F evaluate the economic benefits of futures markets is to examine their effects on the intertemporal allocation of resources. This issue was studied by DeCanio (1980) and Stein (1981Stein ( , 1986) ) who showed that the economic benefits of futures trading can be measured by the ex-post welfare loss from misallocation of resources when futures prices are used instead of perfect foresight spot prices. In particular, the ex-post welfare losses increase proportionally to the mean squared error (MSE) of futures prices.
This article studies the relation between the forecasting accuracy of futures Treasury bill rates (relative to forward Treasury bill rates), the development of the futures market, and the changing monetary conditions. Futures contracts are traded in an open-outcry auction market on an organized exchange with a clearing association. The clearing association employs safeguards to guarantee the financial integrity of futures contracts. Forward contracts are created by buying and short-selling spot Treasury bills with different maturities in an over-the-counter dealer market with no clearing association. Kamara (1988) showed that the futures' market trading structure reduces the liquidity and default premium embedded in Treasury bill futures rates relative to forward rates.' The futures market trading structure reduces the cost of entry into the market and allows the participation of small traders whose credit worthiness is inadequate for using forward contracts. Yet, while the increased risk sharing reduces risk premia, it may add traders who are less informed than traders already in the spot and forward markets and can impair the ability to infer from market prices.' Evidence that futures rates tend to be more accurate forecasts of future spot rates than forward rates is reported in Kamara and Lawrence (1983, Hedge andMcDonald (1986), This article began as a joint research project with Colin Lawrence. I have benefited from his comments and comments by