he theory of efficient markets, resting on the concept of perfect competition T and leading to the conclusion that price changes in an efficient market follow a random walk, clearly predicts that if the price discovery process in a market is operating efficiently, then it is not possible to discover any mechanical trading technique which predicts unknown future price changes with accuracy (Mann and Heifner, 1976). The discovery of any such technique then is strong evidence that the price discovery process in a market is not efficient and must be subject to systematic factors which result in the observed predictability of price changes. Martin and Garcia (1981) test hypotheses relevant to the economic purpose of the live cattle futures market. They are concerned with the dichotomy that one view regards live cattle futures as a forecasting market; while another view regards this market as an agency for the formation of rational prices, providing price signals that assist efficient resource allocation and reducing income variation via hedging. Martin and Garcia (1981) conclude that: "Live cattle futures are found to have inadequate forecasting performance for each hypothesis and do not provide better forecasts than lagged cash prices.. . . The analysis does not support the contention that these futures markets are agencies for rational price formation" (p. 209). Thus, Martin and Garcia's work raises doubt about the economic purpose of the live cattle futures.
This article reports on the discovery of a technique which predicted certain drops in live cattle futures prices with 100-percent accuracy over the period from January 1978 through February 1981, discusses the causes of this phenomenon, analyzes its impact on cash cattle prices, discusses the interconnected trading activities of a group of large traders (holding positions in excess of 50 contracts) whose trading activities appear to contribute to the phenomenon, and presents possible solutions to this market inefficiency problem.' 'The results presented in this article first appeared in Staff (1981).