This paper develops tests of unconditional mean-variance efficiency under weak distributional assumptions using a generalized method of moments framework. These tests are potentially more robust than commonly employed tests which rely on the assumption that asset returns are normally distributed and temporarily i.i.d. Using returns for size-based portfolios from 1926 to 1988, the authors show that the conclusion concerning the mean-variance efficiency of market indexes can be sensitive to the test considered.

This paper develops tests of unconditional mean‐variance efflciency under weak distributional assumptions using a Generalized Method of Moments framework. These tests are potentially more robust than commonly employed tests which rely on the assumption that asset returns are normally distributed and temporarily i.i.d. Using returns for size‐based portfolios from 1926 to 1988 we show that the conclusion concerning the mean‐variance effilciency of market indexes can be sensitive to the test considered.