基于TGARCH-t的混合Copula投资组合风险测度研究

The Research on Mixture Copula Portfolio Risk Measurement Based on TGARCH-t Model

在分析了现有Copula函数在测度投资组合风险不足的情况下,首先充分考虑资产波动的时变性、杠杆效应等特征,选择了TGARCH-t模型进行边缘分布建模.接着引入混合Copula模型来描述投资组合的复杂相关结构,同时利用构造的主对角线距离统计量等方法验证了混合Copula模型的优势.最后通过VaR的蒙特卡洛模拟结果看到,这种方法能更为精确的测度投资组合风险值.

After the deficiency of portfolio risk measurement using Copula currently discussed,TGARCH-t model is selected to describe the time-varying and leverage about asset volatility.Then,mixture copula is to used to model the complex corrlation structure of portfolio,and the main diagonal distance statistics test the superiority of mixture copula.Finally,the result of Monte Carlo simulation of VaR shows that this method measure the risk of the portfolio more accurate.