An analysis of industrial company failure in the UK and Russia for the 1990s

Abstract

This thesis provides an examination of the key determinants of industrial company failure in the UK and Russia, for the 1990s. For the UK, some new empirical evidence, presented for the 1990s recession period, is based on binary logit analyses of a cross-section and unbalanced panel of large quoted companies, using accounting-based indicators. Conventional for cross sectional studies empirical design of modelling the failure determinants separately for various risk-horizons, prior to the event of insolvency, is extended here by allowing for unanticipated changes in the nominal interest rate and in the real exchange rate, and also by controlling for the firm's age effect. We find that cross sectional models, conditioned on changes in overall economic conditions, dominate simpler models, utilising financial inputs alone, for comparisons of ex ante, out-of sample classificatory accuracy. Thus, the UK data suggest that for the years before and during the 1990s recession, shifts in the real exchange rate and rises in the nominal interest rate magnified dramatically the risk of failure of highly geared firms. The estimates from the fixed effects models indicate substantial unobserved heterogeneity across members of the panel and reveal that failing UK companies were less liquid, lacked profitability, and had declining net worth. For Russia, the evidence from binary logit is bootstrap-based and controlled by comparison with a similar random sample drawn for the UK over the recession years 1990-91. The Russian data uncover that, unlike in the UK, gearing and liquidity did not appear to explain enterprise liquidation in the mid-1990s, while lower profitability and smaller size were the key determinants of failure risk.