Forecasting agricultural crops prices: case study of cotton, rice and saffron

The aim of this study was to forecast nominal and real price of some agricultural crops including Cotton, Saffron and Rice. Initially the stationary of the series was tested, then in order to investigate whether series are stochastic, nonparametric test of Vald-Wulfowitz and parametric test of Durbin-Watson were applied. Based on the tests results, all of the selected nominal crops prices and real price of Cotton were found non stochastic and predictable. The study period covers 1971-2005. Models applied to forecast are ARIMA, Single and Double Exponential Smoothing, Harmonic, ARCH and Artificial Neural Network. Based on the lowest forecasting error criterion, ARIMA forecasted nominal prices of Saffron and Rice with lowest forecasting error. In the case of nominal and real price of Cotton prediction of the lowest forecasting error belongs to Artificial Neural Network and Harmonic models respectively. The lowest forecasting errors for nominal prices of Cotton, Rice and Saffron as well as of the real price of Cotton are 30.08, 11.14, 4.46 and 14.78 percent respectively.
JEL Classification: C22، C32، C51، C53، D12، Q11