Resumo

GUTIERREZ, Raúl De Jesús; GONZALEZ, Reyna Vergara  e  CARRENO, Miguel A. Díaz. Volatility forecasting for Mexican crude oil market in the presence of asymmetric effects. Cuad. Econ. [online]. 2015, vol.34, n.65, pp.299-326. ISSN 0121-4772.  https://doi.org/10.15446/cuad.econ.v34n65.48702.

This article assesses the predictive ability of a GARCH-class model family, which can be used to forecast the volatility for Export Mexican Blend crude oil returns over the January 2, 1989 to December 30, 2011 period. The empirical results show a high degree of persistence and the presence of asymmetric effects in the volatility. Although forecast bias test results show that the IGARCH model yields the best fit to capture the infinite persistence of the shocks on the volatility. These findings are not supported by the robust loss functions and Diebold-Mariano test since the GARCH and EGARCH models provide the best accurate out of sample volatility forecasts of the crude oil returns for the 1 and 5 day horizons.

Palavras-chave : Crude oil; volatility; GARCH-class models; optimal forecasting tests.

        · resumo em Português | Espanhol | Francês     · texto em Espanhol     · Espanhol ( pdf )