Forecasting Potato Prices: Application of ARIMA Model

Abstract

Price fluctuations in potatoes (Solanum tuberosum L.) concern consumers, farmers, and policymakers, and its accurate price prediction is important for all the stakeholders. In India, out of a total of 5.34 million ha of land under vegetables, potato occupies nearly 20.8 per cent of area. India produces 12.3 per cent of world potato production (around 45.34 million tons) and is next only to China. The major potato-producing states are highly concentrated in the Indo-gangetic plains of the country. Uttar Pradesh, West Bengal and Bihar account for 32.4, 26.9 and 14.6 per cent of national production of potato. The present study was designed to forecast the prices of potato in these three major potato-producing states of the country. Autoregressive Integrated Moving Average forecasting models - ARIMA (1,0,1) for Varanasi market, ARIMA (2,0,1) for Kolkata market, and the ARIMA (3,0,1) for Patna market were applied. The performance of the ARIMA models produced reliable forecast of prices of potatoes for all three major producing states.