Multivalued Stop-Loss Stochastic Dominance Test

Abstract

Stochastic Dominance tests can be employed to assist decision-makers in ordering uncertain alternatives. Thes tests require specification of alternatives probability distributions and the assumption of the utility function of the decision-maker. With these assumptions, decision alternatives can be partitioned into classes by stochastic dominance or inverse stochastic dominance (stop-loss dominance). This paper notices procedures to identify this class of alternatives in case of multivalued probability distributions.