Abstract
The article explores the possibilities of applying the data of the official financial statements (the Russian national standard) for carrying out financial analysis in order to detect abnormal behavior of economic entities in the course of their activities. Classification of commercial enterprises of the Sverdlovsk region of the Russian Federation was carried out in five branches. In the course of the study, a methodology for financial analysis was developed, including scoring and frequency analysis methods, along with information technologies for processing large amounts of data were also used, in particular, the possibilities of the "nosql" approach. Our empirical research is based on the financial statements of enterprises of the State Statistics Committee. The main conclusions a) are the proposed methodology allowed to identify representative classes for the analysis of anomalous behavior of enterprises; and b) it revealed the relationship between high availability of enterprise registration in the industry and the share of enterprises with super-anomalous financial behavior.