摘要
采用不同时间间隔的金融资产收益率数据可估计出不同的收益相关矩阵,为从这些收益相关矩阵中选出最优收益相关矩阵而建立了最优收益相关矩阵选取方法,即以随机相关矩阵特征值分布作为收益相关矩阵特征值分布的参照分布,使不同收益相关矩阵的特征值分布与之进行对比,具有最大特征值分布偏差的收益相关矩阵为最优收益相关矩阵。同时选取沪市44只股票做了实证检验,检验结果表明,应用上述方法确定的收益相关矩阵在进行投资组合时可表现出最优的有效前沿。这说明采用该方法所选取的最优收益相关矩阵能真实反映资产收益率之间的相关性。
Different return correlation matrices may be obtained by different time-interval return of financial assets. In this paper, the method of selecting the optimal return correlation matrix is established as follows: the eigenvalue distributions of the different return correlation matrices are compared to one of the random correlation matrices and the matrix with the maximal distribution deviation is considered as the best one. 44 stocks from Shanghai Stock Market are selected for the empirical tests. The results show that the optimal return correlation makes the portfolio be in best effective frontier in assets allocation. This indicates that the genuine relevance between the capital return is well reflected by the optimal return correlation matrix mentioned above.
出处
《证券市场导报》
CSSCI
北大核心
2008年第5期65-68,共4页
Securities Market Herald
基金
国家自然科学基金资助项目
项目批准号70471050
关键词
收益相关矩阵
投资组合
投资策略
投资理论
return correlation matrix, portfolio, investment strategy, investment theory