Refining the Best-Performing V4 Financial Distress Prediction Models: Coefficient Re-Estimation for Crisis Periods

Financial distress prediction models have been extensively utilised to assess the financial health of companies. However, their predictive accuracy can be significantly affected by extraordinary economic disruptions, such as the COVID-19 pandemic. Traditional models, particularly those designed for stable economic conditions, necessitate evaluation and potential adaptation to maintain their effectiveness during unprecedented circumstances. This study seeks to evaluate the performance of financial distress prediction models developed by authors from the Visegrad Four (V4) when applied to Slovak automotive companies before, during, and after the COVID-19 pandemic. Initially, the best-performing models from those selected were identified in the pre-pandemic period (2017-2019). The performances of these models were subsequently analysed during the pandemic and post-pandemic periods (2020-2022). Finally, their coefficients were re-estimated to enhance accuracy while preserving the original variables, ensuring the interpretability of any changes. The objective is to identify the models with the highest performance during the pre-pandemic period, assess their reliability under crisis conditions, and suggest improvements through coefficient re-estimation. While the majority of models experienced significant declines in performance during the pandemic, some retained adequate predictive accuracy. The re-estimated coefficients improved the overall accuracy of the models and also enhanced the sensitivity of some, offering stakeholders the option to utilise either the original or adjusted models based on their specific context. To complement the analysis, we also constructed new models for the pandemic and post-pandemic periods, allowing for a more comprehensive evaluation of financial distress prediction under changing economic conditions. This study provides a framework for adapting financial prediction models to unprecedented economic conditions, contributing valuable insights for researchers and practitioners seeking to enhance predictive tools within dynamic economic environments.