股指期貨對韓國股票市場的影響外文翻譯

外文翻譯原文TheimpactofstockindexfuturesontheKoreanstockmarketMaterialSource:/ehost/pdfviewer/pdfviewer?hid=15&sid=6d323da0-58fb-47c7-8da9-3208cc81ed87%40sessionmgr14&vid=8Author:HYUN-JUNGRYOOandGRAHAMSMITHThisarticleinvestigatestheimpactonthespotmarketoftradinginKOSPI200futures.Empiricalresultsshowthatfutures’tradingincreasesthespeedatwhichinformationisimpoundedintospotmarketprices,reducesthepersistenceofinformationandincreasesspotmarketvolatility.Thespotandfuturespricesarecointegratedandthereisbidirectionalcausalitybetweenthetwomarkets.Thelead-lagrelationisasymmetricwithweakerevidencethatthespotindexleadsfuturesandstrongerevidencethatthestockindexfuturesmarketleadsthespotmarket.Tradinginthesestockindexfutureshasgrownremarkably.Stockindexfuturesareperceivedasoneofthemostsuccessfulfinancialinnovationsofthe1980s.TradinginthemwasfirstintroducedinFebruary1982bytheKansasCityBoardofTradeintheUSAandotherdevelopedmarketssoonfollowed.Incontrast,muchofthefuturestradinginemergingmarketsarearelativelyrecentphenomenon.AlthoughKoreaisoneofthefastestgrowingemergingmarkets,itwasnotuntil3May1996thatafuturescontractbasedontheKoreaStockPriceIndex200(KOSPI200)wasintroducedontheKoreaStockExchange(KSE).Tradinginthesestockindexfutureshasgrownremarkably.Bytheendof1998,averagedailytradingvolumewas61279contracts(value1.39millionwons).AccordingtotheFIBV(1999),byDecember1998thestockindexfuturesmarketinKoreahadbecomethesecondlargestintheworldintermsoftradingvolume,aftertheChicagoMercantileExchange(CME).Theimpactofderivativetradingonspotpricevolatilityhasbeenwidelyinvestigatedfordevelopedmarkets.Forstockindexfutures,researchhasfocusedoncomparingspotpricevolatilityinperiodsbeforeandaftertheintroductionoffutures(forexampleEdwards,1988;Harris,1989;AntoniouandHolmes,1995).Inparticular,oneoftheprimaryconcernsofpreviousstudieshasbeentheissueofwhetherfutures’tradingdestabilizestheunderlyingspotmarket.Althoughsomestudiesfindincreasedvolatility,theweightoftheempiricalevidenceshowsnoincreaseinvolatilityfollowingtheintroductionoftradinginstockindexfutures.Amongothers,Freris(1990)examinedtheimpacteffectofHangSengIndexFuturesonthebehaviorsoftheHangSengIndexusingdatafortheperiodfrom1984to1987andfoundthattheintroductionofstockindexfuturestradinghadnomeasurableeffectonthevolatilityofthestockpriceindex.LeeandOhk(1992)examinedtheeffectofintroducingindexfuturestradingonstockmarketvolatilityinAustralia,HongKong,Japan,theUKandtheUSAusingdailyindexdataforperiodsofapproximatelyfouryearsspanningthestartoftradeinindexfutures.Theyfoundthatforthethreelargestmarkets,returnvolatilityincreasedsignificantlyafterthelistingofstockindexfutures.However,fortheAustralianmarkettherewasnosignificantdifference,andforHongKongstockreturnvolatilityactuallydecreased.Usinginternationalportfolios,theyfurtherfoundthatalthoughthecreationofstockindexfuturesgenerallyexertedavolatility-increasinginfluenceonthebehaviorsofcashmarketstockreturns,itmadethestockmarketrelativelymoreefficientbecausevolatilityshocksweremorequicklyassimilatedinthatmarket.Kamaraetal.(1992)investigatedtheeffectoffuturestradingontheS&P500onthestabilityoftheunderlyingindex.Theirresultssuggestedthat,althoughthevolatilityofmonthlyreturnsremainedunchanged,thevolatilityofdailyreturnsinthepost-futuresperiodwashigherthaninthepre-futuresperiod.Oneoftheimportantrolesattributedtofuturesmarketsisthatof‘pricediscovery’;thatis,thefuturesmarketreflectsnewinformationbeforethespotmarket.Ifnewmarketinformationdisseminatesinthefuturesmarketbeforethestockmarket,thentheintroductionofafuturesmarketincreasestheamountofinformationreflectedinthespotprice.Thismightbeexplainedbythefactthattradingfutureshastheadvantagesofahighlyliquidmarket,lowtransactioncosts,easilyavailableshortpositions,lowmarginsandrapidexecution.Thus,informedtradersmayfindtheycanactfasterandatlowercostinthefuturesmarketthantheycaninthecashmarket,resultinginalead-lagrelationshipbetweenfuturesandspotprices.Thelead-lagrelationbetweenmovementsofspotandfuturespriceshasbeenwidelyinvestigatedwiththemethodsusedvaryingacrossstudies.Forexample,Kawalleretal.(1987),Abhyankar(1998)andTangetal.(1992)usemodified/non-modifiedGrangercausalitytests.WhereasWahabandLashgari(1993),Flemingetal.(1996)andPizzietal.(1998)usecointegrationanderrorcorrectionmodels.However,irrespectiveofmethodology,theresultscanbesummarizedconcisely:marketinformationtendstodisseminateinfuturespricespriorto,andatgreaterspeedthan,instockprices.Alloftheresearchsofardiscussedfocusesondevelopedmarkets.Verylittleworkhasinvestigatedtheimpactofstockindexfuturestradinginemergingmarkets.Thisarticlecontributestothissparseliterature;itisthefirsttoexaminetheimpactontheKoreanspotmarketoftradinginfutures.Itfocusesonthreeaspects.First,theimpactoffuturestradingonpricevolatilityinthespotmarketisexamined.Second,itdiscusseslong-runequilibriumandshort-runadjustmentthroughtestsofcointegrationandcausality.Third,lead-lagrelationshipsareanalyzed.Thisarticlediffersfrompreviousstudiesthatuseclosingpricesforfuturesandspotprices,byusingdatawithmatchedclosingtimes.ThisisdesirablebecauseinKoreatradinginindexfuturesandtradinginstocksfinishatdifferenttimes.Byusingmatchedclosingtimes,oneavoidscomparingnon-synchronousclosingpricesofthespotindexandfuturescontracts,whichmightleadtoasignificantsourceoferror.TheimpactoffuturestradingonpricevolatilityintheunderlyingspotmarketindexisexaminedbyadoptingthegeneralizedARCH(GARCH)processinwhichtheconditionalvarianceofuattimetisdependentnotonlyonpastsquareddisturbancesbutalsoonpastconditionalvariances.Empiricalevidence,forexample,Bollerslevetal.(1992),Huangetal.(1995)andRyoo(2001),findsthatreturnsinstockmarketsexhibitheteroscedasticity.Therefore,followingHolmes(1996),aGARCHrepresentationwouldseemtobeanappropriatemeansbywhichtocapturemarket-widepricevolatility.Considerthemodel:（1）（2）whereRs,tisspotpricereturns,thechangeinthelogarithmofthespotpriceindexinperiodt,Rp,tisreturnsonthemarketproxyvariable(forwhichthereisnoassociatedfuturesindex),utisanerrortermrepresentingunexplainedpricechangesandψtistheinformationsetavailableattimet.Sincetheproxyvariablecoversmarket-wideinfluencesonpricechanges,theerrortermcapturestheimpactoffactorsspecifictothemarketonwhichthefuturescontractiswrittenanditsvariance,ht,providesameasureofmarket-widepricevolatility:（3）Byestimatingthemodelforperiodspre-andpost-futurestradingandcomparingtheparametersofthevarianceequations,itispossibletodeterminehowfuturestradingimpactsonvolatility.Thecoefficientα1relatestoday’spricechangestoyesterday’smarket-specificpricechangesandasthesedependonthearrivalofinformationyesterday,α1canbeviewedasa‘news’coefficient.Ifincreasesα1followingtheintroductionoffuturestrading,thissuggeststhatfuturestradingresultsininformationbeingimpoundedintospotpricesmorerapidly.Conversely,ifαldecreasesinthepost-futuresperiod,thissuggeststhatinformationisimpoundedintospotpricesmoreslowly.EngleandBollerslev(1986)andEngleandMustafa(1992)showedfortheGARCH(1,1)model,thatthepersistenceofvolatilityshocksdependsprimarilyonα1+β1.Anincrease(decrease)inα1+β1followingtheintroductionoffuturestradingindicatesincreased(decreased)persistenceofvolatilityshocks.TheKoreaStockPriceIndex200(KOSPI200)istheunderlyingstockindexfortradedfuturesandoptionscontractsontheKSE.TheKOSPI200isacapitalization-weightedindexthattracksthecontinuouspriceperformanceof200activelytraded,largecapitalizationcommonstockslistedontheKSE.Thesesharesaccountforapproximately70–80%ofdomesticmarketcapitalizationsotheindexreflectsoverallmarketperformance.Inordertoavoidunintendedbias,theconstituentstocksarerigorouslyrevisedovertime.Thebasefigurewassetat100.00asof3January1990.TradingofKOSPI200futuresisimplementedunderanorder-driven,continuoustradingsystem.Sincetradingisexecutedthroughacomputerizedsystem,thereisnophysicaltradingfloor.KOSPI200futuresexpirefourtimesayear,inMarch,June,SeptemberandDecember.ThelasttradingdayisthesecondThursdayofeachcontractmonth.Oneindexpointequals500000Koreanwonandsettlementofthecontractisincash.Therearetwotradingsessions,morningandafternoon.Until5December1998,bothstockandstockindexfuturescontractsweretradedonweekdaysbetween9:30a.m.and11:30a.m.inthemorningsession.Intheafternoonsession,stocksweretradedfrom1:00p.m.until3:00p.m.andindexfuturesweretradedbetween1:00p.m.and3:15p.m.exceptionthelasttradingdayofeachcontractmonth,whenfuturestradingclosedat2:50p.m.OnSaturdays,bothstockandindexfuturesweretradedfrom9:30a.m.until11:30a.m.and11:45a.m.,respectively.Since7December1998,therehasbeennoSaturdaytradingandthemorningsessionforweekdayshasbeenextendedfrom9:00a.m.untilmid-day.Withtheintroductionofstockindexfuturestrading,thereweredailypricelimitsforfuturescontractsof±5%oftheprevioustradingday’sclosingprice.On2March1998thiswasrelaxedto±7%andon7December1998to±10%.Thereisalsoasystemofcircuitbreakers.Whenthepriceoftheprevioustradingday’smostactivecontractreaches±5%ofthatday’sclosingpriceforoneminute,thetradingofallfuturescontractsishaltedforthenextfiveminutes.Also,whentheKOSPIcontinues(foroneminute)tolose10%ormoreofitsvaluecomparedtothepreviousday’sclosingprice,futures’tradingishaltedfor20minutes.譯文股指期貨對韓國股票市場的影響資料來源:/ehost/pdfviewer/pdfviewer?hid=15&sid=6d323da0-58fb-47c7-8da9-3208cc81ed87%40sessionmgr14&vid=8作者：玄榮龍，格雷厄姆本文從影響現(xiàn)貨市場的KOSPI200期貨進(jìn)行研究。實(shí)證結(jié)果表明，股指期貨的交易增加了現(xiàn)貨市場價(jià)格信息發(fā)現(xiàn)的速度，降低了信息的持久性，同時(shí)增加現(xiàn)貨市場波動性。現(xiàn)貨和期貨價(jià)格之間存在協(xié)整的雙向因果關(guān)系。研究表明現(xiàn)貨市場和股指期貨市場存在著不對稱性，股票指數(shù)期貨市場領(lǐng)先現(xiàn)貨市場。股指期貨被視為20世紀(jì)80年代的最成功的金融創(chuàng)新之一。1982年2月，股指期貨首次由堪薩斯市教委在美國市場推出，其他發(fā)達(dá)市場緊隨其后。與此相反，在許多新興市場，期貨交易是一種比較新的現(xiàn)象。雖然韓國是發(fā)展最快的新興市場之一，但直到1996年5月3日，基于韓國證券交易所價(jià)格指數(shù)200（KOSPI200）的股指期貨在韓國證券交易所推出。在這些股票指數(shù)期貨的交易增長顯著。到1998年底，平均每日交易量為61279的合同（價(jià)值139萬韓元)。據(jù)國際證券交易所聯(lián)會（1999年），根據(jù)到1998衍生工具交易對現(xiàn)貨價(jià)格的影響已經(jīng)被廣泛運(yùn)用到發(fā)達(dá)市場的研究中去。對于股指期貨的研究主要集中在比較股指期貨推出前后現(xiàn)貨價(jià)格的波動性方面（例如Edwards,1988；Harris,1989;AntoniouandHolmes,1995）。特別是，以往研究的首要問題之一是是否期貨交易對現(xiàn)貨市場起到了本質(zhì)上的影響。盡管一些研究發(fā)現(xiàn)買賣股票指數(shù)期貨增加了現(xiàn)貨是市場的波動性，但是大多是經(jīng)驗(yàn)證據(jù)表明股指期貨推出對波動性沒有影響。其中，F(xiàn)reris（1990）對1984年至1987年之間的恒生指數(shù)進(jìn)行研究，他發(fā)現(xiàn)恒生指數(shù)期貨的引進(jìn)對恒生價(jià)格指數(shù)的波動沒有起到明顯的作用。Lee和Ohk（1992）利用跨越了四年的每日股指期貨數(shù)據(jù)，對在澳洲、香港、日本、英國和美國引入股指期貨后股票市場波動性影響進(jìn)行研究。他們發(fā)現(xiàn)對第三大市場來說，在股指期貨推出之后，波動報(bào)酬率有了明顯的增長。在澳大利亞市場上，波動報(bào)酬率沒有顯著差異，香港股市場卻有著明顯的下降。利用國際投資組合，他們進(jìn)一步發(fā)現(xiàn)，雖然股指期貨的產(chǎn)生對現(xiàn)貨市場的股票收益率波動性的影響越來越大，這是因?yàn)椴▌幼兊迷絹碓届`敏從而使得股市也隨之變得更加有效。Kamara（1992）研究了以S&P500指數(shù)為標(biāo)的物的股指期貨的穩(wěn)定性影響。他的研究結(jié)果表明，雖然每月回報(bào)的波幅維持不變，但是在股指期貨買賣后期的波動收益率明顯高于股指期貨買賣前期。股指期貨在期貨市場上最重要的作用是價(jià)格發(fā)現(xiàn)，也就是說，股指期貨對新信息的反應(yīng)比現(xiàn)貨市場更加超前。如果新的市場信息傳播在期貨市場之前，那么期貨市場引入增加信息量將在現(xiàn)貨價(jià)格中反映出來。這可能解釋為交易期貨具有流動性高，交易成本低，低保證金和快速交易的市場優(yōu)勢。因此，知情交易者可能會發(fā)現(xiàn)，他們可以比現(xiàn)貨市場上更快做出反應(yīng)和以更低的成本在期貨市場交易，這是由于期貨和現(xiàn)貨價(jià)格之間的超前滯后關(guān)系造成的。各國學(xué)者們運(yùn)用不同的方法，對現(xiàn)貨價(jià)格和期貨價(jià)格直接的超前滯后關(guān)系進(jìn)行的調(diào)查研究。如Kawalle（1987年）和Abhyankar（1998年）使用修正或者未修的Granger因果關(guān)系進(jìn)行檢驗(yàn)。Wahab和Lashgari（1993），F(xiàn)leming（1996）還有Pizzi（1998）等都使用協(xié)整和誤差修正模型進(jìn)行研究。然而，不論方法結(jié)果可以簡要概括：市場信息在期貨市場上的傳播往往早于在現(xiàn)貨市場。到目前為止，所有關(guān)于股指期貨的研究都集中于發(fā)達(dá)的市場，只有很少一部分關(guān)于股指期貨交易在新興市場的影響研究。這篇文章對新興市場關(guān)于這方面的研究做出了貢獻(xiàn)，因?yàn)樗堑谝粋€研究韓國股指期貨買賣對現(xiàn)貨市場的影響論文。它主要集中在三個方面的研究：第一，期貨交易對現(xiàn)貨市場價(jià)格波動的影響的研究。其次，通過協(xié)整和因果關(guān)系檢驗(yàn)討論長期均衡和短期調(diào)整的研究。第三，超前滯后關(guān)系的分析。本文不同于過去使用期貨收盤價(jià)格與現(xiàn)貨價(jià)格數(shù)據(jù)，而是使用匹配的關(guān)門時(shí)間價(jià)格數(shù)據(jù)。這種研究方法是可取的，因?yàn)樵陧n國的股指期貨和股票買賣交易在不同的時(shí)間完成。通過使用匹配的關(guān)門時(shí)間，它避免了因非同步收市而可能引起重要錯誤的現(xiàn)貨指數(shù)和期貨合約之間的比較。關(guān)于價(jià)格的相關(guān)現(xiàn)貨市場交易期貨指數(shù)的波動影響審查采用廣義的ARCH（GARCH模型），在條件方差u不僅是過去時(shí)間t平方的干擾，也收到過去的條件方差的影響。例如，Bollerslev（1992