中文摘要：

不同于均值-方差（Mean-Variance）模型,均值-条件风险价值（Mean-Conditional Value at Risk,Mean-CVa R）模型不是以投资组合收益的方差作为风险测度,而是使用了能表征投资收益下侧尾部风险的条件风险价值。同样,Mean-CVa R模型存在优化解微权值数目过多的问题,造成操作性下降。针对这些问题,提出了在Mean-CVa R模型引入权值分离性约束,以保证投资权值不低于某一设定的阈值,结合上证50指数股票进行实例分析。

英文摘要：

Different from the Mean-Variance, the Mean-CVaR （Mean-Conditional value at risk） model didn＇t take the variance of the portfolio returns as a measure of risk. However, it uses the conditional value at risk that express the underside of the tail risk of the investment income. Similarly, the optimal solution of Mean-CVaR problem has a lot of micro weight, which results in the decline of operating. The constraint of weight separation is proposed in this paper into the mean-CVaR model, which can ensures that the value of investment weight is not less than a set threshold. Combined with the SSE 50 index stock, we carry on the example analysis.