摘要：Portfolio theory is used to measure the expected return and risk on the basis of the return ratio, but in fact there is always excessively high or low return ratio caused by some short-term fundamental good or bad news in the history data of return ratio. We introduce the robust statistic idea into the portfolio theory in this paper, thus reduce outliers’ influence on portfolio decision in the history data of return ratios, and bring back the portfolio on its long-term investment value track. We focused on the robust estimate method and apply them to solution processing in the portfolio model and obtained good results.
