STATISTICAL ASSESSMENT OF THE BONDS DEFAULT RISK BY EXAMPLE PJSC “SOLOMENSKII LESOZAVOD”

Vyacheslav Duplyakin,Maxim Boldyrev,,
DOI: https://doi.org/10.15593/2499-9873/2019.3.09
2019-09-30
Applied Mathematics and Control Sciences
Abstract:The application of the previously proposed methodology for the formation of models for quantitative assessment of the risk of default of companies’ securities, taking into account the characteristics of the financial condition of issuing companies in the reporting and previous periods, is considered. The methodology includes a three-stage selection of working variables using statistical hypothesis testing (Student t-test and Fisher test are applied), correlation and one-way regression analysis. The probability of multicollinearity of variables in a developed model is also monitored. The logit-model and probit-model for assessing the risk of default of Russian companies by bonds, developed using the proposed methodology, are presented. A sample of 96 Russian companies that placed bonds on the stock market was used, 48 of which did not fulfill financial obligations on securities was used for developing the models. The results of a comparative analysis of the reliability of the constructed models and well-known models of other Russian and foreign authors, applied to assessing the risk of default of Russian companies on bonds, are also presented. When conducting a comparative analysis of the reliability of the models, the Student t-test, the Mann-Whitney test, the Kruskal-Wallis test and the Wilcoxon test, as well as the modeling efficiency index, are used. An example of applying the logit-model and probit-model to assessing the risk of default of PJSC "Solomenskiy lesozavod" on bonds is demonstrated.
What problem does this paper attempt to address?