Moving in and Out of Financial Distress: Evidence for Newly Founded Service Sector Firms

Abstract

The determinants of transitions between different states of financial distress are analyzed using two versions of Markov chain models: a multinomial logit model without random effects and a multinomial logit model capturing such unobservable factors. The empirical analysis is based on a panel data set containing information on 15,538 East German firms founded between 1994 and 1999. The estimation results indicate that the effect of limited liability depends upon firms' starting state, the existence of corporate shareholders improves firms' financial performance, multiple credit relationships have negative effects and product diversification as well as pos- itive macroeconomic conditions improve firms' financial performance.

Keywords: financial distress, Markov chains, multinomial logit model, simulated maximum likelihood, random effects

JEL Classification: C15, C33, G33