OPTIMAL PORTFOLIO DIVERSIFICATION AND THE EFFECTS OF DIFFERING INTRA SAMPLE MEASURES OF RETURN
证明使用平均数据构建的股票收益率序列会导致均值、方差和协方差的估计偏差，进而扭曲均值-方差有效前沿和投资比例，并基于国际投资组合分散化研究量化了偏差大小及修正效果。
It is shown that share returns series constructed from averaged data lead to biased estimates of the mean, variance and covariances of the underlying returns series. The computed variances and covariances will be only 2/3rds of their true values, whilst the mean will be reduced by 1/6th of the true variance. It is shown that this leads to distortions in the mean‐variance efficient frontier and the implied investment proportions. A number of studies of international portfolio diversification have used averaged data and, therefore, an empirical study of IPD is reported which investigates the magnitude of the biases and the extent to which they can be corrected.