Replication Package for: CAMEL Ratios as Early Warning Signals of Financial Distress in European Retail: Panel Data Evidence with Out-of-Sample Validation

Hand-collected panel dataset of 31 European listed retailers (STOXX Europe 600 Retail constituents), 2015–2024 (309 firm-year observations), built exclusively from publicly available audited annual reports, together with the full Python replication code for the associated article. The dataset includes raw financial statement items (total assets, EBIT, interest expense, net debt, EBITDA, current assets and liabilities, sales, retained earnings, book value of equity) and the five CAMEL ratios used in the study, with the public source documented for every data point. The replication script reproduces all tables in the article: descriptive statistics, correlations, two-way fixed-effects and random-effects panel models, specification tests, subperiod analysis, out-of-sample validation over the 2020–2023 stress period (AUC-ROC = 0.908; Recall = 80.8%; F1 = 0.712), threshold-sensitivity analysis, comparison with the full Altman Z' (1983), construction of the synthetic CAMEL Score, and convergent validation against documented distress events. Author identity withheld pending double-blind peer review; author names and affiliations will be added upon acceptance of the associated manuscript.