This paper uses the Monte Carlo method to randomly simulate the rate of return series, and then studies the characteristics of the long-term investment rate of return and its variance. The computing results show that, when compared with the short-term rate of return and its variance, both the long-term rate of return and variance decrease with the increasing of investment term; and when the short-term rate of return is a relatively minor positive value, the long-term rate of return, might be significantly negative. In addition, this paper also studies the long-term rate of return and variance with stop-loss strategy, and the results showed that the stop-loss strategy improve the long-term investment rate of return greatly. The relation between, the distribution of short-term rate of return and that of long-term rate of return is complicated, it is improper to directly and blindly judge the long-term rate of return and variance only according to the distribution of short-term rate of return.