Bankruptcy prediction using Partial Least Squares Logistic Regression

In the current conditions of economy there is an increasing number of companies that are facing economic and financial difficulties which may, in some cases, lead to bankruptcy. This research is motivated by the inadequacies of traditional forecasting models. The Partial Least Squares Logistic Regression (PLS-LR) allows integrating a large number of ratios in the model; in addition, it solves the problem of correlation, and taking into account the missing data in the matrix. Indeed, the results obtained are very satisfactory and confirm the superiority of this method compared to conventional methods. The proposed model gives the opportunity to consider all the indicators in predicting financial distress, the reduction of the environment's uncertainty, the control's improvement and the coordination between the different company stakeholders.