Sufficient Conditions for Expected Utility to Imply Mean-Standard Deviation Rankings: Empirical Evidence Concerning the Location and Scale Condition.

This paper presents and implements a procedure which examines for empirical support for the location and scale condition. The Kolmogorov-Smirnov multisample test is used to determine if the distribution functions describing the nonsystematic risk component of rate of return for portfolios of common stock are equal to one another except for location and scale. Implications concerning the relationship between mean-variance and expected utility efficient sets are noted. Copyright 1992 by Royal Economic Society.