2024 : 7 : 16
Jamal Ghasemi

Jamal Ghasemi

Academic rank: Associate Professor
Education: PhD.
Faculty: Faculty of Technology and Engineering
Phone: 01135302902


A Comparative Study of the Prediction Stock Crash Risk by using Meta- Heuristic & Regression
Cumulative motion algorithm of particles, genetic algorithm, artificial neural network, stock price risk
Journal International Journal of Finance and Managerial Accounting
Researchers Esfandiar Malekian ، Hossein Fakhari ، Jamal Ghasemi ، Sarveh Farzad


ABSTRACT One of the most important methods of opacity accounting information by management is to accelerate the identification of good news versus delaying the identification of bad news on profits, but there is always a final level of accumulation of bad news in the company, and by reaching that its final level, these bad news will be released, which will lead to a Stock Price Crash Risk. In fact, stock price collapse is a phenomenon in which stock prices are subject to severe negative and sudden adjustments. Accordingly, the first purpose of this research is to model the Stock Price Crash Risk of the listed companies at the Tehran Stock Exchange by using an optimal algorithm The cumulative particles and comparison with the results of logistic regression model. To this, a hypothesis was developed for the study of this issue and the data of 101 listed companies of Tehran Stock Exchange for the period between 2010 and 2014 were analyzed. First, 14 independent variables were introduced as inputs of the combined genetic algorithm and artificial neural network, which was considered as a feature selection method, and 7 optimal variables were selected. Then, using particle cumulative algorithm and logistic regression, predicted The Crashs. To calculate the Stock Price Crash Risk, a Stock Price Crash Period criterion has been used. In The Second Stage, the particle algorithm was used as a feature selection, and this time, to calculate the Crash risk, the NCSKEW criterion was used. Finally, the optimal variables were entered into the Ant Colony algorithm and the results were compared with the multivariable regression. In the second step, MSE and MAE were used to compare the results. The results of the research show that the particle Swarm Optimization and Ant colony are more able than traditional regression (lojestic and multivariable) to predict the Crashs. Therefore, the research hypothesises are confirmed.