Please use this identifier to cite or link to this item: https://hdl.handle.net/10419/39301 
Authors: 
Year of Publication: 
2010
Series/Report no.: 
SFB 649 Discussion Paper No. 2010,022
Publisher: 
Humboldt University of Berlin, Collaborative Research Center 649 - Economic Risk, Berlin
Abstract: 
This paper make an overview of the copula theory from a practical side. We consider different methods of copula estimation and different Goodness-of-Fit tests for model selection. In the GoF section we apply Kolmogorov-Smirnov and Cramer-von-Mises type tests and calculate power of these tests under different assumptions. Novating in this paper is that all the procedures are done in dimensions higher than two, and in comparison to other papers we consider not only simple Archimedean and Gaussian copulae but also Hierarchical Archimedean Copulae. Afterwards we provide an empirical part to support the theory.
Subjects: 
copula
multivariate distribution
Archimedean copula
GoF
JEL: 
C13
C14
C50
Document Type: 
Working Paper

Files in This Item:
File
Size
838.06 kB





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