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.docxVIP

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 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，提出一种新的基于EMD．Copula．GARCH的风险 度量模型用于估计股票投资组合的VaR。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 ABSTRACTCoPULA ABSTRACT CoPULA BASED VAR MEASUREMENT FoR STOCK PoRTFoLIo ABSTRACT With the constant changes in the fiinancial markets，the accurate measurement for risk has become a basis to make rational decisisons for effective risk management and investors．To effectively avoid and hedge risk caused by the price fluctuation of financial assets，and establish a reliable and precise mathematical models to measure risk for financial market portfolio， improving the measurement accuracy of the portfolio is very important． According to the existing risk measurement and assessment studies，the Value—at-Risk method is a popular intemational risk management standards， and it can effectively measure risk of the stock portfolio．Compared with the traditional financial risk measurement model，this approach Can cover a variety of market factors affecting financial assets．Meanwhile，it can also measure non—linear risk，with greater adaptation and scientific．However,in view of the complex and diversified features of financial markets，the correlation between financial assets is significant increasing，particularly in the the market with a downtum(bear)，the correlation between assets is larger than a normal period(bull)，and therefore measuring the dependence between III 万方数据 万方数据 万方数据 北京化工大学硕士学位论文financial 北京化工大学硕士学位论文 financial assets is extremely important for studying the risk management，asset pricing and portfolio analysis．In portraying dependence between random variables，Copula which severd as a effective modeling tool to depicting the r