Is There a J-Curve Effect in the Services Trade in Canada? A Panel Data Analysis

The effects of the real exchange rate changes on the sectoral trade balance (as opposed to the balance of aggregate trade) have received limited consideration in the empirical studies. In this paper we focus on services trade and examine the dynamics of Canada’s bilateral trade balance in services with its 53 major trading partners during 1990-2018. To account for slope heterogeneity and cross-sectional dependence we apply mean group (MG), pooled mean group (PMG), common correlated effects mean group (CCEMG), augmented mean group (AMG) estimators to the trade balance equation, alongside the dynamic fixed effects (DFE) estimator. The results provide strong evidence of a short-run deterioration and a long-term improvement of the services trade balance following depreciation in an aggregate panel as well as sub-panels, hence supporting the J-curve effect hypothesis and satisfying Marshall-Lerner condition. At the level of individual cross-sections, the evidence was mixed: a number of economies experienced long-term improvement of the trade balance, albeit without short-term deterioration.