摘要
本文给出了基于历史收益率数据的均值—平均绝对离差型组合证券投资优化模型。该模型采用收益的平均绝对离差作为风险的尺度 ,可以通过求解线性规划获得最优证券投资组合。在证券收益分布为正态分布时与均值—方差模型的解相似 ,避免了均值—方差模型求解二次规划问题 (尤其在解决大规模的组合证券投资问题时 )的计算复杂性。
The paper raises a mean—absolute deviation porfolio optimal selection model on account of the historical return data. The model uses mean absolute deviation of return as measure of the risk and gains a optimal portfolio by solving the linear programming. If the securitys returns are multivariate normally distributed, its solution is similar to the one of mean—variance models and the computative complication of solving the quadratic programming problom of mean—variance model is avoided. Finally, this paper also raises a portfolio optimal selection model in condition of the trode cost exists.
出处
《预测》
CSSCI
2000年第2期55-57,共3页
Forecasting
基金
国家社会科学青年基金资助项目!(98CJL008)
福建省自然科学基金资助项目!(A97022)