Abstract

Companies persist in encountering difficulties in navigating market conditions. To remain viable in the market, one must adapt rapidly to avoid obsolescence. These characteristics demonstrate increased diversity in the context of emerging markets, such as the Visegrad Group countries (V4). A multitude of financial distress models have been developed within the V4 to date. However, specialists from particular countries have yet to be convened for a comprehensive analysis of these models. The prior assessments overlook several publications. The objective of the article is to provide a novel comprehensive analysis of bankruptcy prediction models developed for the V4 countries and for the whole V4 region. It analyses the scientific publications associated with the development of bankruptcy prediction models from 1990 to 2024. Consequently, 180 works have been examined, encompassing more than 400 models. The review and analysis of bankruptcy prediction studies pertaining to the V4 were conducted by scrutinising the relevant scientific literature. In the study, we selected publications derived mostly from the Google Scholar and ResearchGate databases covering the period from Q1 1990 to Q1 2024. The Kruskal–Wallis test, a nonparametric ANOVA, was employed to analyse the ranks of the measurement values of the ratios. The test was derived from a sample including more than 200,000 companies throughout the period of 2006–2021 from the V4. The results indicate that statistical approaches remain the most prevalent; however, their accuracy is inferior to that of machine learning methods. The number of ratios incorporated in the V4 models fluctuates between three and eight indicators. Nonetheless, only a select few are frequently included in the models, including the total debt ratio (DR), equity ratio (ER), current liquidity ratio (CR), quick ratio (QR), return on assets (ROA) and return on equity (ROE). Thus far, no attempts have been made to utilise text analysis to assess the value of off-balance-sheet information for enterprises operating in the V4. It is noteworthy that, to date, no efforts have been undertaken in Hungary to publish the validation of any previously constructed models by independent academic researchers. The research was executed by specialists from each assessed V4 country, who had previously developed models for their respective countries. Studies published in both English and local languages (Czech, Hungarian, Polish, and Slovak) were considered. To the authors’ knowledge, this is the inaugural comprehensive examination of models developed in the V4. We also checked the statistical significance of the values of the most used ratios in research and models among the V4 countries by using nonparametric ANOVA. Based on the results, we recommend verifying the accuracy of existing models using the most up-to-date large research database, and suggest checking the usefulness of text analysis for assessing the financial condition of companies.