Construction of a Model for Assessing the Probability of Default of Engineering Companies

The master's thesis is devoted to the analysis of default forecasting models in the machine-building sector of the Russian Federation. The object of the study is the machine-building companies not traded in financial markets, engaged in the production of machinery and equipment, as well as mechanical equipment The purpose of this work is to develop a model for assessing defaults on the basis of financial statements, macroeconomic and qualitative indicators. The selection and identification of potentially significant indicators for modeling defaults. In the graduate qualification work 20 thousand companies are investigated on the basis of accounting information for the period 2012-2018. The different forecasting models used are characterized in the work. A large place in the graduate qualification work is given to the construction of econometric methods for assessing defaults using ensemble and simple linear models. According to the results of economic evaluation of the models, the quality of ensemble models is 80%-82%. Graduation qualification work is of interest from the point of view of the use of models for early detection and reorganization of troubled companies. Graduation qualification work consists of an introduction, three chapters, conclusion, appendix, list of references.