The moment index of minima. II. (Q2493804)

[For part I, J. Appl. Probab. 38A, Spec. Vol., 33--36 (2001), see the review Zbl 1022.60014.] The moment index \(k(X)\) of a non-negative random variable \(X\) is defined as \[ k(X) = \sup\{ k\geq 0;\;E(X^k) <\infty\}. \] For the independent, non-negative random variables \(X\) and \(Y\) having finite moment indexes, the moment index has the following property: \[ k(\min(X, Y)) \geq k(X) + k(Y).\tag{MII} \] The aim of this article is to characterize the conditions under which equality holds for the inequality (MII): for the non-negative random variables \(X\) and \(Y\), the tail of the density function of either \(X\) or \(Y\) has to be regularly varying. Extensions to related random variables and their distributions are discussed, and an example in which \(X\) and \(Y\) have discrete supports that are ``increasingly sparse'' and ``well interspersed'' is proved to involve the strict inequality in the (MII) relation.