Tail risk estimation based on extreme Lp-quantiles

Abstract : The notion of quantiles lies at the heart of extreme-value theory and is one of the basic tools in risk management.The alternative family of expectiles is based on squared rather than absolute error loss minimization. Both quantiles and expectiles can be embedded in the more general class of Lp-quantiles as the minimizers of an asymmetric power loss function. In this talk, new methods will be presented for estimating high Lp-quantiles from heavy-tailed distributions.
Type de document :
Communication dans un congrès
Statistics workshop Tilburg University, Dec 2016, Tilburg, Netherlands. 2016
Liste complète des métadonnées

https://hal.archives-ouvertes.fr/hal-01415533
Contributeur : Stephane Girard <>
Soumis le : mardi 13 décembre 2016 - 11:37:26
Dernière modification le : mercredi 11 avril 2018 - 01:58:24

Identifiants

  • HAL Id : hal-01415533, version 1

Citation

Stéphane Girard, Abdelaati Daouia, Gilles Stupfler. Tail risk estimation based on extreme Lp-quantiles. Statistics workshop Tilburg University, Dec 2016, Tilburg, Netherlands. 2016. 〈hal-01415533〉

Partager

Métriques

Consultations de la notice

1088