Partial Stochastic Dominance via Optimal Transport

TL;DR

This paper evaluates various measures of partial stochastic dominance, recommending an optimal transport-based measure as the most natural choice based on multiple perspectives.

Contribution

It provides a comprehensive assessment of partial stochastic dominance measures and proposes an optimal transport-based measure as a new standard.

Findings

Optimal transport measure is recommended as a natural default.

The paper offers axiomatic and computational insights into stochastic dominance measures.

It compares measures from intuitive, axiomatic, computational, and statistical viewpoints.

Abstract

In recent years, a range of measures of partial stochastic dominance have been introduced. These measures attempt to determine the extent to which one distribution is dominated by another. We assess these measures from intuitive, axiomatic, computational and statistical perspectives. Our investigation leads us to recommend a measure related to optimal transport as a natural default.