向量GARCH族模型计算中国金融市场投资组合风险价值(VaR)的比较研究与评述

对由上证综合指数、深证成分指数、上证基金指数、上证国债指数计算的日自然对数收益率组成的数据矩阵,分别建立了残差服从正态分布、t分布的向量ARCH、向量GARCH、纯对角GARCH、BEKK、常条件相关GARCH、主成分GARCH和EWMA模型,基于这些模型,计算了风险价值(VaR),进而通过比较计算结果,得出BEKK—t模型测算中国金融市场投资组合的风险价值(VaR)效果最好等的结论.

To Evaluate VaR of China Money Market Comparatively sing Vector GARCH Family Models and Comment

We established the Vector ARCH,Vector GARCH,Pure Diagonal GARCH, BEKK,Constant Condition Correlation GARCH,Principal Component GARCH and EWMA model of a vector which contained log profit of intra-daily close price on the Shanghai Stock, Fund,and Bond Market and the Shenzhen Stock Market with normal or t distribution,and then we calculate the VaR(value at risk) on these models and compared the results,we found that the BEKK with t distribution was the best one.