Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/155327 
Is replaced by the following version: 
Title: 

New concepts of symmetry for copulas

The document was removed on behalf of the author(s)/ the editor(s).

Year of Publication: 
2017
Series/Report no.: 
FAU Discussion Papers in Economics No. 06/2017
Publisher: 
Friedrich-Alexander-Universität Erlangen-Nürnberg, Institute for Economics, Nürnberg
Abstract: 
This paper introduces two new concepts of symmetry for multivariate copulas with a focus on tails regions. Properties of the symmetry concepts are investigated for bivariate copulas and a connection to radial symmetry is established. Two nonparametric testing procedures for the new concepts are developed using a vector of locally most powerful rank test statistics, applied to a new generalization of the FGM copula which parameterizes every vertex of the unit cube. This vector quantifies deviations from independence in each vertex and the tests for the new symmetry concepts are based on comparisons of these deviations. A combination of both tests can be used to test for radial symmetry. It is shown that this combined test has a similar power in detecting bivariate radial symmetry compared to recently published nonparametric tests. Further, an application to insurance data is provided. Finally, an improvement of the selection process in the context of vine copula fitting is proposed that is based on the elimination of copula families with unsuitable symmetry properties.
Subjects: 
radial symmetry
vertex symmetry
diametrical symmetry
copula
vine copula
rank-based inference
Document Type: 
Working Paper

Files in This Item:
The document was removed on behalf of the author(s)/ the editor(s) on: May 12, 2017


Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.