Price Volatility Dynamics of Cocoa in Major Indian Markets: Evidence from ARCH-GARCH Models
Gali Krishna Chaithanya
Department of Agricultural Economics, Institute of Business and Policy Research, SKUAST-K, Shalimar, Srinagar-190025, J&K, India.
Shabir A. Wani
Department of Agricultural Economics, Institute of Business and Policy Research, SKUAST-K, Shalimar, Srinagar-190025, J&K, India.
Abid Sultan
*
Department of Agricultural Economics, Institute of Business and Policy Research, SKUAST-K, Shalimar, Srinagar-190025, J&K, India.
S. H. Baba
Department of Agricultural Economics, Institute of Business and Policy Research, SKUAST-K, Shalimar, Srinagar-190025, J&K, India.
Imran Khan
Division of Agri. Statistics, SKUAST-K, Shalimar, Srinagar-190025, J&K, India.
Tariq A. Raja
Division of Agri. Statistics, SKUAST-K, Shalimar, Srinagar-190025, J&K, India.
Shelton Peter
Department of Agricultural Economics, SKCAS, Ananthapuram, Andhra Pradesh, India.
*Author to whom correspondence should be addressed.
Abstract
Price volatility can undermine investment in plantation crops and destabilise farm revenues. This study examines the nature and persistence of cocoa-price volatility in three major South Indian markets—West Godavari, Dakshina Kannada, and Idukki—using monthly wholesale price data from January 2008 to December 2023. Each market series comprised 192 observations. Stationarity was assessed before volatility modelling, and the price series were analysed after first differencing. The ARCH-LM test indicated conditional heteroscedasticity in all three series, supporting the use of AR (1)-GARCH (1,1) models. Model selection was based on the Akaike and Schwarz information criteria. The estimated sums of the ARCH and GARCH coefficients were 0.91 for West Godavari, 0.82 for Dakshina Kannada, and 0.72 for Idukki. These values indicate that price shocks decayed most slowly in West Godavari, followed by Dakshina Kannada and Idukki. The findings therefore show market-specific differences in volatility persistence and imply comparatively greater price risk in West Godavari. The analysis is limited to conditional price variance and does not directly estimate the effects of rainfall, production, export demand, competing crop prices, or market institutions. Strengthened market-information systems, price-risk management, crop insurance, and improved coordination among producers, cooperatives, and processors may help market participants respond more effectively to persistent price fluctuations.
Keywords: Cocoa, price volatility, ARCH-LM test, GARCH modelling, volatility persistence, wholesale prices, agricultural markets, market risk, price shocks