Improved inference on risk measures for univariate extremes

DOI10.1214/21-AOAS1555zbMATH Open1498.62096arXiv2007.10780OpenAlexW3045337463MaRDI QIDQ2170408FDOQ2170408

Léo R. Belzile, A. C. Davison

Publication date: 5 September 2022

Published in: The Annals of Applied Statistics (Search for Journal in Brave)

Abstract: We discuss the use of likelihood asymptotics for inference on risk measures in univariate extreme value problems, focusing on estimation of high quantiles and similar summaries of risk for uncertainty quantification. We study whether higher-order approximation based on the tangent exponential model can provide improved inferences, and conclude that inference based on maxima is generally robust to mild model misspecification and that profile likelihood-based confidence intervals will often be adequate, whereas inferences based on threshold exceedances can be badly biased but may be improved by higher-order methods, at least for moderate sample sizes. We use the methods to shed light on catastrophic rainfall in Venezuela, flooding in Venice, and the lifetimes of Italian semi-supercentenarians.

Full work available at URL: https://arxiv.org/abs/2007.10780

Recommendations

• Non-parametric estimation of extreme risk measures from conditional heavy-tailed distributions
• Models and inference for uncertainty in extremal dependence
• Distributionally robust inference for extreme value-at-risk
• Multivariate extremes, aggregation and risk estimation
• On univariate extreme value statistics and the estimation of reinsurance premiums
• ESTIMATION OF RISK MEASURES FROM HEAVY TAILED DISTRIBUTIONS
• Statistical inference for expectile-based risk measures
• Extreme and Inference for Tail Gini Functionals With Applications in Tail Risk Measurement

zbMATH Keywords

profile likelihoodgeneralized Pareto distributionPoisson processextreme value distributiontangent exponential modelhigher-order asymptotic inference

Mathematics Subject Classification ID

Statistics of extreme values; tail inference (62G32) Applications of statistics to environmental and related topics (62P12)

Cites Work

• Title not available (Why is that?)
• Bias reduction of maximum likelihood estimates
• Median bias reduction of maximum likelihood estimates
• Applied Asymptotics
• Title not available (Why is that?)
• Title not available (Why is that?)
• Anticipating Catastrophes through Extreme Value Modelling
• Maximum likelihood estimation in a class of nonregular cases
• Accurate Approximations for Posterior Moments and Marginal Densities
• Title not available (Why is that?)
• Title not available (Why is that?)
• A simple general approach to inference about the tail of a distribution
• Sur la distribution limite du terme maximum d'une série aléatoire
• Likelihood methods in statistics
• Parameter and Quantile Estimation for the Generalized Pareto Distribution
• Estimation of Parameters and Larger Quantiles Based on the k Largest Observations
• A Sturdy Reduced-Bias Extreme Quantile (VaR) Estimator
• Maximum empirical likelihood estimation of the spectral measure of an extreme-value distribu\-tion
• Bayes, reproducibility and the quest for truth
• Is Bayes posterior just quick and dirty confidence?
• On the maximum likelihood estimator for the generalized extreme-value distribution
• An explicit large-deviation approximation to one-parameter tests.
• A simple general formula for tail probabilities for frequentist and Bayesian inference
• Regression Analysis, Nonlinear or Nonnormal: Simple and Accurate p Values From Likelihood Analysis
• The extremogram: a correlogram for extreme events
• The continuous and differentiable domains of attraction of the extreme value distributions
• Tail Index Regression
• Spectral Density Ratio Models for Multivariate Extremes
• Parametric bootstrapping with nuisance parameters
• Improved Likelihood Inference for Discrete Data
• Confidence intervals for extreme Pareto‐type quantiles
• Approximate marginal densities of nonlinear functions
• Bias and size corrections in extreme value modeling
• Approximate predictive likelihood
• Human life is unlimited -- but short
• The plateau of human mortality: Demography of longevity pioneers
• Limits to Human Life Span Through Extreme Value Theory
• Bias-corrected maximum likelihood estimation of the parameters of the generalized Pareto distribution
• Regression‐type models for extremal dependence
• Statistical Models
• Improved inference for the generalized Pareto distribution
• Maximum likelihood estimators based on the block maxima method
• Fast Calibrated Additive Quantile Regression
• A computational study of a quasi-PORT methodology for VaR based on second-order reduced-bias estimation

Cited In (2)

• D. A. S. Fraser: From structural inference to asymptotics
• Distributionally robust inference for extreme value-at-risk

This page was built for publication: Improved inference on risk measures for univariate extremes

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2170408)