An assessment of the reliability of early warning models on the example of small and medium-sized enterprises in the industry and services sector

Abstract

Background: An assessment of the prognostic reliability of early warning models.
 Research purpose: To determine which of the early warning models has the highest prognostic reliability.
 Methods: Literature review and verification of selected early warning tools. Fourteen models were used in the study: eight discrimination models, five logit models, and one scoring model. The study was based on financial data from a sample of 120 entities, classified as small and medium-sized enterprises from the industry and services sector.1
 Conclusions: The research results prove the effectiveness of early warning models in assessing the financial condition of enterprises. All models accurately diagnosed over 60% of the entities surveyed. The highest scores were achieved using the scoring model and discriminative models by Jagiełło. The results that show the percentage of accurate diagnoses will support the future selection of models on choosing the right tools to examine enterprises’ financial standing.