A parametric alternative to the Hill estimator for heavy-tailed distributions

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Despite its wide use, the Hill estimator and its plot remain to be difficult to use in Extreme Value Theory (EVT) due to substantial sampling variations in extreme sample quantiles. In this paper, we propose a new plot we call the eigenvalue plot which can be seen as a generalization of the Hill plot. The theory behind the plot is based on a heavy-tailed parametric distribution class called the scaled Log phase-type (LogPH) distributions, a generalization of the ordinary LogPH distribution class which was previously used to model insurance claims data. We show that its tail property and moment condition are well aligned with EVT. Based on our findings, we construct the eigenvalue plot from fitting a shifted PH distribution to the excess log data with a minimal phase size. Through various numerical examples we illustrate and compare our method against the Hill plot.

Original languageEnglish
Pages (from-to)60-71
Number of pages12
JournalJournal of Banking and Finance
Volume54
DOIs
Publication statusPublished - 2015 May 1

Fingerprint

Phase-type distribution
Eigenvalues
Extreme value theory
Heavy-tailed distribution
Hill estimator
Moment conditions
Sampling
Insurance
Quantile

All Science Journal Classification (ASJC) codes

  • Finance
  • Economics and Econometrics

Cite this

@article{c7795980fa5b45aebdc5d9873f058bd8,
title = "A parametric alternative to the Hill estimator for heavy-tailed distributions",
abstract = "Despite its wide use, the Hill estimator and its plot remain to be difficult to use in Extreme Value Theory (EVT) due to substantial sampling variations in extreme sample quantiles. In this paper, we propose a new plot we call the eigenvalue plot which can be seen as a generalization of the Hill plot. The theory behind the plot is based on a heavy-tailed parametric distribution class called the scaled Log phase-type (LogPH) distributions, a generalization of the ordinary LogPH distribution class which was previously used to model insurance claims data. We show that its tail property and moment condition are well aligned with EVT. Based on our findings, we construct the eigenvalue plot from fitting a shifted PH distribution to the excess log data with a minimal phase size. Through various numerical examples we illustrate and compare our method against the Hill plot.",
author = "Kim, {Joseph H.T.} and Joocheol Kim",
year = "2015",
month = "5",
day = "1",
doi = "10.1016/j.jbankfin.2014.12.020",
language = "English",
volume = "54",
pages = "60--71",
journal = "Journal of Banking and Finance",
issn = "0378-4266",
publisher = "Elsevier",

}

A parametric alternative to the Hill estimator for heavy-tailed distributions. / Kim, Joseph H.T.; Kim, Joocheol.

In: Journal of Banking and Finance, Vol. 54, 01.05.2015, p. 60-71.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A parametric alternative to the Hill estimator for heavy-tailed distributions

AU - Kim, Joseph H.T.

AU - Kim, Joocheol

PY - 2015/5/1

Y1 - 2015/5/1

N2 - Despite its wide use, the Hill estimator and its plot remain to be difficult to use in Extreme Value Theory (EVT) due to substantial sampling variations in extreme sample quantiles. In this paper, we propose a new plot we call the eigenvalue plot which can be seen as a generalization of the Hill plot. The theory behind the plot is based on a heavy-tailed parametric distribution class called the scaled Log phase-type (LogPH) distributions, a generalization of the ordinary LogPH distribution class which was previously used to model insurance claims data. We show that its tail property and moment condition are well aligned with EVT. Based on our findings, we construct the eigenvalue plot from fitting a shifted PH distribution to the excess log data with a minimal phase size. Through various numerical examples we illustrate and compare our method against the Hill plot.

AB - Despite its wide use, the Hill estimator and its plot remain to be difficult to use in Extreme Value Theory (EVT) due to substantial sampling variations in extreme sample quantiles. In this paper, we propose a new plot we call the eigenvalue plot which can be seen as a generalization of the Hill plot. The theory behind the plot is based on a heavy-tailed parametric distribution class called the scaled Log phase-type (LogPH) distributions, a generalization of the ordinary LogPH distribution class which was previously used to model insurance claims data. We show that its tail property and moment condition are well aligned with EVT. Based on our findings, we construct the eigenvalue plot from fitting a shifted PH distribution to the excess log data with a minimal phase size. Through various numerical examples we illustrate and compare our method against the Hill plot.

UR - http://www.scopus.com/inward/record.url?scp=84922464434&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84922464434&partnerID=8YFLogxK

U2 - 10.1016/j.jbankfin.2014.12.020

DO - 10.1016/j.jbankfin.2014.12.020

M3 - Article

VL - 54

SP - 60

EP - 71

JO - Journal of Banking and Finance

JF - Journal of Banking and Finance

SN - 0378-4266

ER -