336. Stochastic dominance intensity
Invited abstract in session MB-33: Scenarios and Stochastic Dominance, stream Decision Analysis.
Monday, 10:30-12:00Room: Maurice Keyworth 1.31
Authors (first author is the speaker)
| 1. | Gaston Yalonetzky
|
Abstract
Stochastic dominance conditions have long been used in myriad areas of finance, operations research and welfare economics to test the robustness of distributional comparisons to alternative evaluation functions satisfying a common set of axioms. Besides being commonly unable to rank all distributions, dominance conditions are deemed uninformative regarding the intensity of a preference relation between distributions. However, for a broad array of dominance conditions for ordinal variables known as cone orderings, we demonstrate that from the set of distributions ordered by a given dominance criterion we can derive a set of ordered pairs ranked by the intensity of the same dominance condition. For that purpose, we derive results describing the conditions whose fulfillment ensures that the preference difference favouring distribution p over q is greater than that favouring r over s for all evaluation functions consistent with the same dominance criterion. The results apply to well-known partial ordering criteria for ordinal variables including first-order dominance, Hammond welfare, Hammond inequality, median-preserving spreads and bipolarisation, among others.
Keywords
- Decision Analysis
- Decision Theory
Status: accepted
Back to the list of papers