协方差矩阵退化情形均值-CVaR模型的有效边界

The Efficient Frontier Feature of Risky Assets with Singular Variance-covariance Matrix

本文利用CVaR方法代替方差或CVaR来度量风险,建立了均值-CVaR模型,首先利用等CVaR线的方法研究了包含无风险资产的均值-CVaR模型的有效边界,然后在无套利假设下研究了当风险资产的协方差矩阵是奇异时的均值-CVaR模型,并得到了正态情形下模型的有效边界及其解析表达式.

This paper establishes the Mean-CVaR model by replacing variance or VaR with CVaR as a measure of risk. First the article studies the efficient frontier feture of risky assets with risk-free securities portfolio in Mean-CVaR model by using the Iso CVaR method. Furthermore, under the arbitrage-free market hypothesis the paper examines the efficient frontier feature and its explicit representation of the Mean-CVaR model, in the case the covariance matrix of the risky assets is singular.