Uma análise preditiva para o uso do instrumento antidumping na Argentina

Abstract

This paper analyzed the influence of macroeconomic factors on the opening of antidumping cases (AD) in Argentina. The resulting model was used to predict the use of AD in this country. Quarterly data, from 1995 to 2010, were used to this end. The models were estimated by Ordinary Least Squares (OLS), Poisson and Negative Binominal. The results revealed that the lagged variables can cause greater impacts on the opening of AD cases than contemporary variables. In addition, a worsening in current account, a currency devaluation or an increase in GDP from Argentina, Brazil and the rest of the world could reduce AD cases. As for predictions, it was found that Poisson was the closest to the real average number of AD measures initiated by Argentina. However, the smaller residuals were produced by OLS estimation. In general, the predictive capability of these models reveals a misuse of AD from Argentina.