Forecasting Futures Trading Volume for Cotton using Vector Auto Regression and ARIMA Model

Abstract

This study involves comparing the forecasting performance of the ARIMAmodel and Vector Auto Regression model. Monthly data of open price, closeprice, mean cash price, open interests, spot price, and trading volume ofcotton futures are used for analysis. Trading volume is taken as a dependentvariable. Data was checked for stationarity using Augumented Dickey-Fullertest based on the p-value. Trading volume was stationary at level and allother variables were stationary at first difference. The optimal ARIMAmodel was selected based on AutoCorrelation Function plot, Partial AutoCorrelation Function plot, and lower Akaike Information Criteria , BayesianInformation Criteria values. The optimal model was found to be ARIMA (2, 0,0). The optimal lag for VAR model was selected based on AIC and BIC. Theforecast is obtained based on the coefficients of the model. The forecastedvalues were compared using RMSE, MAPE, MAE and Theil U Statistic. Basedon the results, ARIMA is better at forecasting the futures trading volume ofcotton than VAR.

DOI

https://doi.org/10.29321/MAJ.10.000550

Pages

1 - 5

Creative Commons

Copyright

© The Author(s), 2025. Published by Madras Agricultural Students' Union in Madras Agricultural Journal (MAJ). This is an Open Access article, distributed under the terms of the Creative Commons Attribution 4.0 License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution and reproduction in any medium, provided the original work is properly cited by the user.

Keywords