On the prediction of financial distress for UK firms: Does the choice of accounting and market information matter?

Research output: Working paper


We assess the contribution of accounting and market-driven variables to the prediction of bankruptcy for UK firms. Using the hazard approach recommended by Shumway (2001), we show that a hazard model that combines both accounting and market information provides more accurate predictions of the probability of financial distress than the accounting ratio-based Z-score model and models exclusively based on market-related covariates. When we decompose the Z-score into its component ratios we find that in the hazard model, half of them do not contribute to the prediction of corporate failure.When we incorporate Z-score as an additional covariate in a model that also containsmarket information, the Z-score does not possess any incremental predictive power. Finally,a comparison of the ability of different information sets (accounting informationonly, market information only and a combination of the two) to predict financial distressboth in-sample and out-of-sample shows that a hazard model combining both accountingand market-driven variables generates the best performance in terms of predictive ability,while a model based exclusively on market-driven variables outperforms that based solelyon accounting variables. Our results suggest that for the UK at least, there are significantgains to be had from predicting financial distress using both accounting and marketinformation.

Bibliographical metadata

Original languageEnglish
Publication statusPublished - 2009