Search for a command to run...
In their review of bankruptcy studies, Ball and Foster [1982] point out that such studies typically used a brute empiricism approach to choose the independent variables (financial ratios) for their models. Since an underlying theoretical rationale was not used to justify the selection of significant ratios, the empirical findings tended to be sample specific and not capable of indicating the most likely predictors of financial distress. As a step toward overcoming this shortcoming we chose a cash-based funds flow model developed in 1972 by Helfert [1982] and suggested in the FASB Exposure Draft [1981] as the basis for our study of bankruptcy prediction. The primary objective of this study is to test such a model by assessing whether cash-based funds flow ratios can adequately classify failed and nonfailed companies and serve as an alternative to financial ratios computed using accrual accounting. As such, our study complements the one by Casey and Bartczak [1985] which investigates cash flow from operations. Our logit findings substantiate Casey and Bartczak's findings that cash flow from operations does not improve the classification results of failed and nonfailed companies. Additionally we found the dividend funds flow component in a logit model was significant in distinguishing between failed and nonfailed companies. In the next section we introduce a cash-based funds flow model to be used for classification purposes. The selection of the companies to be