Riskiness for sets of gambles (Q403706): Difference between revisions

The paper is devoted to the problem of ranking sets of gambles that arise in models of decision under uncertainty. The predecessors of this investigation are papers by \textit{R. J. Aumann} and \textit{R. Serrano} [J. Polit. Econ. 116, No. 5, 810--836 (2008; Zbl 1341.91040)], and \textit{D. P. Foster} and \textit{S. Hart} [``An operational measure of riskiness'', ibid. 117, No. 5, 785--814 (2009; \url{doi:10.1086/644840})], where two economic riskiness measures were suggested in a setup with known probabilities and a large class of expected utility functions. A gamble is understood as a real-valued random variable \(g\) with finitely many values having some negative values with positive probability, i.e. \(P[g<0]>0\) (losses are possible), and positive expectation, i.e. \(E[g]>0\). The author suggests an approach to the ranking problem by merging the Aumann-Serrano and the Foster-Hart approaches with respect to the presence or absence of a specific utility function. In addition, he uses the uniform-rejection criterion of \textit{S. Hart} [``Comparing risks by acceptance and rejection'', ibid. 119, No. 4, 617--638 (2011; \url{doi:10.1086/662222})]. The suggested approach is deeply investigated with respect to the introduced properties of ``wealth-uniformly dominating'' and ``utility-uniformly dominating'' for a set of gambles.