Zeszyty Naukowe Małopolskiej Wyższej Szkoły Ekonomicznej w Tarnowie (Jun 2020)

Evaluation of the effectiveness of early warning models on the example of enterprises operating in SEZ

  • Justyna Chmiel,
  • Karolina Kozioł,
  • Rafał Pitera

DOI
https://doi.org/10.25944/znmwse.2020.02.6981
Journal volume & issue
Vol. 46, no. 2
pp. 69 – 81

Abstract

Read online

The article aims to verify the effectiveness of selected 10 models of discriminant analysis on the example of 30 enterprises operating in special economic zones: Mielec and Tarnobrzeg. The methodology applied for the research was an analysis of existing data and the use of discriminant analysis methods such as systematic review of literature, analysis of public data of the Ministry of Economy and financial data of enterprises (primarily financial statements). Analysis of companies belonging to the Mielec zone, SEZ Euro-Park Mielec and Tarnobrzeg Euro-Park Wisłosan was conducted on a sample of 30 enterprises, including 15 bankrupt and 15 termed “healthy”. The time horizon of the research was 2009–2017, verification was based on 10 early warning models. The conducted analyzes showed that some models correctly reflect the financial situation of the surveyed enterprises (e.g. Artur Hołda’s model—73.3% accurate forecasts), they also revealed the need to use multiple discriminant analysis models to thoroughly analyze the company’s financial situation—using only one lead model maybe to draw incorrect conclusions. The use of discriminatory models to assess the financial situation of enterprises is in many cases based on early warning methods. These methods are characterized by both advantages and certain limitations; one of the disadvantages is the rapid decline in the effectiveness of models due to constant changes in the economic conditions of market players. That is why models created several years ago may be less effective than newer methods. As for the advantages, it should be emphasized above all the simplicity of the use of such tools and unambiguous results—which in comparison to, for example, traditional indicator analysis, allow to avoid errors in the interpretation of results.

Keywords