Affiliations: Department of Business Administration, Mangalore University, Karnataka, India. E-mails: {massoud982000, tmmallik}@yahoo.com
Note: [] Corresponding author: T. Mallikarjunappa, Professor, Department of Business Administration, Mangalore University, Mangalagangothri – 574 199 Mangalore, DK, Karnataka, India. Tel.: +91 824 2287637, +91 824 2287158; E-mail: [email protected]
Abstract: This paper examines the possible effects of commodity futures trading on commodity spot price volatility in Indian market using GARCH family models. We examine daily trading and hedging activities effects by incorporating futures trade volume and open interest, respectively, to represent the two variables. These two variables are decomposed into expected and unexpected components. We also investigate the dynamic relationship between commodities spot volatility and futures trading activity. Following Koch [Journal of Banking and Finance 17(6) (1993), 1191–1205], we posit that VAR and ordinary least square model may lead to spurious results as there could be contemporaneous interdependence in structural relationships between variables. Therefore, we use simultaneous equation model (SEM) and Three Stage Least Square (3SLS). The results of GJR-GARCH model indicate that most of the coefficients of unexpected futures trading activities are statistically insignificant except for a few cases. Also, expected futures volume has positive and expected open interest has negative impact on the spot price volatility. The results of contemporaneous coefficients of the SEM model show that unexpected futures trading volume and open interest have significant positive and negative impact, respectively, on the spot market volatility. The bilateral causal relationships between spot volatility and unexpected trading activity components are observed in most of the commodities.
Keywords: Commodity spot market, commodity futures trading, volatility, trade volume, open interest, GARCH models, VAR model, SEM model
