Identifying Failing Companies: A Re-Evaluation of the Logit, Probit and Mda Approaches
Posted: 17 Mar 1998
Date Written: February 1998
Abstract
This paper examines the causes of bankruptcy for a sample of 949 publicly quoted companies in the UK between 1987-94. The most important determinants of bankruptcy are profitability, leverage, cashflow, company size, industry sector and the economic cycle. In contrast to previous studies, this paper argues that well specified logit and probit models can identify failing companies more accurately than discriminant analysis (DA). Tests for mis specification also reveal that cashflow and leverage have non linear effects on the probability of bankruptcy. Taking account of these non-linear effects significantly improves the explanatory power of the bankruptcy model.
JEL Classification: G33
Suggested Citation: Suggested Citation