The aim of this research was to make predictions for 2007, referring to the price of temporary crops developed in the Rio Grande do Sul state. The data-base of the amout produced of the major grain analyzed were the annual average1995 to 2006 and price produced, from January 1995 to December 2006. To accomplish the forecasts in relation to prices of soybean, rice and sweetcorn in RS state, it was used the Box & Jenkins methodology. To such cultivations, the methodoly showed many significant models, however, through the penalizing criterias, AIC e SBC, in addition to the square average error, it was possible to choose the best model for each serie, being the model ARIMA (0,1,2)(0,1,0) to soybean; the model SARIMA (1,1,0)(1,0,1) to rice and the model ARIMA (2,1,1)(0,1,0) to sweetcorn.
forecast; grains; prices; soybean; rice; sweetcorn
