Evidence of an Asymmetric J-Curve in Brazilian Net Exports by Technology Intensity

Recent international literature has increasingly emphasized the asymmetric effects of exchange rate movements on trade flows, highlighting that currency appreciations and depreciations may affect economic outcomes differently. A growing strand of research has examined how these effects vary across technological intensity levels in international trade, suggesting that the J-curve phenomenon may depend on the technological composition of trade, particularly in Brazil. Motivated by evidence that linear exchange rate models may be overly restrictive, recent studies have adopted nonlinear frameworks to capture distinct adjustment processes associated with appreciations and depreciations. In this context, this study investigates the occurrence of an asymmetric J-curve in the Brazilian trade balance disaggregated by technological intensity, using monthly data from 2000 to 2024 and the Nonlinear Autoregressive Distributed Lag (NARDL) approach proposed by Shin, Yu and Greenwood-Nimmo (2014), alongside the technological classification of Galindo-Rueda and Verger (2016). A distinctive feature of this taxonomy is the inclusion of non-manufacturing economic activities. In the Brazilian case, the trade balance by technological intensity is characterized by an export structure dominated by basic products combined with a strong reliance on imports of high-technology-intensive goods. The empirical results provide evidence of long-run nonlinear exchange rate effects in three of the five technological categories analyzed (High, Medium-Low, and Low Technology). Incomplete J-curve dynamics are observed for High, Medium, Medium-Low, and Low Technology goods. In addition, differences between depreciations and appreciations in terms of magnitude and statistical significance support the presence of nonlinear exchange rate effects on the trade balance, depending on the category of goods considered. Moreover, for the High-Technology sector, exchange rate appreciations are found to be more beneficial to the trade balance than depreciations, allowing for a partial rejection of the Dutch disease hypothesis for technologically intensive goods. By employing a nonlinear framework, this paper contributes new empirical evidence on the asymmetric effects of exchange rate movements and underscores the importance of the real exchange rate in shaping the trade balance when analyzed through the lens of technological intensity in the Brazilian economy. It also highlights the need for stronger policy coordination to support high-technology sectors.