Bank-Specific and Macroeconomic Determinants of Non-Performing Loans in Gulf Cooperation Council Countries: Evidence from Extreme Bounds Analysis

The determinants of bank credit quality have been studied extensively, yet much of the existing evidence rests on a single regression specification, so a variable’s apparent significance may be conditioned on which controls a researcher chooses to include. We confront this problem directly for the Gulf Cooperation Council (GCC) countries, providing a robustness analysis of non-performing loans (NPL) determinants for the region’s banks. We employ a balanced panel of 45 listed commercial banks drawn from all six GCC countries over the period 2010 to 2024. We examine fifteen bank-specific and four macroeconomic potential determinants of NPLs, utilizing two variants of extreme bounds analysis (EBA), namely, the strict criterion of Leamer and the more lenient criterion of Sala-i-Martin, estimated within a panel fixed-effects framework. The findings show that of the nineteen determinants routinely cited in the literature, seventeen prove fragile once their coefficients are tested across the full range of possible model specifications. None survives Leamer’s strict criterion, whereas Sala-i-Martin’s less restricted test suggests that only two variables are robust, namely, asset quality (loan intensity), which enters positively, and capital adequacy, which enters negatively, while all four macroeconomic variables are fragile on both tests. For regulators, bank-level balance sheet indicators, especially loan intensity and capital adequacy, offer a more robust starting point for NPL early-warning and stress-testing frameworks and complement macroeconomic forecasts.