Nonlinear ARDL approach, asymmetric effects and the J-curve

Abstract

Purpose – Previous research that investigated the effects of currency depreciation on the trade balance assumed that the adjustment of all variables in a given model is in linear fashion. The authors wonder if introduction of nonlinearity in the adjustment of some variables such as the exchange rate can shed additional light on evidence of the J-curve. The new approach also allows to test whether exchange rate changes have symmetric or asymmetric effects on the trade balance. Estimates of a trade balance model for Canada, China, Japan, and the USA reveal that the effects are indeed asymmetric. The paper aims to discuss these issues. Design/methodology/approach – The methodology is based on linear and nonlinear ARDL approach. Findings – When nonlinearity is introduced into testing approach for the J-curve, more evidence is found in support of the J-curve. Research limitations/implications – The models are estimated using aggregate trade flows of each country with the rest of the world, hence they suffer from aggregation bias. Using trade flows at bilateral level and at commodity level are highly recommended for future research. Originality/value – This is the first paper that applies nonlinear ARDL approach to test the short-run and long-run effects of currency depreciation on the trade balance.