Proposing a model for assessing Herding behavior in the Iranian capital market using meta-heuristic algorithms

The main purpose of this article is to provide a model for assessing the Herding behavior of investors in the Iranian capital market based on fundamental and non-fundamental factors using meta-heuristic algorithms, based on DNA and racial complement calculations. The statistical population of the present study includes all companies that have been active in the Tehran Stock Exchange during the period of 2009 to 2019, and 129 companies were selected as a statistical sample. Fundamental and non-fundamental factors were identified as factors affecting the mass behavior of investors and after collecting data, meta-heuristic algorithms were used to predict the dependent variable. Comparison of the results of meta-heuristic algorithms also shows that the prediction error according to MSE and MAPE criteria in the shrimp batch algorithm is less than that in the three humpback whale algorithms, the improved genetic algorithm and the simple genetic algorithm; The prediction error rate according to the RMSE criterion in the humpback whale algorithm is less than that in three shrimp batch movement algorithms, improved genetics and simple genetics. On the other hand, the execution time of a simple genetic algorithm is less than the other three algorithms. In general, it can be said that the shrimp batch movement algorithm and the humpback whale algorithm are better than the simple and improved genetic algorithm in predicting the mass behavior of investors in the Iranian capital market and have higher accuracy.