DETECTING FINANCIAL DISTRESS

This paper examines two types of statistical tests, which are multiple discriminant analysis (MDA) and the logit model to detect financially distressed companies. Comparison between the two statistical tests is implemented to identify factors that could differentiate financially distressed companies from the healthy company. Among the fifteen explanators, MDA shows that current ratios, net income to total asset, and sales to current asset, are the indicators of financially distressed companies. Other than net income to total asset, the logit model provides two different ratios which are shareholders' fund to total liabilities, and cash flow from financing to total liabilities, to identify financially distressed companies. It was found that the logit model could accurately predict 91.5% of the estimation sample and 90% of the holdout sample whereas the discriminant model shows an overall accuracy rate of 84.5% and 80% for the estimation and the holdout sampel respectively.