Abstract
Recently a new direction on the relation between the exchange rate and the trade balance has emerged where the emphasis is on application of asymmetry analysis and nonlinear models. Whereas, the linear models have failed to provide strong support for the J-curve effect, the nonlinear models have proved to be fruitful. We add to this later literature by estimating bilateral trade balance models of Tunisia with her six large partners. The results reveal that exchange rate changes do have short-run asymmetric effects in all six models, short-run adjustment asymmetric effects in three models, and significant long-run asymmetric effects in three models.