Corporate bankruptcy prediction for Indian firms: A model based on financial ratios

Abstract

The prediction that a firm may step into bankruptcy is essential for banks to curb credit risks and make better lending decisions. The approach presented in this paper uses accounting-based data and market-based data to calculate financial ratios which are the main determinants of Non-Performing Assets (NPAs). The study examines NPAs from a few top NPA-contributing sectors in India using data after 2008. The model uses the financial ratios throughout three years by the assumption that bankruptcy can be predicted early. Variables underlying the Financial ratios which showed difference in their trends across the years between NPA and non-NPA firms were used to develop the prediction model. Our study also focusses on using dynamic ratios in bankruptcy prediction model development. The empirical results show that solvency, leverage, and asset turnover ratios have high discriminatory power. We used Fisher’s linear discriminant function and solver optimization technique to build the prediction model. Prediction model made using these ratios can monitor the financial health of a firm. It shows strong predictability in identifying a firm which may become NPA. Our results provide the model with strong predictive ability, developed using only three ratios for the Indian firms.