Uncertainty and Risk from Decision-Tree Analysis

Abstract

Decision tree analysis is one of the standard devices for assessing the worth of an exploration opportunity in the hydrocarbon industry. Often just the expected value of the decision-tree is used to decide on the project: a positive expected value usually being a rationale for proceeding, a negative expected value being a rationale for not going ahead. Here we show that uncertainty on the expected value, also easily determinable from decision-tree analysis, is a major factor in evaluating the risk of going ahead with an exploration project. Reliance solely on the expected value can lead to decisions not to participate in exploration projects which have a relatively high risk but also a relatively good chance of being profitable – and this is particularly the case when the expected value is negative but much smaller in magnitude than the standard error in the expected value.