Kernel estimation of extreme risk measures for all domains of attraction

Abstract : Variance are classical risk measures. In statistical terms, the Value-at-risk is the upper α-quantile of the loss distribution where α ∈ (0, 1) is the confidence level. Here, we focus on the properties of these risk measures for extreme losses (where α ↓ 0 is no longer fixed). To assign probabilities to extreme losses we assume that the distribution satisfies a von-Mises condi- tion which allows us to work in the general setting, whether the extreme- value index is positive, negative or zero i.e. for all domains of attraction. We also consider these risk measures in the presence of a covariate. The main goal of this communication is to propose estimators of the above risk measures for all domains of attraction, for extreme losses, and to include a covariate in the estimation. The estimation method thus combines non- parametric kernel methods with extreme-value statistics. The asymptotic distribution of our estimators is established and their finite sample behavior is illustrated on simulated data and on a real data set of daily rainfall.
Type de document :
Communication dans un congrès
Extremes, Copulas and Actuarial Sciences, Feb 2016, Marseille, France. 2016
Liste complète des métadonnées

https://hal.inria.fr/hal-01312846
Contributeur : Jonathan El Methni <>
Soumis le : lundi 9 mai 2016 - 15:43:06
Dernière modification le : mardi 11 octobre 2016 - 13:25:11

Annexe

Identifiants

  • HAL Id : hal-01312846, version 1

Collections

Citation

Jonathan El Methni, Stéphane Girard, Laurent Gardes. Kernel estimation of extreme risk measures for all domains of attraction. Extremes, Copulas and Actuarial Sciences, Feb 2016, Marseille, France. 2016. <hal-01312846>

Partager

Métriques