Some new models for financial distress prediction in the UK

Abstract

In this paper we develop some new models for the prediction of failure in the UK that add to the literature by showing that “dynamic logit” models that incorporate market variables of the form developed by Chava and Jarrow (2004) and Campbell et al (2008) add considerable power to pure accounting-based models. Importantly, we extend the logic of Campbell et al (2008) by showing that incorporating macro-economic variables adds predictive power, both in and out-of-sample, to market-based accounting models. Last, we show that adding industry controls gives a modest improvement to such models for UK firms in the case of a models based on accounting, market and economic variables, but a greater improvement in terms of a pure accounting based model.