pith. sign in

arxiv: math/0609163 · v1 · submitted 2006-09-06 · 🧮 math.ST · math.PR· stat.TH

Estimating heavy-tail exponents through max self-similarity

classification 🧮 math.ST math.PRstat.TH
keywords dataestimatorsratealphaboundsconvergencedistributionestimating
0
0 comments X
read the original abstract

In this paper, a novel approach to the problem of estimating the heavy-tail exponent alpha>0 of a distribution is proposed. It is based on the fact that block-maxima of size m of the independent and identically distributed data scale at a rate of m^{1/alpha}. This scaling rate can be captured well by the max-spectrum plot of the data that leads to regression based estimators. Consistency and asymptotic normality of these estimators is established under mild conditions on the behavior of the tail of the distribution. The results are obtained by establishing bounds on the rate of convergence of moment-type functionals of heavy-tailed maxima. Such bounds often yield exact rates of convergence and are of independent interest. Practical issues on the automatic selection of tuning parameters for the estimators and corresponding confidence intervals are also addressed. Extensive numerical simulations show that the proposed method proves competitive for both small and large sample sizes and for a large range of tail exponents. The method is shown to be more robust than the classical Hill plot and is illustrated on two data sets of insurance claims and natural gas field sizes.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.