Asymmetric relationship between exchange rate and inflation in Tunisia: fresh evidence from multiple-threshold NARDL model and Granger quantile causality

This study examines the dynamic asymmetric relationship between the exchange rate and inflation in Tunisia over the period 2007M1–2018M4. In doing so, we employed a multiple-threshold non-linear autoregressive distributed lag model (MTNARDL). Moreover, we perform a quantile Granger causality test to explore the asymmetric causality between exchange rate fluctuations and inflation variations. Our empirical findings assume an asymmetric effect of the exchange rate on inflation in both the long and short runs. Indeed, the estimated coefficients of exchange rate variations vary across quantiles. In other words, inflation is affected differently depending on the magnitude of exchange rate fluctuations. Furthermore, the quantile causality analysis shows evidence of unidirectional causality from inflation changes to exchange rate variations in various quantiles and confirms a feedback effect at high quantiles. Policymakers should consider asymmetries in the exchange rate–inflation nexus to avoid monetary policy distortions, improve the target inflation strategy, and increase the flexibility of the exchange rate policy, which enhances the effectiveness of the monetary policy.