Designing and Explaining the Systematic Risk Estimation Model using metaheuristic Method in Tehran Stock Exchange: Adaptive Approach to the Model of Econometrics and Artificial Intelligence

Systematic risk is always one of the most important indicators that investors and financial analysts attach importance in their financial decision making. The purpose of this research is to provide a new model based on accounting variables for estimating the systematic risk index (β). The period of study is from 2006 to 2015. The statistical population of the research is the companies accepted in Tehran Stock Exchange. Using the Cochran formula, 174 companies are selected as the research sample. For this purpose, systematic risk beta is first calculated through ARFIMA-FIGARCH, and then, estimated models are compared using stepwise regression econometrics (forward selection) and artificial intelligence (through combination of genetic algorithms and flying birds algorithms in selecting effective factors and its modeling by combining and implementing an evolutionary dynamic data estimator algorithm on the above algorithms). In order to analyze the data, three software of Oxmetrics, Eviews, and MATLAB are used. The prediction accuracy of two models based on econometrics and artificial intelligence is evaluated by calculating the correlation coefficient between estimated betas and beta of ARFIMA-FIGARCH. The AI-based model with a correlation coefficient of 94 percent shows a higher predictive accuracy