The Comparison of Genetic and Weed Algorithms in Portfolio Optimization

In this paper, Genetic and weed algorithms are used to solve constrained mean-semi variance portfolio problem. Then AR model and simple average are compared to predict expected return of stocks. 23 active stocks from June 22, 2014 to June 21, 2016 are used as our sample. The results indicate that, weed algorithm despite its longer time consuming has better performance than Genetic algorithm. And AR (2) model has more accurate prediction than simple average in predicting expected rate of return. Finally, we compare expected and real efficient frontier, the results indicate that, in lower risk, AR model has better prediction accuracy. So in that area, we can allocate our asset with higher certainty