Title: Stochastic dominance inefficiency tests
Authors: Sofia Anyfantaki - Athens University of Economics and Business and Bank of Greece (Greece) [presenting]
Nikolas Topaloglou - Athens University of Economics and Business Research Center (Greece)
Esfandiar Maasoumi - Emory University (United States)
Jue Ren - Texas Christian University (United States)
Abstract: The most common approach to test for stochastic dominance is to posit the null hypothesis of dominance. Rejection of the null, however, does not imply dominance since it can also happen that the test fails to rank the two distributions. A statistical test is proposed for the stochastic dominance efficiency of a given portfolio when the null hypothesis is inefficiency. An analytical characterization of stochastic dominance inefficiency and the null limit distribution for the associated empirical test statistic are derived. Feasible approaches to statistical inference based on bootstrapping and numerical optimization are developed. The test is used to empirically establish whether well known mutual funds are efficient during different market conditions with respect to (1) all possible portfolios constructed from a set of 248 funds and (2) different market proxies.