Abstract: On the foundation of stock portfolio model of tracking error,this paper takes the value at risk (VaR) of portfolio and the uncertainty of future returns to establish the robust optimal portfolio model based on VaR constraints. According to the background of Chinese stock market,we use the linear matrix inequality (LMI) method to carry out a demonstrational computation,and compare the results with the returns of benchmark portfolio and tracking error portfolio model and tracking error robust optimal portfolio model without VaR constraints. Finally,we obtain the conclusion that the robust optimal portfolio model based on VaR constraints is better than other models on the condition of a given stock set.

Key words: portfolio, robust optimization, tracking error, VaR, LMI