Distress and Restructuring Risk in Polish Cooperative Banks: Research Gaps and a Framework for Early-Warning Modelling
Distress and Restructuring Risk in Polish Cooperative Banks: Research Gaps and a Framework for Early-Warning Modelling

Summary/Abstract: The purpose of the article. The article identifies the research gap caused by the lack of empirically validated early-warning models for distress and restructuring risk in Polish cooperative banks. It argues that models developed for commercial banks or non-financial enterprises cannot be transferred mechanically to cooperative banks because of Poland-specific institutional, regulatory and data constraints, including institutional protection scheme (IPS) membership and local-market exposure. Methodology. The study applies a scoping review approach (PRISMA-ScR), critical literature analysis, document analysis and regulatory-institutional mapping. The review covers bank failure, distress and restructuring models from 1970 to 2025 and was supplemented by a targeted screening of review studies and empirical comparisons on logit, discriminant analysis, tree-based methods, machine learning and systemic early-warning systems. Evidence from corporate-bankruptcy and general financial-distress literature is used only for method taxonomy and validation criteria, not as direct evidence for bank-specific predictors. The article is exploratory: it does not estimate a final predictive model but defines the data architecture, candidate predictors and validation path required for future empirical research. Results of the research. The study provides a publishable research agenda for Poland by: (1) updating the description of the cooperative banking sector; (2) distinguishing bank-distress modelling from corporate-bankruptcy prediction; (3) explaining why prudential regulation is not a substitute for statistical prediction; (4) proposing a Poland-specific variable set for an early-warning system; (5) defining a two-layer data blueprint for a bank-year panel combining public and supervisory/IPS data; and (6) translating the verified review literature into method-selection and validation rules, including interpretability, calibration, class imbalance and false-negative costs.

• Details
• Contents