Default Risk Industrial Models Evaluation and its Application for Anti Crisis Financial Management of a Firm

The issue of default prediction has arisen growing interest for more than 50 years. Predicting of corporate default allows management to take preventive measures of anti crisis management in time and to avoid crisis situations. Today a large number of different techniques are developed: starting from statistical methods and ending with artificial intelligence techniques. Practical applicability of different methods is restricted to initial parameters of their construction. This paper is aimed at investigation the problem of default risk evaluation, developing of a prognostic tool with a high accuracy and consideration the ways of its applicability in anti crisis management. Based on the literature review it seems reasonable to take into account industry effect because of specific risks for each sector. Empirical analysis, which based on the data on Russian companies for 2010-2014 years, confirms initial hypothesis. In this paper the probability of default is assessed by multi period logistic regression models that allow revealing stable in time factors. To improve the accuracy of the models we consider financial ratios selected by principal component analysis together with non financial indicators. Investigation of such industries as manufacturing, agriculture and trading shows that factors influencing the probability of default differ in various sectors. Companies belonging to manufacturing industry are exposed to risks related to the lack of own capital and low profitability of production. As for the agricultural enterprises the most probable risk is in the assets structure. And trading companies often face problems with borrowing. In all the sectors small and medium-sized companies with limited experience in doing business and perform in a highly competitive environment are exposed to default risk mostly.