Beyond one-size-fits-all regulation: Do banking resilience patterns vary across countries?

Purpose International banking regulation has long assumed resilience patterns work identically across contexts. This study tests whether resilience patterns of association are context-dependent using data from 142 countries classified under the IMF World Economic Outlook (April 2026) taxonomy of Advanced Economies and Emerging Market and Developing Economies (EMDEs). Design/methodology/approach A multidimensional Banking Resilience Index is constructed from CAMELS components for 142 countries over 2000–2023. Fixed effects models examine associations between resilience and future capital adequacy, asset quality and profitability. Dynamic fixed-effects models account for persistence and macroeconomic conditions. Temporal falsification tests assess forward-looking information. Interaction models evaluate economic-group heterogeneity. Crisis-period change analysis examines three major crises. Findings Fixed effects show associations between resilience and future asset quality (β = −0.095, p = 0.007), profitability (β = 0.034, p < 0.001) and capital adequacy (β = 0.051, p = 0.007). Dynamic models render these insignificant. Temporal falsification confirms persistence. Interaction models show no Advanced–EMDE difference (p = 0.283–0.961). Advanced Economies recorded larger resilience improvements during COVID-19 (4.50, p < 0.001) and the Eurozone Crisis (1.33, p = 0.030). Research limitations/implications Country-level data mask bank-level heterogeneity. Resilience is a persistent summary measure, not a forward-looking early-warning tool. Practical implications Findings support a harmonized core framework with crisis-contingent responses: capital-focused supervision for advanced economies and liquidity and earnings buffers for EMDEs. Originality/value This study provides the first large-scale test of the institutional-contingency hypothesis in banking resilience.