The ultimate aim of any modern corporate is growth with profit maximization. Growth is the first and foremost characteristic of nature and its products which include modern societies with all their industrial, agricultural and service sectors and above all the research organizations to cater to the needs of primary, secondary and tertiary sectors. Governed by the laws of the universe and nature, societies, markets and above all human life are in the constant churn of development in the realm of creativity and innovativeness.

Trading of gold futures in India has been growing by leaps and bounds since its debut at national commodity exchanges in 2002-03. This paper analyzes gold futures prices at National Commodities and Derivatives Exchange of India (NCDEX) from March 2004 to September 2004. Various AutoRegressive Integrated Moving Average (ARIMA) models are fitted into the price series using a Box-Jenkins framework to find out the best model. The validity of the best-fitted model was also checked. The best-fitted model is then used for short-term forecasting which gives superior forecasting result compared to other models.

Futures prices of any commodity, be it that of agricultural and mineral products, precious metals or financial instruments, help in price discovery. The price discovery process can help in determining the present price in the underlying cash market or the expected futures prices, or both. That is, the price quoted in futures market can be used to predict the spot price prevailing at a future date and can also be used to predict the expected futures price. Predictive power of futures prices is more recognized in commodity market than in financial markets as commodity markets in general are less volatile than stock markets. With the introduction of three national level commodity exchanges in India (discussed in later sections), derivative trading in commodities has increased significantly during the last three to four years. Among various kinds of commodity futures, futures trading in gold has shown notable increase. The size of gold futures market can be judged from the viewpoint expressed by Jignesh Shah, Managing Director, Multi Commodity Exchange Limited as “Indian gold futures market is expected to grow to a staggering size of 40,000 tons of Gold, valued at Rs. 24,00,000 cr in next three to four years” (Press Release, January 2004, MCX Ltd.).

The objective of this paper is to forecast the short-term gold futures prices using ARIMA models. The best ARIMA model is fitted to the given time series of gold futures prices and is used for short-term forecasting. This paper is organized as follows. In section 2, we briefly highlight the historical development of commodity derivative trading in India. Section 3 highlights certain important aspects of Indian gold market.

Section 4 surveys the literature on forecasting models on gold futures prices and other relevant aspects. Section 5 gives a brief overview of the various econometric models used in the present study. Section 6 presents the data description and pre-estimation analysis. Section 7 discusses the selection of an appropriate ARIMA model. Section 8 gives the results and presents the forecasting results. And finally, Section 9 concludes the paper.