Abstract
Recently, the topic of forecasting the financial insolvency of an organization based on public open data has become quite popular in domestic analytical science and practice. In general, all models are built according to the same principle: a sample of organizations from the bankruptcy register reveals certain dependencies between individual coefficients that must satisfy certain statistical tests. In fact, the researcher makes extensive use only of the mathematical (statistical) apparatus, not attaching importance to the logic and economic meaning of the represented ratios. Without at all belittling the role of mathematical tools in research of this kind, we note that any model should have a certain economic core that describes the logic and the nature of the influence of the proposed factors on the final result. The article is devoted to testing the most popular forecasting models at enterprises of the real sector in order to study the possibility of using these models in economic practice.