his article examines the pricing efficiency of the gold futures market relative T to the Treasury bill futures market. Futures contracts for gold, a relative newcomer, have only been traded since 1975. T-bill futures contracts, also a relative newcomer, have been traded on organized exchanges since 1976.To the extent that prices for different delivery dates on the gold contract should reflect some interest rate which is related to the T-bill rate (as a measure of the opportunity cost of storing gold), gold futures prices should exhibit a well-defined relationship with T-bill futures prices in equilibrium. This article demonstrates that the futures markets for gold and T-bills are less than completely efficient relative to each other insofar as the interest rates implied by gold futures prices do tend to stray from their equilibrium level relative to the forward rates implied by the corresponding T-bill futures prices. If the market is inefficient from time to time and the implied gold rate is sometimes too high or too low relative to the implied T-bill rate, then one should be able to make abnormal profits by using trading rules based on the implied rates; evidence of such an ability is presented.Section I provides a brief review of the literature concerning efficiency in futures markets. Section I1 discusses the nature of futures markets and the mechanics of the "arbitrage" which produces potential profits from observed disequilibria in the gold and T-bill markets. A trading procedure is proposed which is novel because it utilizes a "tail" and a hedge ratio for the gold spreads so as to isolate profit arising solely from changes in relative implied rates. The results of simulated trading over *University of Illinois at Chicago and University of Hartford respectively. The authors are grateful to Fred Arditti for inspiring the research presented in this article, and for his comments on a previous draft, and to the Chicago Mercantile Exchange for providing the data employed in the empirical analysis. We wish to thank Michael Henry for computational assistance. We are, of course, solely responsible for any errors.
everal studies such as Fama (1965) and Oldfield and Rogalski (1980) document that S equity returns are more volatile during trading hours than during non-trading hours.In a recent paper, French and Roll (1986) examine the behavior of the daily (close to close) returns of all NYSE and AMEX stocks. They find that trading hour return variance is much higher than non-trading hour variance and conclude that the extra trading hour variance in equity returns can be attributed predominantly to private information trading. Some traces of noise trading are also documented.The purpose of this study is to test whether the extra trading hour volatility which has been documented in stock markets also exists in the gold futures market and to document variations in the distribution of intraday futures returns over the course of a trading day.' A discussion of how these might be related to information and noise trading is included.
I. DATA AND METHODOLOGY
A. The DataThe empirical work employs gold futures transactions data obtained from the Chicago Mercantile Exchange. Gold is chosen because it represents a universally-recognized and widely traded commodity.The study focuses on the behavior of the December 1978 and the March 1979 gold contracts. For each of these contracts, price quotes from the four months prior to (and not including) the delivery month are investigated. The resulting December 1978 contract price file contains 59,420 observations and the March 1979 contract price file contains 53,786 observations. The comments of Mike Long and an anonymous referee are gratefully acknowledged. All remaining errors are our own.'The decision to study intraday returns reinforces the decision to focus on futures markets. Futures exchanges employ a uniform market structure throughout the trading day while the NYSE does not (see Amihud and Mendelson (1987)). This uniformity of market structure is important for a study which seeks to document variations in the return distribution throughout the day.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.