Business Failure Prediction Using Discriminant Analysis

Abstract

Business firms are formed to operate into the foreseeable future. However there are cases where a business firm fails to fulfil the objectives for which it was formed or even to survive in the environment in which it was formed. In such cases, the stakeholders in the firm loose substantially, both in monetary and non - monetary terms. But if the failure of a business could . be predicted with reasonable accuracy, then the stakeholders can act in good time to avoid or minimize the possible losses. In this paper, the aim was to formulate a model to predict business failure. Data on some Kenyan companies that failed within the period 1980- 1990, was collected. These companies were matched with comparable successful firms;and ratios from their financial statements subjected to discriminant analysis. The results showed that it is possible to predict failure with upto 90% accuracy two years before the

event. In this case, current ratio, fixed c harge coverage , retained earnings to total assets, return on total assets, return on net worth, average collection period wid sales to total assets; were identified as the the critical ratios that discriminate failed from non failed firms in Kenya.