摘要：Short-term spot price forecasting of Brazilian commodities coffee and live ca ttle are usually performed with ARIMA models. Unfortunately, most of these works do not use volatilities models and in the few articles that use it, the volatility is modeling separately. If the data presents heteroskedasticity, such estimation violates the regression model assumption, since the estimated residual variance covariance matrix is not correctly specified. In order to solve this problem, this paper, estimates jointly the ARIMA model and the volatility model (GARCH, GJR and EGARCH). Results show heteroskedasticity problem using only ARIMA to estimate both series. The best models using joint approach for live cattle are ARIMA(1,1,1) with GARCH(2,2), GJR(5,6), and EGARCH(4,5) and for coffee ARIMA(2,1,0) with GARCH(1,1), GJR(1,1,1) EGARCH(1,1,1).
Loading...
