资产组合的集成风险度量及其应用——基于最优拟合Copula函数的VaR方法

Portfolio integrated risk measurement and its application-VaR method based on goodness-of-fit copula functions

通过对不同族类、不同种类Copula函数之间的比较分析,提出了最优拟合Copula函数的一种选择方法.基于沪深两市经验数据的实证检验与分析表明,Frank-Copula和Clayton-Copula分别适用于计算低置信度和高置信度下资产组合集成风险的VaR.在各自置信度下,根据这两种Copula函数的计算方法优于其它Copula函数方法,更优于使用多维正态分布或者多维t分布的传统方法.

The key to applying the copula function to measure integrated risk of portfolio is to search for the Goodness-of-fit copula. This paper proposes a method to determine the Goodness-of-fit copula by comparing and analyzing copulas from different families and types. The test and analysis on empirical data of Shanghai and Shenzhen security exchange testify, Frank-Copula and Clayton-Copula respectively apply to calculate VaR of portfolio integrated risk under low and high confidence level. Under respective confidence level, the methods based on these two Copulas are better than other Copula functions, and much better than the traditional methods based on multivariate Gaussian distribution and multivariate t distribution.