基于POT-VaR混合模型的投资组合风险分析

在简要分析金融产品投资组合理论的基础上,尝试将超门限极值法(POT)与传统VaR模型中的方差-协方差法相结合的方式,改进传统单纯基于投资组合计算历史收益率数据的VaR模型.通过划分沪深股价前100支股指期货高频收益率数据,运用POT-VaR混合模型进行实证研究.结果表明:该混合模型可以解决收益率方差估计随时间推移的动态问题,为降低投资组合市场风险提供科学的方法,对我国期货市场投资组合风险管理提供了可靠的依据.

Risk analysis of the investment portfolio based on the POT-VaR hybrid model

With a brief theoretical analysis of the investment portfolio of financial products,this paper tries to improve the traditional simple calculation based on the VaR model of the past portfolio returns by integrating the POT with the traditional VaR model of the variance- covariance method. Through the study of the division of the high-frequency data of CSI 100 stock index futures,this paper uses the POT- VaR hybrid model for empirical researches. The results show that the hybrid model can solve dynamic problems of the estimate of variance for the rate of returns with the passage of time,and then it provides the scientific method of reduce for the investment portfolio market risk and a reliable basis of the investment portfolio in the risk management of China's futures market.