IS LOGISTIC REGRESSION RELIABLE IN BANKRUPTCY PREDICTION?

Abstract

Background: The fundamental foundation of financial and economic analysis is the implementation of financial and statistical models in business entities. Financial indicators help assess the company's financial situation and, at the same time, represent the most used tool in the detection of bankruptcy. A still significant method is the logistic regression, which ranks among classic statistical methods. Aims: The aim of the paper is to propose a scoring model for non-financial corporations in the electrotechnical industry and to estimate the probability of bankruptcy of enterprises in the electrotechnical industry of the Slovak Republic. Sample: The research sample consists of 1,241 companies operating in the electrotechnical industry of the Slovak Republic, according to SK NACE 26, 27 in the year 2020. Methods: We used logistic regression and designed a logistic model. Financial data were sourced from The Register of Financial Statements of the Slovak Republic. Results: In the bankruptcy model for the electrotechnical industry, the estimated odds ratios showed that the chances of bankruptcy significantly reduce the financial indicators EBITDA/T, ROA, L2, and NWC/A. When these indicators increase, the company's probability (or chance) of bankruptcy decreases. Conclusions: The estimate of a risky company or the probability of its bankruptcy is always very important. Therefore, it seems appropriate to look for applicable models. Assessing the financial health of business entities using various models is an important area not only in scientific research but also in the practice of business entities. Implications: The benefit of the paper is the construction of an early warning model for the Slovak electrotechnical industry. We constructed an appropriate model and evaluated the financial situation in the selected branch of industry.