Identifying the set of SSD-efficient mixtures of risky alternatives

Target MOTAD and other direct utility-maximization models provide one way of computing SSD-efficient mixtures. These models are appropriate when the utility function is known and can also be used to identify part of the set of SSD-efficient mixtures even when the utility function is not known. However, they do not always identify all SSD-efficient mixtures. A grid method was proposed by Bawa, Lindenberg, and Rafsky. A third approach, which extends the work of Dybvig and Ross, is presented here. It is illustrated by applying it to data from Anderson, Dillon, and Hardaker.