ommodity trading decisions, like all decisions, must be made under one of C three broad conditions: certainty, risk, or uncertainty. When a decision is made under certainty, the results of all possible actions are known in advance. When a decision is made under risk, the results are not known in advance. However, the possible results of any given action are known, and the relevant probabilities are known or can be estimated. When a decision is made under uncertainty, the possible results may not be known or may be known incompletely. When the possible results are known, the probabilities are not.
Futures traders obviously need many types of decision-making techniques. Unfortunately, while there is a rich and interesting literature about decision making under certainty and risk, there are few techniques for decision making under uncertainty. Of course, there is a large literature that claims to be concerned with decision making under uncertainty but even the most casual perusal reveals that the techniques are really only valid under conditions of risk or, worse, certainty. Schlaifer's (1969) Analysis of Decisions Under Uncertainty is typical. As far as I can tell, not one of the book's 700 plus pages deals with decision making under uncertainty.
To date, there are only five techniques available for decision making under uncertainty. More correctly, to the best of my knowledge, only five techniques are mentioned in the literature. Worse, the techniques available are deeply flawed and unusable in many, perhaps most, real world situations. This article will present the known techniques and discuss their limitations. It will also present several new techniques that, while flawed in their own way, should be more useful than those now available. All of the published techniques implicitly assume that the uncertainty in the decision is inherent; that is, the uncertainty cannot be reduced to risk or certainty. Fortunately, this is not always the case. However, the relevant information cannot always be purchased or cannot be purchased at a reasonable price and techniques for making decisions under uncertainty are needed.
The published techniques assume further that, given a set of investments, that all of the states of nature that might affect those investments and all of the possible results are known with certainty. This is rarely the case, of course.