Performance Assessment of Sandwich and Block Bootstrap Estimators for Temporally Dependent Bivariate Extremes

Ignoring temporal dependence when modelling sequences of extreme observations yields underestimated standard errors which can lead to inaccurate risk assessment of extreme phenomena such as floods and economic crises. One remedy is to inflate standard errors with sandwich or block bootstrap estimat...

Full description

Saved in:
Bibliographic Details
Main Author: Alexander Engberg
Format: Article
Language:English
Published: Instituto Nacional de Estatística | Statistics Portugal 2025-05-01
Series:Revstat Statistical Journal
Subjects:
Online Access:https://revstat.ine.pt/index.php/REVSTAT/article/view/444
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Ignoring temporal dependence when modelling sequences of extreme observations yields underestimated standard errors which can lead to inaccurate risk assessment of extreme phenomena such as floods and economic crises. One remedy is to inflate standard errors with sandwich or block bootstrap estimators. In this study, four such standard error estimators are investigated and compared, through simulation, when modelling extremes from bivariate sequences with the logistic extreme-value model. Block bootstrap estimators generally yield the most correct standard errors, but suffer from being computationally intensive and a sandwich estimator might be a good alternative due to relatively good performance and computational cheapness.
ISSN:1645-6726
2183-0371