Forecasting wheat prices in India

Abstract

Wheat is the single most important crop after paddy, which plays a vital role in the Indian economy. In this paper, ARIMA model was used to predict the future harvest period prices of wheat before the sowing date to facilitate farmers to make a decision on their acreage under wheat. For this, a monthly data of modal prices from January 2006 to June 2017 were used for model fitting and forecasting. The best fit ARIMA models were selected based on autocorrelation function and partial autocorrelation function at various lags. Forecasting performance of this model was evaluated with 95% confidence interval using criterions like MAE, MAPE, and RMSE. Model parameters were estimated by using the Statistical Packages for Social Sciences (SPSS) software. Empirical results showed that ARIMA (0,1,1) (0,1,1) model was found suitable to forecast the future prices of wheat in India during harvesting season with 95% accuracy level. This model can facilitate the farmers and wholesalers in effective decision making. In India, wheat is harvested during the month of February to May. The forecast shows that market prices of wheat would be ruling in the range of Rs. 1,620 to 2,080 per quintal during harvesting season, 2017-18. The farmers are
advised to take sowing and marketing decision accordingly.