Estimation of high-dimensional extreme conditional expectiles

Stéphane Girard 1 Gilles Stupfler 2
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 : Expectiles are least-square analogues of quantiles. They have received a fair amount of attention due to their potential for application in financial, actuarial, and economic contexts. Some recent work has focused on the application of extreme expectiles to assess tail risk, and on their estimation in a heavy-tailed framework. We investigate the estimation of extreme expectiles of a heavy-tailed random variable $Y$ given a high-dimensional covariate $X$. We derive generic conditions under which the limiting behaviour of our estimators can be established. Applications are presented to some regression models. A finite-sample study illustrates the behaviour of our procedures in practice.
Document type :
Conference papers
Complete list of metadatas

https://hal.inria.fr/hal-02099370
Contributor : Stephane Girard <>
Submitted on : Sunday, April 14, 2019 - 8:57:21 PM
Last modification on : Tuesday, April 16, 2019 - 3:27:50 PM

Identifiers

  • HAL Id : hal-02099370, version 1

Collections

Citation

Stéphane Girard, Gilles Stupfler. Estimation of high-dimensional extreme conditional expectiles. CRoNoS & MDA 2019 - Final CRoNoS meeting and 2nd workshop on Multivariate Data Analysis, Apr 2019, Limassol, Cyprus. ⟨hal-02099370⟩

Share

Metrics

Record views

56