Premium
Value‐at‐Risk Bounds With Variance Constraints
Author(s) -
Bernard Carole,
Rüschendorf Ludger,
Vanduffel Steven
Publication year - 2017
Publication title -
journal of risk and insurance
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.055
H-Index - 63
eISSN - 1539-6975
pISSN - 0022-4367
DOI - 10.1111/jori.12108
Subject(s) - variance (accounting) , portfolio , aggregate (composite) , constraint (computer aided design) , connection (principal bundle) , stability (learning theory) , value at risk , value (mathematics) , mathematics , point (geometry) , complement (music) , mathematical optimization , econometrics , computer science , economics , risk management , statistics , finance , machine learning , complementation , composite material , gene , phenotype , biochemistry , materials science , geometry , accounting , chemistry
We study bounds on the Value‐at‐Risk (VaR) of a portfolio when besides the marginal distributions of the components its variance is also known, a situation that is of considerable interest in risk management. We discuss when the bounds are sharp (attainable) and also point out a new connection between the study of VaR bounds and the convex ordering of aggregate risk. This connection leads to the construction of an algorithm, called Extended Rearrangement Algorithm (ERA), that makes it possible to approximate sharp VaR bounds. We test the stability and the quality of the algorithm in several numerical examples. We apply the results to the case of credit risk portfolio models and verify that adding the variance constraint gives rise to significantly tighter bounds in all situations of interest.