Tails of higher-order moments with dominatedly varying summands

被引：6

作者：

Leipus, Remigijus ^{[1
]}

Siaulys, Jonas ^{[2
]}

Vareikaite, Ieva ^{[2
]}

机构：

[1] Vilnius Univ, Inst Appl Math, Naugarduko Str 24, LT-03225 Vilnius, Lithuania

[2] Vilnius Univ, Inst Math, Naugarduko Str 24, LT-03225 Vilnius, Lithuania

来源：

LITHUANIAN MATHEMATICAL JOURNAL | 2019年 / 59卷 / 03期

关键词：

asymptotic independence; tail-moment; heavy tails; dominatedly varying distribution function; sum of random variables; weighted sum; RANDOMLY WEIGHTED SUMS; DEPENDENT RANDOM-VARIABLES; INDEPENDENT RANDOM-VARIABLES; GOOVAERTS RISK MEASURE; TIME RUIN PROBABILITY; ASYMPTOTICS; MODEL; FINITE; INSURANCE; EXPECTATION;

D O I：

10.1007/s10986-019-09444-x

中图分类号：

O1 [数学];

学科分类号：

0701 ; 070101 ;

摘要：

Let xi(1), . . . , xi(n) be dependent real -valued random variables with dominatedly varying distribution functions. We investigate the asymptotic behavior of the tail-moment E((S-n(xi))(m)1({Sn xi >x})), where m is a nonnegative integer, and S-n(xi) = xi(1) + . . . + xi(n). We consider the dependence structure, similar to pairwise strongly quasiasymptotic independence among the random summands. We also study the case where each summand of S-n(xi) can be expressed as the product of two random variables.

引用

页码：389 / 407

页数：19

共 43 条

[1]

[Anonymous], 1987, ENCY MATH APPL

[2] Asymptotics for risk capital allocations based on Conditional Tail Expectation [J].