A parsimonious multivariate copula for tail dependence modeling

Gildas Mazo 1 Stephane Girard 1 Florence Forbes 1
1 MISTIS - Modelling and Inference of Complex and Structured Stochastic Systems
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : Copulas are increasingly studied both in theory and practice as they are a convenient tool to construct multivariate distribution functions. However the material essentially covers the bi-variate case while in applications the number of variables is much higher. Furthermore, when one wants to take into account tail dependence, a desirable property is to have enough flexibility in the tails while avoiding the exponential growth of the number of parameters. We propose in this communication a one-factor model which exhibits this feature.
Complete list of metadatas

Cited literature [6 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-00863540
Contributor : Stephane Girard <>
Submitted on : Thursday, September 19, 2013 - 10:36:54 AM
Last modification on : Wednesday, April 11, 2018 - 1:58:14 AM
Long-term archiving on : Friday, December 20, 2013 - 3:06:34 PM

File

EVT2013presentername.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-00863540, version 1

Collections

Citation

Gildas Mazo, Stephane Girard, Florence Forbes. A parsimonious multivariate copula for tail dependence modeling. EVT 2013 - Extremes in Vimeiro Today, Sep 2013, Vimeiro, Portugal. ⟨hal-00863540⟩

Share

Metrics

Record views

597

Files downloads

231