Determining returns to storage: does data aggregation matter?

摘要

This article examines aggregate and grain elevator data from to determine if aggregate data underestimate returns to storage in Oklahoma, USA. The micro-level data comes from three grain elevators located in western Oklahoma spanning nine crop years from the spring of 1992 through the spring of 2001. Market-level data are from the Oklahoma Department of Agriculture data. There is no difference between the mean returns estimated with aggregate data and the mean returns estimated with transaction level data from grain elevators in Oklahoma. Storage at a loss does occur in Oklahoma but it cannot be explained by data aggregation. In conclusion, storage at a loss is not due to data aggregation. It is suggested that the problem could be in the choice of a 12-month benchmark for Oklahoma. Oklahoma is closer to the Gulf and therefore producers have an incentive to sell early in the marketing season. But a 12-month benchmark might work reasonably well for corn produced in Illinois since it is near the centre of U.S. corn production.