Expected Value A Logic for Decision Making

Abstract

Introduction A vital aspect of our business is exploration for new petroleum reserves. Petroleum exploration involves risk and uncertainty, and over the years management has had to contend with these factors in its analysis of investment strategies. In earlier times the decision maker's analysis of each capital investment decision was probably quite informal and subjective. Most wells were shallow and inexpensive to drill, and drilling prospects were relatively numerous and easy to find. Although never certain of recovering his drilling capital, most of the decisions were probably quite obvious to the decision maker. Undoubtedly, important factors in his analysis were his intuition, judgment, hunches, experience and luck. Decision making could perhaps be described as an "art" in the early days of our country's oil exploration effort. But times have changed. Today the decision maker is faced with capital investment choices that are very complex to evaluate and understand. Spiraling exploratory and development drilling costs, the increased difficulty of finding large new reserves, stringent government controls, and fierce competition have all become very critical factors in the potential success of his investment policies. The decision maker must formally consider risk and uncertainty, as well as complex cash flow factors. Rarely is the decision maker confronted with "obvious" decision choices. These stresses and swains on the management decision-making precess have been the impetus in recent years to find new and better ways to evaluate and compare the desirability of various decision alternatives. The objectives of research in this area of decision theory have been to develop analysis procedures that would capture as much information and judgment as possible in quantitative terms. One of the new concepts available to the decision maker is mathematical expectation, or expected value, as it is usually called. The expected value concept provides a means of combining probability estimates with quantitative estimates of the degree of risk (probability numbers) to yield a risk-adjusted decision. criterion or parameter. This parameter can be used to compare parameter. This parameter can be used to compare the desirability of dissimilar investment alternatives available to the decision maker. The concept is not a substitute for managerial judgment, but rather a method of analysis whereby the various consequences of each decision option can be evaluated and compared. It is the fundamental basis for all quantitative analyses of decisions under uncertainty. These new formal methods of decision analysis have stimulated widespread attention within the business community. A new, interdisciplinary branch of learning variously called decision analysis, statistical decision theory, or management science has developed within the last 10 to 20 years. The general consensus seems to be that these methods have become (or will become) a very important part of the over-all decision-making process. However, as might be expected, process. However, as might be expected, implementation of a new concept is not without its share of frustrations and misunderstanding. And the expected value concept seems to be no exception. Some management reactions to the expected value idea include the following. The decision maker sees no need to formally quantify risk in selecting drilling prospects; hence, expected value criteria are of little orno value. The decision maker accepts the logic of expected value, but feels that he has no meaningful way to assess the probabilities, or risk estimates. Thus the approach has little practical value. The decision maker accepts the meaning of expected value (as the average value recovered per decision over repeated trials), but does not see how the repeated trial connotation can be applied to drilling decisions, where each decision can be made only once. An independent operator, acting on his own behalf, chooses not to use expected value approaches because his organization is too small to justify such a sophisticated analysis. The decision maker accepts the expected value concept, but feels there are more restrictive constraints that overide expected value criteria, such as the firm's asset position, corporate objectives, etc. P. 87