© 2000 John Petroff

5)- Insolvency predictors

Because of the importance of assessing correctly potential default of companies, several authors have proposed individual ratios or combinations of ratios as predictors of financial difficulty.

Hickman found that fixed charges coverage (labeled times-charges-earned ratio) was correlated with default rate on bonds between 1900 and 1943. If the ratio was 3 or above, the default rate was 2.1%. If the ratio was below 1.4, the default rate rose to 69.1%.

Beaver found that the ratio of cash flow to total debt was most accurate in predicting bankruptcy and bond default up to five years ahead. Capital structure ratios were also useful, while liquidity ratios somewhat less.

Altman's Z discriminant function is one of the best known predictors. It is the combination of
- working capital to total assets (percent)
- retained earnings to total assets (percent)
- EBIT to total assets (percent)
- market value of equity to book value of debt (percent)
- sales to total assets (times)
Values of Z above 2.675 predict no bankruptcy, while values below 2.675 predict bankruptcy. The model gives accurate predictions for 95% of the firms included in the sample used. A subsequent modification known as the ZETA model, included additional ratios such as log of times-interest-earned and current ratio, as well as total assets, but the exact calculation is proprietary and has not been divulged publicly.

One should be careful in placing too great importance on such predictors. They definitely can save a lot of time in screening when looking for one or a number of firms that deserve further analysis. The detrimental effects of poor performance and excessive debt are certainly known to management of companies, as it is to researchers. The difference is that management can take action. Discovering whether management is devising an appropriate strategy is far more important and rewarding than reaching quick conclusions to accept or reject a prospect.

See review questions Q-11C5.1 through Q-11C5.5.

See research assignment R-11C5.1.