Corn Price Modeling and Forecasting Using Box-Jenkins Model

Abstract

Corn harvesting is one of the most complicated problems which farmers need information to make decision prior farming. Corn price is the main factor for farming and there are many factors that affect the corn price vice versa. Knowledge of the factors affecting the corn price and the ability to forecast the corn price in advance would benefit farmers in the context of harvesting. The factors that affect the corn price in Thailand include chicken export rate, corn import rate, weather, soybean price, corn production, stock-to-use, season and planting area. The Cause Tree diagram has been constructed to demonstrate the linkage of such factors and all related data have been collected and analysed by using SPSS software. The Box-Jenkins model has been implemented to establish a time series forecasting model. And performance comparisons among the ARIMA model with Holt-Winters multiplicative seasonal model and Holt-Winters additive seasonal model methods. The results of this research indicated that the corn price can be forecast by using its two lag data with current period soybean price data. The resulting forecasting equation with the ARIMA model generate the lowest errors with Root Mean Square Error (RMSE) at 0.8678, Mean Absolute Percent Error (MAPE) at 12.1009 and Mean Absolute Error (MAE) of 4.7592.