Shape measures based on the convex transform order

Abstract

We introduce in this paper three functional measures to examine the shape of univariate distributions that satisfy some desirable properties, including consistency with the convex transform order. The first two are weighted tail indices that characterize location-scale families of distributions and the third one is a skewness measure. We study some classical families of distributions in terms of these measures, characterizing, in particular, the generalized Pareto distribution. The problem of estimation is also considered and illustrated with real data.