Vol. 12, Issue 3 (2023)

Author(s):
K Nirmal Ravi Kumar, M Jagan Mohan Reddy, Adinan Bahahudeen Shafiwu and B Venkataviswateja

Abstract:
This study employed autoregressive integrated moving average (ARIMA) model to forecast daily modal prices of major vegetables viz., bhendi and brinjal in Andhra Pradesh, India. ARIMA with explanatory variable (ARIMAX) model was also considered in this study through considering ‘market arrivals’ of selected commodities as explanatory variable, as it further improves the forecasting performance. This will also facilitate to compare the forecasting performances from ARIMA and ARIMAX models. Findings from this study revealed (1,1,2) and (2,1,1) are the best fit models of ARIMA and (1,1,2, a-bh) and (2,1,1,a-br) are the best fit models of ARIMAX for bhendi and brinjal respectively. Further, ARIMAX model outperformed ARIMA model both in terms of model fit (as indicated by lower error metrics) and forecasting performance. Diebold-Mariano test and Harvey, Leybourne and Newbold test statistics also highlighted the forecast performance from ARIMAX model is statistically superior than of ARIMA model for both bhendi and brinjal. As the modal prices showed prominent declining trend based on ARIMAX model, the same should be considered by the officials of Rythu Bazars to safeguard the interests of farmers and other stakeholders.