Genetic Algorithms in Determining Optimal Capital Structure of Firms Accepted in Tehran Stock Exchange

Authors

Abstract

Capital structure or a mixture of debts and equity constitutes an effective factor on shareholder wealth maximization, as the main goal of the firms. This research reviews the correlation between capital structure and profitability of 300 accepted firms in 12 industries and verifies confidence and significance of relationship between these two variables, determining the optimal capital structure in total levels of firms and various industries. The correlation results indicate that the relationship between capital structure and profitability depends on definition of profitability variable. Due to the significance of relationship between capital structure and return on assets (ROA) in total levels of firms and various industries, this variable has been used as a criterion of profitability and determining factor of optimal capital structure in genetic algorithms. For modeling input (capital structure) and output (ROA) data, support vector (SV) regression is used, and for optimization of capital structure, genetic algorithms are applied. The results obtained by genetic algorithms application, indicate that the maximum of profitability is achieved in lieu of using less financial leverage (debt). The finding agrees with correlation results indicating a negative relationship between capital structure and ROA.

Keywords