PRICE FORECAST FOR MAIZE IN MEXICO

Abstract

Maize (Zea mays) is the most important grain in Mexico because it is an essential part of the Mexican diet and it has economic importance in both livestock and industrial production. Farmers face volatility in grain prices and lack of information because there is no predictive signal of futures price for white maize in Mexico. Under the hypothesis that United States prices have an influence on maize prices in Mexico, the objective of this research was to build white-maize price forecasting estimators with price series from the State of Mexico and Sinaloa. Data were from the period 2000 to 2018 and the Box-Jenkins methodology with the Autoregressive Integrated Moving Average (ARIMA), Vector Autoregression (VAR) and Vector Error correction (VEC) models were used on the Mexican data, and the series of maize futures prices and US physical maize prices. Multivariate models provided forecasts closer to the observed values due to the influence of US prices on Mexican prices. The predictive capacity of the models was evaluated with the mean absolute percentage error (MAPE) and Theil’s U coefficient. The VAR model provided price predictions in Sinaloa with lower MAPE and Theil’s U; whereas with the series of the State of Mexico, the ARIMA model rendered forecasts with lower values of mean absolute percentage error and Theil’s U. Thus, there is influence of US prices on Sinaloa prices; but the use of multivariate models is not decisive to obtain closer forecasts to the observed values.