Efficient Algorithms for Conducting Stochastic Dominance Tests on Large Numbers of Portfolios: A Comment

Abstract

In their recent paper Porter, Wart, and Ferguson (PWF) [6] discuss the factors which they allege to be responsible for “the time-consuming nature of empirical tests of” stochastic dominance (SD) efficiency relative to Markowitz (EV) efficiency. Based upon their analysis, a number of algorithms for improving computational efficiency are offered. The “tricks” upon which applications would be based are of no concern here. Our concern is with the logical inconsistency of any attempt to compare and contrast SD and EV efficiency criteria on empirical grounds.