Predictive Power of Financial Risk Factors: An Empirical Analysis of Default Companies

This paper provides empirical evidence on the significance of financial risk factors in predicting default companies. Traditionally, credit decision process is built on accounting ratios derived from financial statements of the borrower. Combining various ratios through application of multivariate statistical techniques and testing their predictive power has been popular in credit risk quantification. Altman's Z-score model is the most acceptable model in this category. In this paper, three forms of Z-score models are applied: The first equation is developed by surveying the internal credit rating models of the Indian banks and the ratios selected are: current ratio, debt-equity ratio, and operating margin. The second equation is similar to that of Altman's (1968) original equation with a slight modification: instead of debt-to-market value of equity, debt-to-book value of equity is considered. The other three ratios of the second equation are working capital to total assets, retained earnings to total assets, and earnings before interest and taxes to total assets. The third equation is called as Altman, Hartzell and Peck's ‘Emerging Market Score Model.’ Except the asset turnover ratio, all the ratios of the second equation are considered. In all the three equations, the coefficients are estimated by using the development sample of 112 companies. The dominant variables discriminating the default companies from non-default ones are: current ratio, debt-equity ratio, operating margin, working capital to total assets, earnings before interest and tax to total assets, net worth to debt, and asset-turnover ratio. The classification accuracy of the second and the third equations is 82 per cent while that of the first equation is only 57 per cent. It implies that the most widely used two ratios — current ratio and debt-equity ratio — are relatively poor in predicting the default companies. Similarly, the ROC accuracy ratio is the highest for Altman's equation whereas the variables considered in internal credit rating models of banks is having a relatively low accuracy ratio. To test the ability of the model in identifying the defaulting companies correctly, an unbiased diagnostic test of the model is conducted on two separate sets of defaulted firms. The results reveal the following : The Altman's model is capable of predicting default in most of the sample companies. The hold-out sample accuracy results show that the selected variables are capable of predicting default. The analysis shows that the financial risk factors being considered by banks in their internal rating models are not very effective in comparison to other two models in discriminating the firms into default and non-default categories. Banks can map the internal ratings with the Z-scores and scale this up to assign various credit ratings. By arriving at the coefficients on the basis of their own database, banks can develop Z-score calculators for various segments of borrowers.