Statistics and Its Interface

Volume 7 (2014)

Number 3

Special Issue on Extreme Theory and Application (Part I)

Guest Editors: Yazhen Wang and Zhengjun Zhang

Copula function’s concentration set and its concentrated partition

Pages: 319 – 329

DOI: http://dx.doi.org/10.4310/SII.2014.v7.n3.a2

Authors

Lujun Li (Department of Financial Mathematics, Peking University, Beijing, China)

Yijun Wu (Department of Financial Mathematics, Peking University, Beijing, China)

Jingping Yang (LMEQF, Department of Financial Mathematics, School of Mathematical Sciences and Center for Statistical Sciences, Peking University, Beijing, China)

Abstract

The research on the local correlation structure of copula function is an attractive topic. This paper investigates bivariate copula function’s local correlation structure by defining its concentration set. The concentration set of a copula function is defined in $[0, 1]^2$ with restrained Lebesgue measure such that the samples of the copula fall in the set with the largest probability. The method for finding the concentration set is provided and the properties of the concentration set are discussed. Based on the concentration set, a concentrated partition of $[0, 1]^2$ for the copula function is introduced, and one measure for quantifying copula function’s local correlation is defined by applying our concentrated partition. An empirical study is provided to support our idea of proposing the concentration set.

Keywords

copula function, local correlation structure, concentration set, concentration measure

2010 Mathematics Subject Classification

Primary 60A10, 62H20. Secondary 62P05.

Full Text (PDF format)