基于方差或VaR方法度量投资组合风险的不足,在Roli的均值一跟踪误差模型基础上引入CVaR总风险约束,同时考虑到现实交易市场中存在交易费用约束和多元权值约束条件等因素,构建了基于CVaR和多元权值约束下的积极投资组合模型。结合我国股票市场数据采用非线性优化算法对模型进行实证分析,结果表明,交易费用和多元权值约束会显著影响投资组合的风险收益空间;在CVaR总风险可控情形下,无论是样本内还是样本外市场交易数据,本文模型都能获得持续超越基准投资组合的阿尔法收益。
Sue to the deficiencies of Variance and VaR methods for measuring the portfolio risk, we takethe CVaR method to control the total risk of portfolio under the framework of Roll's mean-tracking errormodel. Considering that transaction costs and multiple weights constraints exist in the real market, wepropose an active portfolio model with these factors. With real data of China stock market, we conduct anempirical analysis of the model through nonlinear algorithm. The results show the transaction costs andmultiple weights constraints in the proposed model notably affects the portfolio' s space of returns andrisks. The model can obtain sustained alpha returns beyond the market 's benchmark in both in-sample andout-of-sample tests when the total risk of CVaR is controlled.