PREDICTING BANKRUPTCY OF COMPANIES FROM THE LOGISTICS SECTOR OPERATING IN THE PODKARPACIE REGION

Abstract

Research on effectiveness of various concepts for modelling the bankruptcy of companies from the logistics sector is described in this article. In order to present this issue more completely the above-mentioned prediction of possible negative effects for the conducted business activity was conducted for all companies operating in that sector in the Podkarpacie region. The study was supported by the data from the database EMIS (Emerging Markets Information Service ). A wide range of 28 financial indicators was grouped into five groups i.e. liquidity ratios, profitability, debt, performance, and financial respectively. The above mentioned research trial was divided into a group of companies – so-called ill - in relation to which the bankruptcy was declared and healthy ones (of good financial condition). Such an approach allows for a better and right assessment of the methods in modeling bankruptcy. The purpose of this publication was to find factors (models) describing the risk of bankruptcy of enterprises in terms of their effectiveness prediction in one - and two year- horizon. The logistics regression models, classification trees and two lunatics artificial neural networks were applied. A full evaluation of the models application were made in the validation process. The primary tool used in this case to study the effectiveness of models classification are matrices of correct classification. It was made an estimation of the correct and wrong indications in both the above mentioned models. Finally, an assessment of the method was done as well as the overall condition of the logistics sector in the Podkarpacie region.

Keywords

• bankruptcy, logistic sector, modeling, financial indexes