English

Reflected maxmin copulas and modelling quadrant subindependence

Statistics Theory 2018-12-12 v2 Probability Risk Management Statistics Theory

Abstract

Copula models have become popular in different applications, including modeling shocks, in view of their ability to describe better the dependence concepts in stochastic systems. The class of maxmin copulas was recently introduced by Omladi\v{c} and Ru\v{z}i\'{c}. It extends the well known classes of Marshall-Olkin and Marshall copulas by allowing the external shocks to have different effects on the two components of the system. By a reflection (flip) in one of the variables we introduce a new class of bivariate copulas called reflected maxmin (RMM) copulas. We explore their properties and show that symmetric RMM copulas relate to general RMM copulas similarly as do semilinear copulas relate to Marshall copulas. We transfer that relation also to maxmin copulas. We also characterize possible diagonal functions of symmetric RMM copulas.

Keywords

Cite

@article{arxiv.1808.07646,
  title  = {Reflected maxmin copulas and modelling quadrant subindependence},
  author = {Tomaž Košir and Matjaž Omladič},
  journal= {arXiv preprint arXiv:1808.07646},
  year   = {2018}
}

Comments

30 pages, 2 figures

R2 v1 2026-06-23T03:41:39.239Z