Forecasting of Day-Old Chick Monthly Price in East Azarbaijan Province: Application of Seasonal Time Series Models

Document Type : Research Paper

Author

Abstract

The broiler industry is one of the main subsectors of agriculture in Iran. This industry is a risky business. Reduction and market risks cause fluctuations in the broiler producer’s income. In this context, price fluctuation of day-old-chick is one of the main sources of the market risk. The objective of present study is modeling and forecasting monthly prices of day-old-chick in East Azarbaijan. To this end, the behavior of day-old-chick prices was specified with considering its characteristics, spatially seasonality by using regression-based model by applying seasonal unit root test (BM test) and seasonal Box-Jenkins models as the primary nominates for forecasting model. Results of seasonal unit root test indicated that the monthly prices of day-old chick follow a non-stationary stochastic seasonal process. Accordingly, the regression-based model is an appropriate modeling framework. While SARIMA is an alternative modeling approach, the RMSE of forecasting error suggested the superiority of the regression-based model over the SARIMA model. Therefore, the estimated parameters of the regression-based model can be used to predict the monthly prices of day-old chick in Iran.

Keywords