AI & Computingarticle2026-10-15

Rao-Blackwellized e-variables

Open access0 citations

Abstract

Summary We show that for any concave utility, the expected utility of an e-variable can only increase after conditioning on a sufficient statistic, echoing the classic Rao-Blackwell theorem. The simplest form of the result has a straightforward proof using Jensen’s inequality, but we prove more sophisticated versions. We provide several applications of this insight, including a derivation of the log-optimal e-variable for testing the variance in a linear regression model.

// Source

View paper (DOI)Open access versionOpenAlexBiometrikaPublished 2026-10-15

Institutions: Stanford University, Centrum Wiskunde & Informatica, Carnegie Mellon University