基于Copula理论的VaR算法与实证分析

针对在金融市场中资产收益的非正态分布以及资产组合中不同的风险资产的非线性相关,将Copula理论应用于资产组合的VaR估计之中。上证指数和深证综合指数的算例表明,由于Copula方法允许去选择更恰当的边际分布,并且Copu-la理论对相关性分析特别是尾部的相关性分析有特殊优势,可以跟更广泛的捕获各风险资产的相关信息,基于Copula理论的VaR在估算的精度方面优于传统的方法。

VaR algorithm and empirical analysis using Copula

Copula theory is put forward to solve the problem of non-gauss distribution of assets returns and non-linear correlation of the risky portfolios.It is applied to calculating the VaR.The computing example of SH index and SZ index are presented and the computation results indicate that VaR algorithm based on Copula outperformed the conventional numerical method on the aspect of accuracy,since it allows us to select the right marginals of distribution.