Too much of a good thing?

This paper explores a seemingly paradoxical phenomenon associated with the use of expected-utility theory in capital-budgeting and risk-sharing decisions under uncertainty. As an investment prospect becomes better and better, decision makers using classic decision-analysis techniques may, in fact, prefer less and less of it. We explore this phenomenon in the very real context of petroleum company drilling-investment decisions and demonstrate that the phenomenon is pervasive. When we examined the prospect inventory of a major oil company, we found that an increase in the upside payoff would lead to a lower optimal working interest for the grand majority of its prospects. We also explore the underlying factors leading to this phenomenon to develop both intuition and understanding. These factors have to do with degree of risk aversion (the more risk averse you are, the less likely you are to increase your holdings of an improving prospect) and managerial perspectives on risk. Once understood, the result no longer seems surprising.