Stability of the Smoothed Taylor Rule in Crisis Periods: Evidence from Bank Al-Maghrib on Quarterly Data (2008 Q1–2025 Q2)

Résumé

This paper examines the stability of the monetary policy reaction function in Morocco by estimating a smoothed Taylor rule for Bank Al-Maghrib using quarterly data over the period 2008Q1–2025Q2 (n = 70). The objective is not to propose a fully fledged structural model, but rather to provide a transparent and reproducible empirical diagnosis of the following question: does the central bank’s response to inflation and the real cycle remain stable during crisis episodes?

Ordinary least squares estimates with HAC standard errors (Newey–West, 4 lags) indicate a high degree of interest rate inertia (ρ = 0.962; p < 0.001) and a positive response to inflation (  = 0.059; p = 0.030). These results are robust to addressing endogeneity through instrumentation (IV/2SLS): inertia remains remarkably stable (ρ = 0.952 under IV), and the response to inflation remains positive and statistically significant (  = 0.062; p = 0.018 under IV). Given the high degree of smoothing, the paper explicitly distinguishes between the implied short-run and long-run responses, with a cautious interpretation of the long-run response to inflation.

Stability is assessed through triangulation: Chow tests at crisis dates (2020Q1, 2022Q1), an endogenous Quandt–Andrews (supF) test with bootstrap, and multiple break tests à la Bai–Perron. The dated tests indicate significant breaks in 2020Q1 (p = 0.0002) and, more prominently, in 2022Q1 (p < 0.001), robust to wild bootstrap inference (p = 0.014 with 999 replications). The Zivot–Andrews test (unit root with endogenous break) detects a level break in 2022Q1 and rejects the unit root null (statistic = −7.53; 5% critical value = −4.80), consistent with a stationary process around a structural break. Regime-specific estimation shows that the inflation response shifts from non-significant before 2022Q1 (  = −0.014; p = 0.48) to strongly positive after 2022Q1 (  = 0.112; p < 0.001), confirming a substantial change in behavior.

The main contribution is policy-relevant: assessing fiscal–monetary coordination (policy mix) without accounting for regime changes in the monetary reaction function may bias the causal attribution of macroeconomic effects during crisis periods.

Keywords: Taylor rule; stability; inertia; structural breaks; Bank Al-Maghrib; Morocco; crises; policy mix
JEL Classification: E52, E58, C22, C52, C26