Approximation and estimation of very small probabilities of multivariate extreme events

TL;DR

This paper introduces a novel approach using large deviation principles to estimate extremely small probabilities of multivariate extreme events, providing a more accurate asymptotic bound compared to classical tail limit methods.

Contribution

It develops a tail LDP framework for multivariate extremes, linking dependence and marginals, and proposes a simple, consistent estimator that avoids complex rate function estimation.

Findings

The tail LDP provides asymptotic bounds for log-probability ratios.

The proposed estimator is strongly consistent for very small probabilities.

Simulations and real data demonstrate advantages over classical methods.

Abstract

This article discusses modelling of the tail of a multivariate distribution function by means of a large deviation principle (LDP), and its application to the estimation of the probability of a multivariate extreme event from a sample of n iid random vectors, with the probability bounded by powers of sample size with exponents below -1. One way to view classical tail limits is as limits of probability ratios. In contrast, the tail LDP provides asymptotic bounds or limits for log-probability ratios. After standardising the marginals to standard exponential, dependence is represented by a homogeneous rate function. Furthermore, the tail LDP can be extended to represent both dependence and marginals, the latter implying marginal log-GW tail limits. A connection is established between the tail LDP and residual tail dependence (or hidden regular variation) and a recent extension of it. Under a smoothness assumption, they are implied by the tail LDP. Based on the tail LDP, a simple estimator for very small probabilities of extreme events is formulated. It avoids estimation of the rate function by making use of its homogeneity. Strong consistency in the sense of convergence of log-probability ratios is proven. Simulations and an application illustrate the difference between the classical approach and the LDP-based approach.