Application of Liquidity Ratios in Predicting Corporate Financial Crisis; Comparison of Support Vector Machine Model and Neural Network in Cement Industry

Abstract

Investors need to evaluate and analyze financial statements for reasonable decisions. Using financial ratios is one of the most common methods. The main purpose in this research is to predict the financial crisis of companies using liquidity ratios. Support vector machines and neural network of back propagation error are compared. In this study, liquidity ratios have been investigated over the period of 93-89. The research method is quantitative and qualitative and casual comparative. The results indicate that the accuracy of the neural network with a significant difference of 0.001 in year t with a significant difference of 0.005 in year t-1 and with a difference of 0.30 in year t-2 is greater than the support vector machine. This shows that the neural network has the ability to predict correctly up to 2 years before bankruptcy. Also, the results showed that the ratio of capital to total assets had the greatest effect on bankruptcy prediction.