Distributionally robust tail bounds based on Wasserstein distance and $f$-divergence

TL;DR

This paper develops robust bounds on tail probabilities and tail indices for heavy-tailed distributions under model misspecification, using Wasserstein and $f$-divergence measures, with applications to insurance data.

Contribution

It introduces explicit asymptotic bounds for tail behavior based on Wasserstein and $f$-divergence ambiguity sets, highlighting the impact of discrepancy choice.

Findings

Explicit bounds derived for Wasserstein and $f$-divergence neighborhoods.

Comparison of bounds demonstrates the influence of discrepancy measure.

Application to insurance claims illustrates practical relevance.

Abstract

In this work, we provide robust bounds on the tail probabilities and the tail index of heavy-tailed distributions in the context of model misspecification. They are defined as the optimal value when computing the worst-case tail behavior over all models within some neighborhood of the reference model. The choice of the discrepancy between the models used to build this neighborhood plays a crucial role in assessing the size of the asymptotic bounds. We evaluate the robust tail behavior in ambiguity sets based on the Wasserstein distance and Csisz\'ar $f$-divergence and obtain explicit expressions for the corresponding asymptotic bounds. In an application to Danish fire insurance claims we compare the difference between these bounds and show the importance of the choice of discrepancy measure.