Author(s): Upasana D Bhopala, MG Dhandhalya, Mohit Kumar and Bhoomi Suthar

Abstract: Cotton is one of the most important fibre crops playing essential role in the history of mankind and civilization. India is the largest producer of cotton followed by China. High volatility in prices of agricultural commodities over the time generate the need for the accurate price forecasting for policy makers for effective planning and monitoring. For the present study, monthly time series data on cotton prices in six major national markets were collected from January, 2001 to December, 2021. ARIMA and SARIMA model was employed to predict the future prices of cotton using the SPSS software. By comparison of forecast performance i.e., Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), best fit model was used for prediction of future cotton prices. Results of cross validation of each fitted model was compared and model with lesser percentage of absolute error margin was selected as best fit model. On the basis of comparison, SARIMA model was used to predict cotton prices in Adoni and Sendhwa market, whereas in Parbhani, Budalada, Rajkot and Gondal market, ARIMA model was found better for forecast. Results of price forecasting revealed that in Parbhani, Budalada, Sendhwa and Rajkot markets, the predicted values were found higher in November with Rs. 9380.47/qtl, Rs. 8440.35/qtl, Rs. 8809.70/qtl and Rs. 8927/qtl, respectively. Moreover, in Gondal market, the predicted prices were found higher in month of December.

DOI: 10.22271/maths.2023.v8.i6b.1480

Pages: 172-176 | Views: 116 | Downloads: 14

Download Full Article: Click Here