A behavioral analysis of financial ratios.

I. INTRODUCTION

Financial ratios have been a topic of a number of empirical studies in the past decades. One group of empirical studies utilizes ratios derived from failed and non-failed firms' balance sheets and income statements. The purposes of these studies are to obtain discriminant functions

with the smallest classification error or logistic functions with the best fit possible and to utilize the estimated discriminant or logistic function for the prediction of corporate failures. Indeed, Altman (1968) has identified several statistically useful financial ratios for predicting corporate failures.

Since then, a number of empirical studies similar to Altman's have been conducted to advance Altman's bankruptcy prediction model in several directions. Such attempts include Edmister's study of small business corporate failure (1972), Casey and Bartczeck's investigation of cash flow ratios in bankruptcy prediction (1985), Ohlson's logit regression (1980), and Zjmijewski's profit analysis (1984). The original discriminant analysis of corporate failure was subsequently updated by Altman et al. into the Zeta analysis (1977). More recently, Altman also developed a computer program to predict corporate failures based on a Recursive Partitioning Algorithm (1985).

Another direction which emerged in research into financial ratio analysis was the classification and reduction of a large number of ratios to a small subset. For instance, Pinches et al. (1975) utilized factor analysis to determine the dimensionality of information contained in financial ratios. Chen and Shimerda (1981) reconciled the findings of past studies and highlighted seven groups of ratios: 1) return on investment, 2) financial leverage, 3) capital turnover, 4) short term liquidity, 5) cash position, 6) inventory turnover, and 7) receivables turnover.

Several studies have shown that financial ratios reflect differences in underlying industry characteristics. Lev (1969) suggests that financial managers strive to achieve the industry averages as a firm's target ratios. Gupta and Huefner (1972) and Johnson (1979) found that financial ratios were significantly different between retailers and manufacturers. More recently, Gombola and Ketz (1983) asserted that these two industries are at opposite ends of the spectrum of ratio characteristics and that firms in all other industries are found between these two extremes.