基于Copula理论的股票投资组合VaR风险度量研究-管理科学与工程专业论文.docx

北京化工大学硕士学位论文进行扩展,提出的二元EMD算法,可有效解决EMD方法在处理二元数据分析时所存在的模式混叠与尺度不对齐等问题。对此,为提高估计精度,本文将引入二元EMD算法,以构建新的多尺度股票投资组合风险度量模型。因此,本文将基于有效的风险度量Copula-GARCH模型,以及多尺度分析二元EMD算法,。首先,结合二元EMD算法与Copula理论探讨股市间微观相关结构的特征。其次,构建一种基于二元EMD—Copula—GARCH的VaR风险度量模型,对股票市场中所存在的风险进行预测,并与现有的风险度量模型的VaR结果相比较,最后通过实证分析证明了EMD-Copula-GARCH模型在股票市场风险预测方面的有效性。关键字:VaR,Copula,EMD,股票投资组合,风险度量Ⅱ万方数据ABSTRACTCoPULABASEDVARMEASUREMENTFoRSTOCKPoRTFoLIoABSTRACTWiththeconstantchangesinthefiinancialmarkets,,andestablishareliableandprecisemathematicalmodelstomeasureriskforfinancialmarketportfolio,,theValue—at-Riskmethodisapopularintemationalriskmanagementstandards,paredwiththetraditionalfinancialriskmeasurementmodel,,itcanalsomeasurenon—linearrisk,,plexanddiversifiedfeaturesoffinancialmarkets,thecorrelationbetweenfinancialassetsissignificantincreasing,particularlyinthethemarketwithadowntum(bear),thecorrelationbetweenassetsislargerthananormalperiod(bull),andthereforemeasuringthedependencebetweenIII万方数据北京化工大学硕士学位论文financialassetsisextremelyimportantforstudyingtheriskmanagement,,CopulawhichseverdasaeffectivemodelingtooltodepictingtherelationshipbetweenrandomvariablesCannotonlyreflectlinearornon—linear,symmetricorasymmetricdependencebetweenvariables,butalsocatchthetailcorrelationbetweenthem,,themulti—plexsystem,(EMD)modelwhichisaneffectivemulti—uratelyreflectthephysicalcharacteristicsoftheoriginalsignal,especiallyindealingwiththenon—linearstationarydata