InternationalStockCorrelationsandMacroFluctuations国际股票市场间相关系数与宏观经济波动YijieCaiJinheCenterforEconomicResearchXi’anJiaoTongUniversity,ChinaRayYeutienChou∗InstituteofEconomics,AcademiaSinicaandNationalChiao-TungUniversityDanLiJinheCenterforEconomicResearchXi’anJiaoTongUniversity,China联系方式:1、周雨田研究员,研究领域:数量经济学,金融市场,时间序列分析联系地址:台北市南港研究院路二段128号中研院经济所。邮编:11528Email:rchou@econ.sinica.edu.tw,联系电话:886-2-278227912、蔡义杰在读博士生,研究领域:数量经济学联系地址:西安交通大学金禾经济研究中心。邮编:710049Email:caidapi@163.com,联系电话:135725366843、李丹在读博士生,研究领域:数量经济学联系地址:西安交通大学金禾经济研究中心。邮编:710049Email:lidan@mail.xjtu.edu.cn,联系电话:13468822269∗Correspondingauthor.Contactaddress:InstituteofEconomics,AcademiaSinica,#128,Yen-Jio-YuanRoad,Sec2,Nankang,Taipei,Taiwan.Telephone:886-2-27822791ext321,fax:886-2-27853946,email:rchou@econ.sinica.edu.tw.1AbstractInthispaper,realandfinanciallinkageistobeinvestigated.Wefocusonsixtypicalstockmarketsaftertimezoneeffecttakenintoconsideration.WeselectmonthlyannualCPIrateastransitionvariableinSmoothTransitionConditionalCorrelationCARR(namedSTCC-CARRforshort)modeltoscrutinizeinterdependenceamonginternationalstockmarkets.Asitistestified,correlationsamongthemarefluctuantwithdifferentinflationcyclesandcouldnotbeignoredarbitrarily.Thehighestcorrelationscomeoutbetweencountrieswhenbothareincontractionaryphase,whilethelowestcorrelationsdowhenbothareinexpansionaryphase.Keywords:EconomicFundamentals,InternationalStockMarkets,CPIRates,SmoothTransition,CARR.摘要本文深入调查了国际金融市场间相关性与实际面的联系。我们选择了6个典型的国际股票市场为研究对象,并考虑了时区效应的影响,同时选取月度年CPI变化率做为转移变量,利用平滑转移条件相关系数CARR模型(简称为STCC-CARR)仔细研究了国际股票市场间的相依性。通过本文的证明,我们发现两市场的条件相关系数将随着两市场的通货膨胀周期而波动且这种影响不能被忽视。当两个国家同时处于通货紧缩期时,两市场具有最高的相关性;而当两国同时处于通货膨胀期时,市场间具有最低的相关性。关键字:经济基础,国际股票市场,CPI,平滑转换,CARR21.IntroductionFinancialandtraditionaltradelinkagesaretwotypicalapproachestodispersecontagionacrosscountries.Withtheintegrationandglobalizationoffinancialmarkets,internationalstockmarketscorrelationshavebeenpaidmoreattentionto.Awealthofqualitativeliteratureshavebeendevotedtotheintriguingconnectionbetweenfinancialmarketsandeconomicfundamentals,andprovidesufficientevidencesthatco-movementofbusiness-cyclefluctuationsdoimpactoncorrelationsofinternationalfinancialmarketsmoreorless.However,thecontroversyhasneverstopped.Debatesonwhetherinternationalfinancialcorrelationsaresignificantlyaffectedbyeconomicfundamentals,suchasbusinesscycleindicatorsandsoon,cameforthintheearly1990s,andhaven’treachedaconsistentagreementyet.Erbetal.(1994)findthatcorrelationsbetweentwoequitymarketsarevaryingaccordingtobothcountries’economiccycles,whostandsthepositivepointthateconomicfundamentalsaffectremarkablyonstockmarketcorrelations.TakingG-7countriesforinstance,analysisshowsthatthehighestcorrelationsappearwhenbothcountriesstandincontractionaryphaseandthelowestcorrelationsdowhenbothcountriesareinexpansionaryphase.Correlationsarevaryingbetweenthesetwoextremestateswhentheyareoutofphases.Longinetal.(1995)drawaconclusionthatthereisagreatandincreasingcorrelationbetweeninternationalstockmarketsinperiodsofhighervolatility1.Internationalintegrationtightensthefinanciallinkage1LonginandSolnik(2001)findthatcorrelationwouldincreaseinbearmarkets,butnotinbullmarkets.Theypointoutthatcorrelationisnotrelatedtomarketvolatilitybutmarkettrend.3progressively.Thecorrelationscouldbepredictedtoclimbinperiodsoflowdividendyieldandhighinterestrates,andthiswasattributedtotheevidencethateconomicvariablessuchasthedividendyieldandinterestratescontaintheinformationaboutfuturevolatilityandcorrelationthatisnotincludedinpastreturnsalone2.Ragunathanetal.(1999)researchonwhetherandtowhatextentbusinesscyclesimpactoncorrelationsandintegrationbymakingasurveyoftheAustralianandAmericanmarketsdata.TheysuggestthatcorrelationsbetweenthesetwomarketswouldbethehighestwhenUSAwasinacontractionaryphase,andalsopointoutthatthemarketsaremorelikelytobesegmentedandintegratedincontractionaryandexpansionaryphaseofbusinesscyclesrespectively.Dumasetal.(2003)highlightthestatisticalevidencethatoutputcorrelationsandstockmarketscorrelationsarepositivelyrelated.Forbesetal.(2004)showushowimportanttherealandfinanciallinkagesbetweentwocountriesare.Theyconsiderthatdirecttradeseemstobethepredominantmovementoftheworld’slargestmarketsthataffectfinancialmarkets,andtheyalsohaveobservedthatUSAplayadominantroleagainstJapanandUK.Onthecontrary,thereareliteraturesstandingwithnegativeviewpoints.Kingetal.(1994)suggestthatco-variancesbetweeninternationalstockmarketsaredifficulttobeinterpretedbytheobservableeconomicvariables,reverselycanbechangedbythevariablesunobservable.AmmerandMei(1996)discoverthatco-variancesbetweeninternationalstockmarketsareinfluencedbycontemporaneousco-movementinmacroeconomicvariables.Buttheythinkitcanbeneglected,becausethereal2Althoughtheeffectisnotobvious,itisprovedstatisticallytobesignificantinthatpaper.4linkagesaremuchstrongerfromalong-runthanashort-runperspective.Kizysetal.(2006)supportAmmerandMeiandshowthatthelinkagebetweenmonthlyconditionalinternationalequitycorrelationsandco-movementofbusiness-cyclefluctuationsisnotsignifica