摘要
由于经典均值-方差(Mean-Variance)模型对输入参数的变动非常敏感,输入参数的微小波动会造成最优化结果的巨大偏差。投资组合再抽样方法通过模拟输入参数以增加样本的信息量,能够大大降低MV模型对输入参数的敏感性。本文在对沪市A股10只股票进行再抽样方法及其改进研究中证明:在均值-方差分析中运用再抽样方法进行组合权重修正,能够取得更接近于真实有效边界的再抽样有效边界,是对均值—方差模型有效改进。
Since classical Mean-Variance(M-V)model is sensitive to the change of input parameter, a small change of input often leads to a big change in the weight of optimized portfolios. The method of resampling can add more information to the M-V model and decrease its sensitivity. The empirical research in Shanghai stock market proves that the method of resampling, together with its improvement, are more effective than classical M-V analysis.
出处
《管理评论》
2006年第5期3-8,共6页
Management Review
基金
"新世纪优秀人才支持计划"项目资助(教技司[2005]2号)。
