Monetary Policy and Credit Risk in Azerbaijan: An ARDL Bounds-Testing Approach

AHEAD OF PRINT
Download

Summary

This paper examines the long-run and short-run relationship between credit risk and monetary policy in Azerbaijan using quarterly data over 2005Q1–2025Q4. Credit risk is measured by the ratio of overdue loans – loans overdue by more than 30 calendar days, following the Central Bank of Azerbaijan (CBAR) definition – to total loans, while the monetary policy stance is proxied by broad money (M3), which is more informative than the policy rate under the country’s de facto fixed exchange-rate regime. Applying the autoregressive distributed lag (ARDL) bounds-testing approach and its nonlinear (NARDL) extension, the study finds a stable long-run (cointegrating) relationship between credit risk and monetary aggregate. Monetary expansion is associated with lower credit risk in the long run, while the credit-to-money ratio emerges as a strong predictor of credit risk. The error-correction term is negative but small, reflecting the high persistence of problem loans. Short-run asymmetry is significant, capturing the sharp impact of the 2015 monetary contraction and devaluations. The findings carry macroprudential implications for monetary and financial stability policy.