Empirical likelihood based confidence intervals for copulas

JOURNAL OF MULTIVARIATE ANALYSIS

卷:100

Article

Chen, Jian²  Peng, Liang²  Zhao, Yichuan¹ 

[1] Georgia State Univ, Dept Math & Stat, Atlanta, GA 30303 USA

[2] Georgia Inst Technol, Sch Math, Atlanta, GA 30332 USA

关键词: Confidence interval;    Copula;    Empirical likelihood;    Normal approximation;   

DOI  :  10.1016/j.jmva.2008.04.005

来源: Elsevier

【 摘 要 】

Copula as an effective way of modeling dependence has become more or less a standard tool in risk management, and a wide range of applications of copula models appear in the literature of economics, econometrics, insurance, finance, etc. How to estimate and test a copula plays an important role in practice, and both parametric and nonparametric methods have been studied in the literature. In this paper, we focus on interval estimation and propose an empirical likelihood based confidence interval for a copula. A simulation study and a real data analysis are conducted to compare the finite sample behavior of the proposed empirical likelihood method with the bootstrap method based on either the empirical copula estimator or the kernel smoothing copula estimator. (C) 2008 Elsevier Inc. All rights reserved.

【 授权许可】

Free