Forecasting financial distress for French firms: a comparative study

The aim of this paper is to compare three statistical methods predicting corporate financial distress. We use discriminant analysis, logistic regression and random forest (RF) methods. These approaches are evaluated based on a sample of 800 companies, composed of 400 healthy companies and 400 failed companies. This study covers the period from 2006 to 2008 using 33 financial ratios. The results show the superiority of the RF approach, which gives better results in terms of classification. It allows for better forecast accuracy because it minimizes type I and type II errors.