Dynamic Implications of Fiscal Policy on NPLs: Theoretical Analysis and Panel-Regression Empirics

This paper investigates the interaction between fiscal policy and non-performing loans (NPLs), a nexus often overlooked in banking stability literature. By proposing a generalized theoretical framework that augments the industrial organization (IO) theory of banking with liquidity preference theory, this study explains why a fiscal contraction (an improvement in the primary balance from deficit toward surplus) can decrease NPLs in a bank’s portfolio. Using bank-level quarterly data from Guyana (2009: Q4 to 2024: Q4) and a Panel Autoregressive Distributed Lag Pooled Mean Group (ARDL-PMG) model, we find that a fiscal contraction reduces NPLs in the long run. Specifically, a one-percentage-point improvement in the seasonally adjusted primary balance (as a % of GDP) is associated with a 0.473 percentage point decrease in NPLs in the long run. This finding contrasts with the existing literature, which often suggests that fiscal consolidations increase credit risk. In the short run, however, the results indicate a divergent effect where fiscal contractions lead to a temporary increase in NPLs, with a coefficient of 0.103, likely because of immediate pressure on borrower debt-service capacity. This study contributes to the literature by extending the IO theory of banking to the fiscal policy–NPL relationship in a developing, resource-rich economy. Notably, while higher oil prices and bank efficiency significantly lower NPLs, traditional macroeconomic drivers such as GDP growth, inflation, and the real effective exchange rate—as well as the COVID-19 pandemic—are found to be statistically insignificant in this framework.