“Don't put all your eggs in one basket,” is a familiar adage. Economists, such as Marschak, Markowitz, and Tobin, who work only with mean income and its variance, can give specific content to this rule—namely, putting a fixed total of wealth equally into independently, identically distributed investments will leave the mean gain unchanged and will minimize the variance.

Questions have been raised in recent years concerning the interpretation of previous studies purporting to show that distributed earnings have had a consistently greater impact on equity prices than have retained earnings. Miller and Modigliani have convincingly argued that if capital markets are perfect and rational behavior of market participants is assumed, the price-earnings ratio of the shares of a firm with a given investment policy should be invariant to alternative earnings-payout ratios. They also point out, however, that with the present tax subsidy on capital gains and the existence of substantial brokerage fees and flotation costs, dividend policy might be expected to have an effect on share prices, even though the amount and direction of this effect is an empirical matter and not determinable a priori.

The origins of the application of quantitative methods to the determination of optimum solutions to managerial problems can be traced back at least as far as 50 years when calculus techniques were first suggested for solution of economic lot size problems. However, the great increase in applications of optimizing techniques occurred after the Second World War with the popularization of the operations research concept and development of a wide variety of mathematical tools — among them linear programming, queuing theory, dynamic programming, and network flow theory. The development of digital computers in the late 1940's substantially reduced computational problems associated with the use of many mathematical methods and also made possible the use of simulation techniques which, without digital computers, are not ordinarily practical because of the large amount of computation involved.

In a recent article, Donald R. Hodgman set forth a framework for analyzing commercial bank lending behavior which emphasized the relationship of customers to their banks as both depositors and borrowers. Specifically, Hodgman links the borrower's contract rate of interest to the profitability of his deposit account, i.e., banks compete for profitable deposit customers by offering the customer a rate which is lower than the comparable open market rate (adjusted for risk). Hodgman uses the terms deposit relationship and customer relationship to describe this behavior. He argues that compensating balance requirements, the prime rate convention, and provision of services below cost may be viewed as a systematic, rational attempt by commercial banks to maximize long-run profit under existing institutional arrangements when viewed from the perspective of the customer relationship.

Econometric models have often been used to explain and predict demand, production and price patterns of agricultural commodities. With computers generally available, estimation and application of these models is a simple matter. In spite of this, there has been nothing written, at least nothing that we are aware of, which details the use of such models as an aid to speculation in commodities futures. This brief note reports successful use of an econometric model and a time-sharing computer system for this purpose.