Handling asymmetries in the trade balance

The existing literature in the US-Mexico trade balance reports mixed evidence on the effects of exchange rate shocks. In this paper we propose a new approach of studying exchange rate effects that allows for asymmetries in the balance of trade. Our results show that using either linear modelling or non-linear modelling and zero threshold, there is no evidence of co-integration for the US-Mexico trade. However, using a quantile analysis and median threshold, for the first time in the literature on trade balance, we obtain evidence of co-integration. Indeed, our findings show the existence of a long-run relation between the trade balance and its determinants except for periods of significant deficits, where the trade balance is detached from underlying fundamentals and follows a random walk. A USD depreciation, especially a small one, is effective in the short run in improving the trade balance, while in the long run, any depreciation worsens the trade balance. These findings hold due to the complementarity of the US and Mexican economies and the change in the structure of the Mexican economy towards higher value-added products that have led to income effects outweighing in the long run, any intending short-run effects, of exchange rate depreciation.