Simplified vine copula models: Approximations based on the simplifying assumption

Fabian Spanhel, Malte S. Kurz

Electronic Journal of Statistics · 2019 · 13 citations · 29 references

DOIFull text

Open access

Concepts

TL;DR

Vine copulas are a key tool for modeling high‑dimensional dependence, and the simplifying assumption that a copula can be represented by a simplified vine copula is widely used. The paper investigates the simplifying assumption and examines how well multivariate copulas can be approximated by simplified vine copulas. The authors introduce the partial vine copula, a specific simplified vine copula that assigns a j‑th order partial copula to each edge, and analyze how accurately it captures the dependence structure of the underlying copula. They find that the PVC generally fails to minimize Kullback–Leibler divergence when the simplifying assumption is violated, yet stepwise estimators converge to the PVC regardless of that assumption, making the PVC often the best practical simplified vine copula approximation.

Abstract

Vine copulas, or pair-copula constructions, have become an important tool in high-dimensional dependence modeling. Commonly, it is assumed that the data generating copula can be represented by a simplified vine copula (SVC). In this paper, we study the simplifying assumption and investigate the approximation of multivariate copulas by SVCs. We introduce the partial vine copula (PVC) which is a particular SVC where to any edge a $j$-th order partial copula is assigned. The PVC generalizes the partial correlation matrix and plays a major role in the approximation of copulas by SVCs. We investigate to what extent the PVC describes the dependence structure of the underlying copula. We show that, in general, the PVC does not minimize the Kullback-Leibler divergence from the true copula if the simplifying assumption does not hold. However, under regularity conditions, stepwise estimators of pair-copula constructions converge to the PVC irrespective of whether the simplifying assumption holds or not. Moreover, we elucidate why the PVC is often the best feasible SVC approximation in practice.

References

29