I analyzed how the existence of futures markets affects the amount of a good stored for sale in a subsequent period, when (1) individuals holding the good face price risk because of demand fluctuations, and (2) they make their decisions using the correct (i.e., equilibrium) price distribution. Randomness is alleged to be due to cyclical factors impinging on demand. The two-period model presented does not lend itself to analyzing multiperiod strategies that would normally have to be formulated if disturbances are indeed cyclical or serially correlated in their effects on demand. Hence, the only source of risk in the model is that demand next period may differ from the demand expected. Furthermore, any such forecast error is random and without predictive significance for the demand expected for periods thereafter, given the way in which expectations are assumed to be formed.
Rudimentary features of the model outlined by Britto deserve to be recapitulated to appreciate additional limitations. There is a storable commodity, good 2, produced and held by merchant-farmers; and, there are other goods, numbered 1, which are produced and consumed by consumers and traders. Consumers and traders are endowed with a nonrandom amount of good 1 in period 0 and a random amount in period 1. Randomness of the latter amount will be apparently the only cause of the randomness of demand for good 2 in period 1. Traders alone who, unlike consumers, consume only good 1, will enter into futures trading with merchant-farmers about good 2 if they can achieve thereby a more desirable distribution of consumption, or perhaps one should say wealth, reckoned in terms of good 1. Merchant-farmers alone do the storing.
Having summarized the mis-en-s&ne, we turn to some of the results. What are the implications of the assumptions that merchants and traders act so as to maximize expected utility? Since merchant-farmers are certain of their (individual and I he author IS solely respoiisible tor the views expressed.