Taxa real de câmbio e diferencial de produtividade : testes do efeito Balassa-Samuelson para a economia brasileira
Nenhuma Miniatura disponível
Data
2026-02-20
Autores
Cordeiro, Felipe Eduardo Megale
Título da Revista
ISSN da Revista
Título de Volume
Editor
Universidade Federal do Espírito Santo
Resumo
This article investigates the empirical validity of the Balassa-Samuelson (BS) effect in the Brazilian economy over the period 2003–2023. Employing the BEER framework, as proposed by Clark and MacDonald (1998), and Autoregressive Distributed Lag (ARDL) models with the Bound Test, following Pesaran et al. (2001), the study examines the long-run relationship between the real exchange rate (RER) and productivity differentials using alternative proxies. The results provide evidence consistent with the BS effect, albeit sensitive to the measurement employed. The proxy based on Total Factor Productivity (BSPTF) yields coefficients in line with the theoretical hypothesis—according to which productivity gains in the more productive economy (USA) relative to the less productive economy (Brazil) lead to a real depreciation of the brazilian exchange rate—whereas the labor productivity proxies (BSPROD and BSREPR) display signs inconsistent with the theory, in line with recent findings in the international literature (GUBLER; SAX, 2014, 2019). Furthermore, the findings indicate the predominance of financial factors, such as global risk (VIX) and the international investment position (IIP), over real fundamentals in determining the RER. It is concluded that, during the period under analysis, the risk premium exerted a stronger influence than productivity differentials, suggesting a high sensitivity of Brazil’s exchange rate dynamics to external volatility and institutional fragilities
Descrição
Palavras-chave
Balassa-Samuelson , Produtividade , Taxa real de câmbio , Macroeconomia , Câmbio , Behavioral Equilibrium Exchange Rate (BEER) , Autoregressive Distributed Lag (ARDL) , Productivity , Real exchange rate