Analysis of Predictors in Bankruptcy Prediction Models for Slovak Companies

Abstract

The creation of bankruptcy prediction models is in the last years a topic, which much attention has been dedicated to. Researchers and economists in many countries have created prediction models which are useful for failure prediction of companies in that country. These prediction models used various financial ratios or other predictors to reach the best bankruptcy prediction. The effort of researchers leads to build a strongly prediction model that is able to predict a bankruptcy of companies or can, with some probability level, classify the companies into a group of prosperous or a group of non-prosperous ones. Previous works have shown that these models are then less effective in application in another country or in another time. Our work will lead to a creation of bankruptcy prediction model for Slovak companies. One of the first steps in this process is to choose an appropriate set of predictors, such as financial ratios of companies or characteristics of the environment, in which the company operates. For this purpose we do the preliminary statistical analysis of financial ratios of real Slovak companies. This analysis is made separately in different regions of Slovak Republic in order to analyze which regions are sufficiently similar in their characteristics and therefore could be analyzed together and, on the contrary, which regions are so different that we have to analyze them separately. Then, we can apply cluster analysis on basic statistical characteristics of financial ratios and get the clusters of Slovak regions that are for predicting bankruptcy appropriate to be analyzed together. This result will be very useful during the process of failure prediction model development in the future.