Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
1146312 | Journal of Multivariate Analysis | 2010 | 15 Pages |
Abstract
In a recent paper Fermanian (2005) [9] studied a goodness-of-fit test for the parametric form of a copula, which is based on an L2L2-distance between a parametric and a nonparametric estimate of the copula density. In the present paper we investigate the asymptotic properties of the proposed test statistic under fixed alternatives. We also study the impact of different estimates for the parameters of the finite-dimensional family of copulas specified by the null hypothesis and illustrate the performance of a parametric bootstrap procedure for the approximation of the critical values.
Keywords
Related Topics
Physical Sciences and Engineering
Mathematics
Numerical Analysis
Authors
Axel Bücher, Holger Dette,