학술논문

Mean-risk tests of stochastic dominance.
Document Type
Journal
Author
Dentcheva, Darinka (1-STIT) AMS Author Profile; Stock, Gregory J. (1-STIT) AMS Author Profile; Rekeda, Ludmyla (1-FRX) AMS Author Profile
Source
Statistics \& Decisions. International Journal for Statistical Theory and Related Fields (Statist. Decisions) (20110101), 28, no. 2, 97-118. ISSN: 0721-2631 (print).
Subject
62 Statistics -- 62G Nonparametric inference
  62G05 Estimation

90 Operations research, mathematical programming -- 90C Mathematical programming
  90C15 Stochastic programming
Language
English
Abstract
Summary: ``We propose a new approach to testing whether one randomvariable is stochastically non-dominated by another one. The testscompare mean-risk differences of two unknown distributions usingindependent samples. The test can be used for comparison of thecoherent risk measures of the distributions, as well as to rejectstochastic dominance relation of first, second, or higher order betweenthe two distributions. We consider several law-invariant coherentmeasures of risk which are consistent with the stochastic dominancerelation of first and higher order. Numerical comparisons with theMann-Whitney test and with the F-test for comparison of variance areprovided. The numerical study indicates that most of the mean-risktests are more powerful than the Mann-Whitney test.''

Online Access