基于多维Copula函数的投资组合CVaR分析

Portfolio CVaR Analyses based on the Multi-dimensional Copula Function

目前国内有关Copula函数的实证研究主要以二种资产的相关性为主。根据Copula函数在构建反映随机变量实际分布与相关性的联合分布函数上的优势,构建了反映多个资产收益实际分布和相关性的联合分布函数,并使用蒙特卡罗模拟技术,分析在不同置信度下投资组合的最小条件风险价值(CVaR)。实证表明,根据所提出的模型度量资产的风险,可以使投资者选择的资产更加稳健,同时也有利于投资者对投资组合整体风险进行分散和监管。

Based on the fact that the current domestic empirical research on Copula is mainly to examine two kinds of assets, we＇ll construct a joint distribution according to the advantage of the Copula function in reflecting the actual distribution of random variables associated. With the use of Monte Carlo simulation technology, the minimum conditional value at risk（CVaR） of the different confidence level and the composition of assets related are also analyzed. From the evidence, investors cannot only choose more stable assets, but also is conducive to the risks of dispersion and supervision. This paper presents evidence that the use of the model of this paper will lead investors choose more stable assets, which will also be conducive to the overall portfolio risk dispersion and supervision.