Analysis and prediction of insolvency in the property-liability insurance industry: A comparison of logit and hazard models

This article compares the performance of the logit and hazard models in predicting insolvency and detecting variables that have a statistically significant impact on the solvency of property-liability insurers. The empirical results indicate that the hazard model identifies more significant variables than the logit model and that both models have comparable forecasting accuracy. We conclude that the combined use of both models provides a more complete analysis of the insurance insolvency problem.