Copula理論在相關(guān)性分析中的應(yīng)用及其多元擴(kuò)展.pdf

1、天津科技大學(xué)碩士學(xué)位論文Copula理論在相關(guān)性分析中的應(yīng)用及其多元擴(kuò)展姓名：盧穎申請(qǐng)學(xué)位級(jí)別：碩士專(zhuān)業(yè)：管理科學(xué)與工程指導(dǎo)教師：杜子平200904ABSTRACTnispaperfirstlysummarizesCopulatheoryanditsmodelingmethodsystematicallySubsequently，thegraphtheoryisintroducedinordertoshedsomelightonmode

2、lingmultivariabledependencestructurewhichalSOprovidesthepossibilityforthemultiplevariablesgroupingdecompositionFinallytheapplicationinmultivariablefinancialseriesanalysisiSinvestigatedIntheanalysisofthefourfinancialserie

3、snotonlythebivariateseriesdependenceisstudied，butalsothedependencestructureoffourvariablesismodeledusingvinesmethod，SOtheadvantageofvinemethodofgraphtheoryinmultivariabledependencestructuremodelingCanbeprovedThekeypoints

4、andmainachievementsofmisworkarelistedasfollow：(1)ThispapersummarizesCopulatheoryanditsmodelingmethodsystematicallyItintroducesCopula’SconceptandpropertyconditionalCopulageneralform，therelevantindexeswhicharederivedbyCopu

5、latheoryandtheparameterestimationinCopulamodeling(2)Thegraphtheoryisintroducedtoprovideanewwayofmultivailabledependencestructuremodelingwhichalsocreateanalgorithmtosolvethepositiveproblemwhenmodelingmultivariabledependen

6、cestructure，andstudiesthedecompositionofthemultiplejointdistilbutionsusingvinesCombiningCopulatheorywitllvinetheorythispaperinnovativelyconstructsamultivariabledependencestructurewiththevine’Siterativestructurelayeruponl

7、ayerFurthermore，thefeasibilityofmultiplevariablesgroupingdecompositionusingvinesmethodiSdiscussedAtthesametime，themethodofparameterestimationinCopulavinemodelingisshownBesides，relevantinformationisintroduced，whichiSusedf

8、orinterpretingthevinedescriptionofdependencestructure(3)CopulaandvinetheoryareinnovativelyappliedinfinancefieldThepaperchoosesthefourexchangeratedataassampleseriesBasicstatisticsofthesamplearecalculatedfirstlyandthenGARC

9、HmodeliSconstructedThedependencestructurebetweenbivariateseriesiSanalyzedusingCopulafunctionandthefittingresultsofdifferentCopulasareinvestigatedAtlastdependencemodelofthefourvariantsiSconstructedusingvines，whichcompares

10、withGaussianCopulamultivariabledependencemodelThentheCopulavinemodeliSimprovedTheimprovedmodelnotonlychoosesdifferentCopulasinthesamevinebutalsotriestochoosedifferentvines，andthencomparewiththeformermodelbyfittingtestThe

11、processnotonlyshowsthevalidityofmultivariabledependencestructuremodelingwithvinestheorybutalsoshowstheagilityofvinestructurebychoosingdifferentCopulafunctiononeverylayerandchangingdifferentvinestructureKeywords：Copula，mu