Early warning models against bankruptcy risk for Central European and Latin American enterprises

This article is devoted to the issue of forecasting the bankruptcy risk of enterprises in Latin America and Central Europe. The author has used statistical and soft computing methods to program the prediction models. It compares the effectiveness of twelve different early warning models for forecasting the bankruptcy risk of companies. In the research conducted, the author used data on 185 companies listed on the Warsaw Stock Exchange and 60 companies listed on Stock Exchange markets in Mexico, Argentina, Peru, Brazil and Chile. This population of firms was divided into learning and testing setdata. Each company was analyzed using the absolute values of 14 financial ratios and the dynamics of change of these ratios.The author's developed models are characterized by high efficiency. These studies are one of the world's first attempts at comparing differences in forecasting this phenomenon between the regions of Latin America and Central Europe. Additionally, a comparison of the effectiveness of discriminant analysis, decisional trees, and artificial neural networks models was made.