The present article is an attempt to empirically investigate the long-term market efficiency and price discovery in Indian commodity futures market. The study has been conducted with eight commodities which include two agricultural commodities, two industrial commodities, two precious metal and two energy commodities. Sophisticated statistical methods like restricted cointegration and vector error correction model (VECM) are used to analyse the spot and futures prices time series. Restricted cointegration test shows that near-month futures prices for all the commodities are cointegrated with the spot prices but futures prices of all the commodities are inefficient to predict the future spot price. Indian commodity futures market evidenced as the thinly traded market (Kumar & Pandey, 2013, Journal of Indian Business Research, 5(2), 101–121) rejects the null hypothesis of efficiency and unbiasedness for all the eight commodities which reconfirms the result of Fortenbery and Zapata (1997, Journal of Futures Markets, 17(3), 279–301). The presence of short-term biases in the Indian futures market is evidenced in the results of VECM model which indicates the presence of informational efficiency. The statistically significant value of past prices of spot and futures confirm the short-term inefficiency and biasedness. The significant value of error correction term (ECT) of futures prices suggests that commodity futures are the most important indicator of commodity price movements. The important implication of the results is for market traders. They can use the futures prices to discover the new equilibrium and earn profits by transmitting it to the spot market. The better understanding of the interconnectedness of these market would be useful for policymakers who try to establish stability in the financial markets.
Agricultural commodity futures prices prediction via long- and short-term time series network
