Abstract

The current study investigates the time-varying explicatory potential of accounting data and financial ratios as explainers of bankruptcy by corporations. Using quantitative research methodology, the analysis utilizes a comprehensive panel data set of 65,000 firm-year observations made by Polish manufacturing firms (2000-2013) to estimate logit and probit regression models of the 64 financial ratios, using quite rigorous variable selection procedures in order to guarantee the robustness. The results show a good degree of temporal consistency in bankruptcy explanation: there was a common set of indicators, working capital, operating profit margin, adjusted gross margin, and retained earnings to total assets, that are significant variables throughout the period. More importantly, the explanatory ability of such ratios changes over time, with leverage indicators giving an early warning, whereas profitability and liquidity ratios dominate in the pre-bankruptcy period. These findings illustrate the need for dynamic modelling to clarify bankruptcy more accurately and offer valuable insights for developing early warning systems on a stage-by-stage basis.

Keywords: Bankruptcy, Financial Ratios, Dynamic Analysis, Logit Regression, Probit Regression.