Predicting corporate failure: empirical evidence for the UK

Abstract

The main purpose of this paper is the development and validation of a failure classification model for UK public industrial companies using current techniques: logit analysis and Neural Networks. Our dataset consists of 51 matched-pairs of failed and nonfailed UK public industrial firms over the period 1988-1997. Prediction models are developed for up to three years prior to the failure event. The models are validated using an out of sample period ex-ante test and the Lachenbruch technique. Our results indicate that a parsimonious model that includes three financial variables, a profitability, an operating cash-flow and a financial leverage variable can yield an overall correct classification accuracy of 83% one year prior to failure. In summary, our models can assist managers, shareholders, financial institutions, auditors and regulatory agents in the UK to forecast financial distress.

More information

Identifiers

Catalogue record

Export record