISPROCESSISREFERREDTOASINKINDTRADINGONCETHEETFSHARESAREPROVIDEDTOTHEAUTHORIZEDPARTICIPANT,THEYCANBETRADEDONSTOCKEXCHANGESJUSTLIKESTOCKSMORERECENTLY,SOMEETFSHAVEBEENDESIGNEDTOMIRRORFIXEDINCOMEASSETSARBITRAGEPREVENTSTHEPRICEOFTHEETFFROMDEVIATIN

7、GSIGNIFICANTLYFROMTHENETASSETVALUEAND,THEREFORE,LIMITSTRACKINGERRORTHEEXPENSESOFETFSASSOCIATEDWITHMANAGINGTHEPORTFOLIOAREVERYLOWETFSCANBESHORTED,JUSTLIKESTOCKSAND,THEREFORE,PROVIDEAMEANSBYWHICHMARKETPARTICIPANTSCANSPECULATEORHEDGEBASEDONPESSIMISTICEXPECTATIONSABOUTASPECIFICMARKETORSECTORHOWEVER,EFTS

8、MUSTBEBORROWEDFROMBROKERAGEFIRMSBEFORETHEYCANBESOLDSHORTINSTITUTIONALINVESTORSHAVEEASIERACCESSTOBORROWINGETFS,ANDALSOTENDTOBEMORECOMMONPLAYERSINTHESHORTSELLINGPROCESSCOLLATERALMUSTBEPOSTEDWITHTHEBROKERAGEFIRM,WHICHCANBEINCASHORTREASURYBILLSTHEMINIMUMAMOUNTOFCOLLATERALREQUIREDIS50OFTHETRANSACTION,BUT

9、SOMEBROKERAGEFIRMSREQUIREAHIGHERPERCENTAGEUNLIKESTOCKS,ETFSCANBESHORTEDONANUPTICKALTHOUGHSHORTSELLERSARENOTNORMALLYCHARGEDANEXPLICITFEEFORSHORTINGETFS,THEYINCURANOPPORTUNITYCOSTWHENPOSTINGCASHASCOLLATERALCLOSEDENDFUNDSCANALSOBESHORTED,BUTTHEYARESUBJECTTOPRICINGDISCREPANCIESBECAUSETHEIRPRICESMAYCONTA

10、INPRONOUNCEDDISCOUNTSORPREMIUMSRELATIVETOTHEIRNETASSETVALUESEEPONTIFF1995,1997THEDISCOUNTSORPREMIUMSCANCHANGEOVERTIMEANDPOSSIBLYOFFSETANYBENEFITSFROMACHANGEINTHENETASSETVALUETHATOCCURSDURINGTHEPERIODDURINGWHICHTHESHARESARESHORTEDINADDITION,THETRADINGVOLUMEOFMANYCLOSEDENDFUNDSISLIMITED,WHICHCOULDMAKE

11、ITCOSTLYTOOFFSETASHORTPOSITIONTHUS,ETFSCANBEMOREEFFECTIVELYUSEDTOBETAGAINSTASPECIFICMARKETORSECTORETFSHARESSOLDREPRESENTABOUT19OFTOTALOUTSTANDINGSHARESONAVERAGE,WHICHISMUCHHIGHERTHANTHELEVELREPORTEDFORINDIVIDUALSTOCKSSHORTINGETFSHASRECENTLYBECOMEVERYPOPULAR,BUTTHEREISALACKOFRESEARCHREGARDINGWHYSHORT

12、POSITIONSVARYSUBSTANTIALLYAMONGETFSORWHETHERTHESHORTPOSITIONCANSERVEASAUSEFULSIGNALFORINVESTORSOUROBJECTIVEISTOIDENTIFYTHECHARACTERISTICSOFETFSTHATATTRACTSHORTSELLERS,ANDTODETERMINEWHETHERTHELEVELOFSHORTINTERESTINETFSSERVESASANEFFECTIVESIGNALOFBEARISHSENTIMENTRESULTSOFOURSTUDYOFFERIMPLICATIONSREGARD

13、INGTHEBEHAVIOROFSHORTSELLERS,HOWSHORTSELLERSIDENTIFYTHEIRTARGETS,ANDWHETHERTHESHORTINTERESTLEVELOFETFSSERVESASANINDICATORFORINVESTORSWEFINDFIRSTTHATSHORTINTERESTISLARGESTFORSECTORBASEDETFS,ANDSMALLESTFORINTERNATIONALETFSSECOND,SHORTINTERESTISLARGERFORETFSTHATHAVEAHIGHERTRADINGVOLUMEANDALOWERMARKETCA

14、PITALIZATIONTHESERESULTSHOLDFORTHEENTIRESAMPLEANDFORSUBSAMPLESSHORTINTERESTISRELATIVELYLOWFORETFSREPRESENTINGINDEXESTHATHAVETRADABLEDERIVATIVES,BUTRELATIVELYHIGHFORINTERNATIONALETFSREPRESENTINGINDEXESTHATHAVETRADABLEDERIVATIVESWEALSOFINDTHATTHELEVELOFSHORTSALESISAUSEFULSIGNALABOUTTHEFUTUREPERFORMANC

15、EOFETFSWHENASSESSINGTHEENTIRESAMPLEHOWEVER,THISISNOTTHECASEWHENTHEANALYSISISPARTITIONEDBYTYPEOFETFIIRELATEDLITERATUREONSHORTINTERESTSTUDIESHAVEATTEMPTEDTOEXPLAINWHYSOMESTOCKSATTRACTMORESHORTSALESTHANOTHERSBRENTETAL1990SHOWTHATINDIVIDUALSTOCKSWITHHIGHBETASANDTRADABLEOPTIONSTENDTOHAVEHIGHERLEVELSOFSHO

16、RTINTEREST,WHICHTHEYARGUE,ISCONSISTENTWITHARBITRAGEEFFORTSDECHOWETAL2001FINDTHATSHORTSELLERSUSUALLYTAKETHEIRPOSITIONSINOVERPRICEDSTOCKSTHATHAVELOWFUNDAMENTALSTOPRICERATIOS,INCLUDINGCASHFLOWTOPRICE,EARNINGSTOPRICE,BOOKTOMARKET,ANDVALUETOMARKET,SINCETHESELOWRATIOSAREPREDICTIVEOFFUTUREDECLINEINSTOCKPRI

17、CESSOMESTUDIESONSHORTSELLINGHAVEASSESSEDTHECHARACTERISTICSOFSTOCKTHATAPPEALTOSHORTSELLERSANGELETAL2003FINDTHATSHORTSELLERSUSUALLYTARGETTHEMOSTVOLATILEANDACTIVELYTRADEDSTOCKSTHEIRSUBSEQUENTSTUDYIN2004ONSHORTSELLINGPRIORTOFIRMSEARNINGSANNOUNCEMENTSFURTHERSHOWSTHATSHORTSELLERSAREMOSTLYINTERESTEDINGROWT

18、HSTOCKSANDSTOCKSWITHBETTERPASTPERFORMANCESTOCKSWITHPOOREARNINGSQUALITYAREALSOATTRACTIVETOSHORTSELLERSSINCEFIRMSWITHAHIGHLEVELOFOPERATINGACCRUALSTENDTOHAVELESSPERSISTENTEARNINGS,ACCRUALSCANBEUSEDTOIDENTIFYMISPRICEDSTOCKSWITHPOOREARNINGSQUALITYDESAIETAL2004FINDASTRONGANDPOSITIVERELATIONSHIPBETWEENTHRE

19、EDIFFERENTMEASURESOFACCRUALSANDCHANGESINSHORTINTERESTMANYOFTHESTUDIESONSHORTSELLINGTESTWHETHERTHEREISARELATIONSHIPBETWEENTHESHORTINTERESTLEVELANDSUBSEQUENTSTOCKPRICEPERFORMANCEAITKENETAL1998FINDTHATHIGHLEVELSOFSHORTINTERESTPRECEDEWEAKPERFORMANCEFORSTOCKSINAUSTRALIADESAIETAL2002FINDTHATHIGHLEVELSOFSH

20、ORTINTERESTPRECEDENEGATIVEABNORMALRETURNSFORSTOCKSTRADEDONTHENASDAQMARKETACKERTANDATHANASSAKOS2005FINDTHATHEAVILYSHORTEDCANADIANSTOCKSPERFORMPOORLY,WHICHSUGGESTSTHATTHESHORTINTERESTLEVELSERVESASANEFFECTIVESIGNALTHEYSUGGESTTHATTHESHORTINTERESTLEVELINTHECANADIANMARKETMAYBEAMOREEFFECTIVEINDICATORTHANTH

21、ESAMEWOULDBEINTHEUSMARKET，BECAUSEOFFEWERSHORTSELLINGRESTRICTIONSINTHECANADIANMARKETCHANGETAL2007ASSESSSHORTSELLINGCONDITIONSINHONGKONGANDFINDTHATSHORTSELLINGCONSTRAINTSCAUSEAGREATERDEGREEOFOVERREACTIONOFINDIVIDUALSTOCKSIIITESTOFRELATIONBETWEENSHORTINTERESTANDSUBSEQUENPERFORMANCEOFTHEETFSTOEXAMINETHE

22、RELATIONSHIPBETWEENSHORTINTERESTANDSUBSEQUENTPERFORMANCEOFTHEETFS,WETESTWHETHERTHETRADINGSTRATEGIESBASEDONTHEPREVIOUSSHORTINTERESTLEVELRESULTINABNORMALGAINSTHETRADINGSTRATEGYINVOLVESTWOSTEPS1FORMINGDECILEPORTFOLIOSBASEDONTHEPREVIOUSSHORTINTERESTLEVEL,AND2ASSESSINGHOWTHEDECILEPORTFOLIOSPERFORMINSUBSE

23、QUENTHOLDINGPERIODSTHEBEGINNINGOFEACHMONTHOFTHESAMPLEPERIODISLABELEDTHEPORTFOLIOFORMATIONMONTHTATTHEBEGINNINGOFEACHPORTFOLIOFORMATIONMONTHT,WEOBTAINTHESHORTINTERESTRATIOOFTHEETFSWEFORM10DECILESOFETFSBASEDONTHESHORTINTERESTLEVELOFTHEETFSTHEDECILEPORTFOLIOSARETHENASSESSEDTODETERMINETHEIRPERFORMANCEOVE

24、RAHOLDINGPERIODSINCETHEREISNOSINGLEHOLDINGPERIODTHATPERFECTLYFITSALLINVESTORS,WEOBTAINTHEABNORMALHOLDINGPERIODRETURNSONTHEETFSINEACHDECILEFORFOURDIFFERENTHOLDINGPERIODS,STARTINGINQUARTER1THEQUARTERIMMEDIATELYFOLLOWINGTHEPORTFOLIOFORMATIONPROCESSANDACCUMULATINGONAQUARTERLYBASISFORTHREEADDITIONALQUART

25、ERSTHUS,OURMETHODOLOGYASSESSESPORTFOLIOSINOVERLAPPINGHOLDINGPERIODS,WHICHISCONSISTENTWITHTHEMETHODUSEDBYJEGADEESHANDTITMAN1993,CHANETAL1996,BAYTASANDCAKICI1999,ANDLEEANDSWAMINATHAN2000JEGADEESHANDTITMAN1993SUGGESTTHATTHEUSEOFOVERLAPPINGPERIODSINCREASESTHEPOWEROFTHESTATISTICALTESTSHOWEVER,WEALSOEMPLO

26、YAMETHODOLOGYTHATASSESSESTHEINVESTMENTSTRATEGIESINNONOVERLAPPINGPERIODSASAROBUSTNESSTESTSPECIFICALLY,ATTHEBEGINNINGOFEACHJANUARYANDJUNE,ALLETFSARECATEGORIZEDINTODECILEPORTFOLIOSBASEDONTHEIRPREVIOUSSHORTSELLINGLEVELTHESEPORTFOLIOSARETHENHELDFORTHENEXT6MONTHS,AFTERWHICHTHEDIFFERENCEBETWEENTHEABNORMALR

27、ETURNOFTHELOWESTANDHIGHESTDECILESISDETERMINEDFOREACHNONOVERLAPPING6MONTHHOLDINGPERIODANDTESTEDFORSIGNIFICANCESINCETHERESULTSFORNONOVERLAPPINGPERIODSAREVERYSIMILARTOTHOSEFOUNDFOROVERLAPPINGPERIODS,THEYARENOTREPORTEDHERETHEABNORMALHOLDINGPERIODRETURNISTHEDIFFERENCEBETWEENTHEETFDECILEPORTFOLIOHOLDINGPE

28、RIODRETURNANDACORRESPONDINGBENCHMARKHOLDINGPERIODRETURNTHEHOLDINGPERIODRETURNSHPRSFORTHEETFSARECALCULATEDONAQUARTERLYCOMPOUNDEDBASISDIVIDENDSAREACCOUNTEDFORINTHECALCULATIONOFQUARTERLYRETURNSMARKETBENCHMARKHOLDINGPERIODRETURNS,MHPR1TOMHPR4,ARECALCULATEDINTHESAMEMANNERMHPRSFORBROADBASEDANDSECTORBASEDE

29、TFSARECALCULATEDUSINGTHECRSPEQUALLYWEIGHTEDINDEXMHPRSFORINTERNATIONALETFSARECALCULATEDUSINGTHEMSCIWORLDINDEXEXTRACTEDFROMTHEDATASTREAMDATABASETHEABNORMALHOLDINGPERIODRETURNAHPRIFOREACHETFISMEASUREDASTHEDIFFERENCEBETWEENHPRIANDTHECORRESPONDINGMHPRIWHEREI1TO4QUARTERSAFTERTTHEAVERAGEABNORMALHOLDINGPERI

30、ODRETURNAAHPRIISCALCULATEDFOREACHDECILEPORTFOLIOBYTHEFOLLOWINGFORMULAWHEREIINDICATESTHENUMBEROFQUARTERSAFTERTI1TO4ANDKINDICATESTHENUMBEROFETFSINEACHDECILEPORTFOLIOK1TONTHEPROFITFROMTHETRADINGSTRATEGYEQUALSTHEDIFFERENCEBETWEENTHEAVERAGEABNORMALHOLDINGPERIODRETURNSOFTHEPORTFOLIOOFETFSREPRESENTINGTHEHI

31、GHESTDECILEANDTHOSEOFTHEPORTFOLIOOFETFSREPRESENTINGTHELOWESTDECILEIVRESULTSOFRELATIONSHIPBETWEENSHORTINTERESTANDPERFORMANCEAMONGETFSRESULTSFROMCREATINGETFPORTFOLIOSACCORDINGTOTHESIZEOFTHESHORTPOSITIONEACHMONTHARESHOWNINTABLE6PANELACOVERSTHEENTIRESAMPLE,PANELSB,C,ANDDCOVERTHESUBSAMPLESFORTHEENTIRESAM

32、PLE,THEDIFFERENCEINAVERAGEABNORMALHOLDINGPERIODRETURNSOFDECILESREPRESENTINGETFSWITHTHELARGESTVERSUSSMALLESTSHORTPOSITIONSISNEGATIVEANDSIGNIFICANTWHENAAHPRIINTHREEANDFOURQUARTERSAFTERTHEPORTFOLIOFORMATIONMONTHARECONSIDEREDTHATIS,THEDECILESWITHTHELARGESTSHORTPOSITIONSATTHETIMEOFPORTFOLIOFORMATIONPERFO

33、RMEDWORSEINTHOSEHOLDINGPERIODSFORTHESAMPLEOFBROADBASEDETFS,THEDIFFERENCEINAVERAGEABNORMALHOLDINGPERIODRETURNSBETWEENTHEDECILESOFTHESMALLESTANDLARGESTETFSISINSIGNIFICANT,REGARDLESSOFTHELENGTHOFTHEHOLDINGPERIOD,WHICHISALSOTRUEFORTHESECTORBASEDANDINTERNATIONALETFPORTFOLIOSTHUS,THESIGNIFICANTDIFFERENCES

34、INPERFORMANCEOFDECILESFOUNDFORTHEENTIRESAMPLEVANISHWHENISOLATINGAPARTICULARTYPEOFETFTHISIMPLIESTHATTHESIGNALFROMASHORTPOSITIONISMOREEFFECTIVEWHENASSESSINGALLETFSASAGROUPTHANWHENFOCUSINGONAPARTICULARTYPEOFETFVCONCLUSIONSHORTSELLINGETFSHASINCREASEDINPOPULARITYINRECENTYEARSANDISNOWEVENMOREPOPULARTHANSH

35、ORTSELLINGSTOCKSWHENMEASUREDINPROPORTIONTOTOTALSHARESOUTSTANDINGWEATTEMPTTODETERMINE1THECHARACTERISTICSTHATMAKESOMEETFSMOREATTRACTIVETHANOTHERSTOSHORTSELLERS,AND2WHETHERTHESHORTSALESLEVELSERVESASAUSEFULSIGNALABOUTFUTUREPERFORMANCEOFETFSOURANALYSISISFOCUSEDONTHEENTIRESAMPLEOFETFS,BUTWEALSOSEPARATELYA

36、SSESSSUBSAMPLESOFETFSCLASSIFIEDASBROADBASED,SECTORBASED,ORINTERNATIONALFIRST,SHORTINTERESTISUNUSUALLYLARGEFORSECTORETFS,ANDUNUSUALLYSMALLFORINTERNATIONALETFSSECOND,SHORTINTERESTISLARGERFORETFSTHATHAVEAHIGHERTRADINGVOLUMETHIRD,SHORTINTERESTISLARGEFORETFSWITHALOWERMARKETCAPITALIZATIONANDALOWEREXPENSER

37、ATIOTHOUGHTHISSPECIFICRELATIONSHIPWITHEXPENSERATIOONLYHOLDSFORTHEENTIRESAMPLETHERESULTSFORTRADINGVOLUMEANDMARKETCAPITALIZATIONHOLDFORTHEENTIRESAMPLEANDFORALLTHREESUBSAMPLESFOURTH,SHORTINTERESTISRELATIVELYLOWFORETFSREPRESENTINGINDEXESTHATHAVETRADABLEDERIVATIVES,BUTRELATIVELYHIGHFORINTERNATIONALETFSRE

38、PRESENTINGINDEXESTHATHAVETRADABLEDERIVATIVESFINALLY,WEFINDTHATTHELEVELOFSHORTSALESISAUSEFULSIGNALABOUTTHEFUTUREPERFORMANCEOFETFSWHENASSESSINGTHEENTIRESAMPLEHOWEVER,ITISNOTAUSEFULSIGNALABOUTTHEFUTUREPERFORMANCEWHENTHEANALYSISISPARTITIONEDBYTYPEOFETF毕业论文外文翻译外文题目SHORTINTERESTINEXCHANGETRADEDFUNDS出处SWIS

39、SSOCIETYFORFINANCIALMARKETRESEARCH作者JEFFMADURATHANHNGO译文在交易型开放式基金中的卖空摘要卖空交易型开放式基金（ETFS）已经成为投机或针对某个特定市场或行业的悲观预期进行对冲的常用手段，因为ETFS的卖空利息超过了个别股票平均水平的10倍。我们确定，以部门为基础的ETFS有一个异常大的卖空利率水平，而国际ETF通常是一个非常小的利息水平。对具有更高交易量和市值较低的ETFS来说，卖空的利息水平是高的，也不需要顾及ETFS类型的评估。代表衍生品交易指数的ETFS的卖空利息水平是低的，但是，代表衍生品交易指数的国际ETFS的卖空利息水平是高的。

40、我们判定，当考虑了所有的ETFS时，它作为了一种悲观情绪的信号；当隔离了任何特定类型的ETFS时，卖空利息不再是一个有效的信号了。关键词交易型开放式基金、买空、卖空一、引言ETF是由专门的投资信托公司建立的证券投资组合。大部分ETFS代表了股票投资组合，依据他们模拟的索引类型不同，可分为广义的、以部门为基础的和国际的。他们不同于开放式基金，因为它们不断在交易所买卖，并且可随时购买或出售。当被授权的参与者（如一位专家）获得了股票组合并在保管银行储存，ETF就被创建了。作为回报，保管银行会提供一些ETFS股份给被授权的参与者。这个过程被称为实物交易。一旦ETF的股票提供给被授权的参与者，他们可以在

41、证券交易所像股票一样买卖。最近，一些ETFS被设计的目的是为了反映固定收益资产。套利阻止了ETFS价格偏离资产净值的价值，因此，可以限制误差。ETF的投资组合与管理的相关费用都非常低。ETF可以卖空，就像股票一样，因此，它提供了一种手段，使市场参与者可以基于对一个特定的市场或行业的悲观预期推测或对冲。然而，他们必须从经纪公司借入ETFS才可以做空。机构投资者更容易借到ETFS,并在卖空过程中扮演了一个最普遍的角色。抵押品必须在经纪公司公布，它可以是现金或国库券。所要求的抵押物的最低金额为交易的50，但一些经纪公司要求更高的比例。不像股票，ETFS可以在价格上升中卖空。虽然卖空者通常不会被收取卖

42、空ETFS的确定费用，但当拿出现金作为担保时，也承担了机会成本。封闭式基金也可做空，但他们受到价格差异的影响，因为它们相对其净资产值的价格可能含有明显的折价或溢价（见PONTIFF1995年、1997年）。这个折扣和溢价随时间而变化，并可能抵消来自于股票卖空期间发生的净资产价值变化的任何利益。此外，许多封闭式基金的交易量是有限的，这可能使抵消卖空头寸的费用是昂贵的。因此，ETFS可以有效地用于对抗特定市场和行业。ETF份额的出售代表了大约占已发行股份的19高于被报道的个别股票水平。做空的ETFS最近变得非常流行，但是关于为什么卖空头寸会变化和是否卖空头寸作为投资者的有用信号缺乏研究。我们的目的

43、是确定吸引空头的ETFS的特征，并确定是否在ETFS中的卖空利息作为悲观情绪的有效信号。我们研究的结果提供了关于做空行为、做空者如何确定他们的目标以及ETFS的卖空利息水平是否作为投资者的指示等的有关影响。我们发现第一，对以部门为基础的ETFS，卖空利息是最大的；对国际的ETFS卖空利息是最小的。第二，对有高交易量和市值较低的ETFS的卖空利息是较高的。这些结论是基于所有的样本的。代表衍生品交易指数的ETFS的卖空利息是相对低的，但是代表衍生品交易指数的国际ETFS的卖空利息相对较高我们还发现，当评估所有样本时，买空水平是一个关于ETFS将来表现的有用的信号。然而，当分析结果根据ETF的类型划

44、分时，情况并不是如此。二、做空利息的相关文献这些研究试图解释为什么一些股票比另一些股票更吸引做空者。BRENTETAL1990显示，高贝塔值和流通的个股偏向于高的卖空利息水平，他们认为，这与套利是一致的。DECHOWETAL2001发现，卖空者通常持有股票价格过高的头寸，这些股票的基本价格比率低，包括现金流与价格的比率、收益价格比、账面市值、市场价值。因为这些低比率可以预测未来股票价格的下降。一些关于做空的研究评估了吸引空头的股票的特征。ANGELETAL2003发现卖空者以最不稳定和交易活跃的股票为目标。他随后在2004年关于公司将来的盈利公告的研究中显示了卖空者最感兴趣的是成长型股票和过去

45、表现良好的股票。收益较差的股票也吸引了做空者。由于使用高层次管理水平的公司往往持续收入较低，预提费用可以用于识别收益差的被错误定价的股票。DESAIETAL2004发现卖空利息应计项目的三种措施和它的变更之间存在强有力的和积极的关系。卖空的许多研究探讨了卖空利息水平和股票价格表现之间是否存在关系。AITKENETAL1998发现，在澳大利亚，高卖空利息水平的股票优于表现不佳的股票。DESAIETAL2002发现在纳斯达克市场上高卖空利息水平的股票优于有负的异常回报的股票。ACKERT和ATHANASSAKOS2005发现被严重抛空的表现不佳加拿大股票表明，卖空利息水平可以提供一个有效的信号。他

46、们认为，与在美国市场上的同样的股票相比，在加拿大的卖空利息水平是一个更有效的指标，因为加拿大市场的限制较少。CHANGETAL2007评估了在香港的卖空条件，发现卖空限制引起了了个别股票的过度反应。三、卖空利息与ETFS实施之间关系的研究为了探讨卖空利息与ETFS实施之间的关系，我们研究了基于以前卖空水平的交易策略是否会带来异常回报。这个交易策略包括两个方面（1）基于先前的卖空利息水平形成的投资组合；（2）评估在随后的持有期内这个投资组合的表现如何。持有样本期间的每个月初，投资组合的形成会标志T。每个投资组合形成初期，我们会获得ETFS卖空利率。我们组成10份基于卖空利息水平的ETF。这些等份

47、组合通过评估去决定它们的持有期的表现。由于没有一个适合所有投资者的单一信号，我们得到了各等份四个不同持有期的ETFS异常报酬。从第一季度开始（紧接本季度的投资组合形成过程），在上一季度的基础上增加三个季度的累积。因此，我们的方法评估了持有期内重叠的投资组合。这与JEGADEESH和TITMAN1993,CHANETAL1996所使用的方法一致。BAYTAS、CAKICI1999和LEE、SWAMINATHAN、JEGADEESH、TITMAN1993认为，重叠期间的使用提高了统计的检验功效。然而，我们也采用了一种方法，作为评估在不重叠时期稳健性检验的投资策略。具体来说，在每年的一月和六月初，所

48、有的ETFS根据它们先前的卖空水平被分为10等份组合。这些投资组合使用于未来的6个月，最高和最低的的等份的异常回报之间的区别由每个不重叠的6个月持有期和对重要性的检验决定的。由于非重叠期间的结果与重叠期间的发现非常相似，它们就不再报告了。持有期间的异常报酬不同于ETF的等份组合利润和相应的基准投资组合的利润。ETF的持有期报酬（HPRS）是按复利每季度计算的，股息是一季度计算一次。持有期收益以市场为基准，MHPR1到MHPR4以同样的方法计算。广义的和以部门为基础的ETFS的MHPRS用平均加权指数计算。国际ETFS的MHPRS从数据库提取数据流，使用摩根士丹利世界指数计算得到的。每个ETF持

49、有期异常报酬（AHPRI）是用HPRI和相应的MHPRI的差别衡量的（其中I1至4季度）。持有期的平均异常收益（AAHPRI）是按下列公式计算等分组合这里I代表T后的季度数据（I1到4），K代表ETFS投资组合的每等份数（K1到N）。来自交易策略的利润等于代表最高等份的ETFS的投资组合的持有期异常报酬和代表最低等份ETFS投资组合持有期异常报酬的差异。四、卖空利息与ETFS实施关系的结果根据每个月的卖空头寸规模创建ETF投资组合的结果显示在表6中，A组涵盖了所有样本，B,C,D组是子样本。从整个样本来看，代表最大与最小的卖空头寸比较，ETFS的平均持有期异常报酬的差异是负的和意义重大的，那时，投资组合形成后的三四个月内的AAHPRI考虑在内。也就是说，投资组合形成时最大卖空头寸的等份在持有期内表现恶劣。对于广义的ETFS样本，最小的和最大的ETFS之间，持有期异常报酬的差异是无关紧要的，而且忽视了持有期的长短，在这期间以部门为基础和国际ETFS投资组合是存在的。因此，当一个ETF的特定类型孤立时，等份基金的显著性差异消失了。这意味着，作为一个群体评估所有ETFS时比集中于ETF的一个特定类型时相比，来自卖空头寸的信号是非常有效的。五、总结卖空ETFS在近几年逐渐流行，现在甚至比卖空以发行股票流行，什么试图确定（1）比卖空股票更能吸引人的ETFS的特征；（2）卖空水平对ETFS