Modelagem e previsão da volatilidade dos retornos do café arábica produzido no Brasil

• Autor(a) principal: SOUZA, Marcela Verônica Alves de
• Data de Publicação: 2005
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UFRPE
• Texto Completo: http://www.tede2.ufrpe.br:8080/tede2/handle/tede2/5198
• topic: Café; Previsão de volatilidade; Risco de mercado; Coffee; Volatility forecast; Market risk; CIENCIAS EXATAS E DA TERRA::PROBABILIDADE E ESTATISTICA

The main aim of the present dissertation is to present a modeling methodology and forecast of the volatility of the Arabic coffee, with the intent not only to predict and explain its conditional heteroskedasticity, but also to obtain the market risk for its investments. For modeling and prevision of volatility in this dissertation, the temporal series of Arabic coffee returns from September 1996 to December 2004, was used. The ARIMA models were analyzed, together with non-linear ARCH models, and their generalization GARCH. The results predict high volatility and investment loss for up to nine steps ahead. The results also indicate that the GARCH model, adjusted to the residuals of the AR model, presents the largest forecast power one step ahead