Investigating the Effect of Study Period Selection on Solving Portfolio Optimization Based on Different Risk Criteria Using Meta-Heuristic Algorithms

The fluctuation of stock returns in the capital market in different periods is significant and the choice of the study period is very important in solving the stock portfolio optimization problem. hence, in this study, we solved the stock portfolio optimization problem based on the mean-variance model and the mean-percentage of Value at Risk model, which examines the VaR criterion from another perspective. for this purpose, we selected three different study periods in 1393-1398, 1396-98, and 1391-98 in Tehran Stock Exchange as study periods, and used NSGA II and MOABC algorithms. Then while analyzing different periods in terms of the desirability of the proposed portfolios and data quality, we selected the most efficient period. In order to ensure the results, we compared the proposed solutions resulting from solving the studied models were analyzed using the algorithms used in each of the studied periods, independently and also a fixed period. The results of this study indicate that the selection of the study period has a significant effect on the quality of the proposed solutions resulting from solving the portfolio optimization problem. The longer the study period and the less fluctuating the average and variance of stock returns of the companies under review in the said period, the more reliable the selected period is and the higher the desirability of solving the stock portfolio optimization problem.