English

Dimension Reduction in Multivariate Extremes via Latent Linear Factor Models

Methodology 2026-02-27 v1

Abstract

We propose a new and interpretable class of high-dimensional tail dependence models based on latent linear factor structures. Specifically, extremal dependence of an observable vector is assumed to be driven by a lower-dimensional latent KK-factor model, where KdK \ll d, thereby inducing an explicit form of dimension reduction. Geometrically, this is reflected in the support of the associated spectral dependence measure, whose intrinsic dimension is at most K1K-1. The loading structure may additionally exhibit sparsity, meaning that each component is influenced by only a small number of latent factors, which further enhances interpretability and scalability. Under mild structural assumptions, we establish identifiability of the model parameters and provide a constructive recovery procedure based on a margin-free tail pairwise dependence matrix, which also yields practical rank-based estimation methods. The framework combines naturally with marginal tail models and is particularly well suited to high-dimensional settings. We illustrate its applicability in a spatial wind energy application, where the latent factor structure enables tractable estimation of the risk that a large proportion of turbines simultaneously fall below their cut-in wind speed thresholds.

Keywords

Cite

@article{arxiv.2602.23143,
  title  = {Dimension Reduction in Multivariate Extremes via Latent Linear Factor Models},
  author = {Alexis Boulin and Axel Bücher},
  journal= {arXiv preprint arXiv:2602.23143},
  year   = {2026}
}

Comments

36 pages

R2 v1 2026-07-01T10:54:06.726Z