BANKRUPTCY PREDICTION APPLYING MULTIVARIATE TECHNIQUES

Authors

  • Martina Mokrišová University of Prešov in Prešov, Faculty of management
  • Jarmila Horváthová University of Prešov in Prešov, Faculty of management

Keywords:

Bankruptcy, Multidimensional Scaling, Prediction, Principal Component Analysis

Abstract

The paper focuses on the analysis of the corporate bankruptcy prediction using selected statistical multidimensional methods. Existing multidimensional methods are a suitable tool for predicting the bankruptcy of companies, for their graphical representation in space, the identification of clusters of companies with the same bankruptcy preconditions, as well as the identification of bankruptcy factors. The research was carried out on a sample of 343 heat management companies in Slovakia. All of these companies operate local district heating systems. Within this group, there are companies that have a monopoly position in a given geographical area. Of the multidimensional methods, the Principal Component Analysis (PCA) method and the Multidimensional Scaling (MDS) method were used. The resulting graphical representation of both methods yielded significant results. The paper identified the main factors in predicting bankruptcy. It has been found that it is possible to predict bankruptcy of the analyzed sample of companies using three main factors that capture 70% of the information from the applied indicators. It follows that it is not necessary to apply a large number of indicators to reveal the financial situation of companies. In addition, similar characteristics of enterprises make it easier to predict bankruptcy in larger samples.

Downloads

Published

2020-12-31

How to Cite

Mokrišová, M., & Horváthová, J. . (2020). BANKRUPTCY PREDICTION APPLYING MULTIVARIATE TECHNIQUES. Journal of Management and Business: Research and Practice, 12(2), 52–69. Retrieved from http://journalmb.eu/JMB/article/view/10