The Prediction of Bankruptcy of Small- and Medium-Sized Industrial Firms

Focusing on predicting the bankruptcy of small and medium size enterprises (SMEs) through the use of publicly available financial data, bankruptcy models are created that attempt to establish the relationship between bankruptcy and a number of financial ratios. Bankruptcy models were developed using methods of multiple discriminate analysis and neural networks. Seventy-three financial ratios were used and classified as activity, profitability, solvency, and liquidity. Various combinations were sought for use with the models. Two hypotheses were tested: (1) In firms heading toward bankruptcy, downward movements can be seen in the activity and profitability ratios, followed by the solvency and then liquidity ratios. (2) It is easier to predict bankruptcy of young firms than established firms. Data on 1,369 SMEs in Belgium was compiled from annual reports. No support was found for the first hypothesis: no fixed order of financial ratios category was predictive. Results confirm the second hypothesis: it is difficult to predict a young firm’s bankruptcy because its bankruptcy is unexpected. Also found that virtually every ratio evaluating a dimension of a firm's financial position has some similar predictive power many years before bankruptcy. Finally, the univariate and multivariate importance of ratio stability were not very high. (TNM)