Forecasting of Maize Production in Bangladesh Using Time Series Data

Maize has been gaining importance as one of the major grain crops in Bangladesh in recent years. Due to its multiple uses, i.e., food, feed and other industrial uses, maize production and its possible trend have created great interest among the policy planners. This study aims to forecast future production of maize in Bangladesh using both Box-Jenkins autoregressive integrated moving average (ARIMA) and mixed model approach (dynamic regression model) using secondary yearly data, for the growing seasons 1970-71 to 2019-20, published by the Bangladesh Bureau of Statistics (BBS). Our analyses suggest that ARIMA (0, 2, 1) is the best model for forecasting maize production all over Bangladesh. However, when the area of maize is considered the mixed model with ARIMA (1, 0, 0) performs better than the univariate ARIMA (0, 2, 1) model. The length of the 95% confidence interval of the forecast values of the mixed model is smaller than that of the ARIMA model indicating its better predictive performance. These forecast values will be useful for planning resources and making appropriate decisions regarding imports and exports by the government before harvesting.