Abstract The article focuses on some new approaches to risk. A great deal can be said on the subject of risk and how it might be identified, measured and evaluated. It is not the purpose of this article to distinguish rigorously between different categories or dimensions of risk. Rather the author conceptualizes "risk" as emerging from the fact that some of the information, which is pertinent to a decision, can at best be known only in the form of specified probability distributions. The resulting possibility of deviations from any estimate of the events governed by such probability distributions is then the basic phenomenon, which they shall suppose gives rise to risk. Of course, more than one probability distribution may be applicable and a combination of these distributions may then also require consideration prior to effecting choices between investment alternatives. This kind of phenomenon can supposedly be handled, at least in principle, by suitable theorems or algorithms in probability and statistics. At any rate, given this assumption, one version of a more classical approach would next proceed to reduce each alternative to a single-number basis for comparison. In more sophisticated analyses this might be accomplished via a "utility function" approach.
Byrne et al. (Mon,) studied this question.