Forecasting of monthly cardamom price using long memory time series modelling technique

Abstract

Cardamom is an important spice crop known for its aromatic qualities. Kerala is a major producer state of small cardamom in India. This study aimed to develop a long memory time series model for forecasting of monthly average farm wholesale price of small cardamom in Kerala. Monthly price data collected from Directorate of Economics and Statistics, Government of Kerala for the period January 2007 to December 2020. Upon analysis of data, ARFIMA (0, 0.22, 1) model was selected which in turn compared with traditional ARIMA model. The results emphasised the superior performance of ARFIMA (0, 0.22, 1) model over ARIMA model in forecasting the long memory price time series of small cardamom.