不允许卖空情况下均值-方差和均值-VaR投资组合比较研究  

The Comparison between Mean-Variance and Mean-VaR Portfolio Models without Short Sales

作　　者：张鹏[1] 

机构地区：[1]武汉科技大学管理学院,湖北武汉430081

出　　处：《中国管理科学》2008年第4期30-35,共6页Chinese Journal of Management Science

基　　金：国家自然科学基金资助项目(70471077);武汉科技大学校基金项目(2008XY33)

摘　　要：文章研究了不允许卖空情况的均值-方差和均值-VaR两种投资组合模型,并运用不等式组的旋转算法并结合序列二次规划法进行求解。最后,通过实证研究验证了上述算法的有效性。计算结果还表明,在不允许卖空情况下,均值-VaR投资组合的有效前沿为均值-方差投资组合有效前沿的子集;置信度越低,投资者越倾向于选择收益率大而风险也大的投资组合。The paper studied mean-variance and mean-VaR models without short sales respectively, then used pivoting algorithm and sequence of quadratic programming method to solve those models. The algorithms were proved efficient by the empirical research. The result indicated that the efficient frontiers of mean-VaR model were the subset of the efficient frontiers of mean-variance model. The smaller the credit value was, the more the investors were interested in the portfolio that the expected return and variance were all big.

关 键 词：投资组合 不允许卖空 均值-VAR 序列二次规划 旋转算法 

分 类 号：F830[经济管理—金融学]