Methods for predicting the number of bankruptcies in the Ukraine.

Abstract

This article discusses methods for predicting the number of bankruptcies at the macroeconomic level. This indicator is important for the formation of economic policy, but its definition is still poorly understood. The information base of the research grounds on analyzing a number of internal and external factors affecting the financial situation of enterprises. On a base of this factor was formed a system of macroeconomic indicators. Correlation analysis of data for the years 1997-2013 allowed determining the individual indicators impact on a bankruptcy level in the country. For the analysis of bankruptcies are proposed and researched four economicmathematical models using various modeling tools, and a different number of factors. The authors analyzed two linear regression models and two models based on neural networks. The best results in the input data approximation are showed by neural networking seven-factors model. Using these models was made the forecast of bankruptcies level in 2015. Concluded that the number of bankruptcies will significantly increase and exceed the level of 12000 per year.