• DocumentCode
    1698050
  • Title

    Stressed Value-at-Risk

  • Author

    Dash, Jan

  • Author_Institution
    Strategic Risk Res., Bloomberg L.P., New York, NY, USA
  • fYear
    2012
  • Firstpage
    1
  • Lastpage
    1
  • Abstract
    Stressed Value at Risk (Stressed VAR) in its advanced framework provides a realistic measure of market risk tailored for stressed market environments. The simpler regulatory version of Stressed VAR is a special case. Stressed VAR corrects various deficits of ordinary VAR in times of market stress. Stressed VAR incorporates scenario analysis in a VAR setting in a sophisticated and consistent fashion. The mathematical framework is familiar and simple, designed to be understandable in a practical way by risk managers, traders, and regulators. Specifically, the familiar Gaussian (normal) probability formalism is employed, but in a completely different way than for ordinary VAR, designed to account for tail risk and collective behavior. The two main ingredients for Stressed VAR are "fat-tail volatilities" that account for outlier events in the risk factors, and stressed correlations between risk factors that account for collective market participant behavior in stressed markets. This information is provided as input to standard Monte Carlo simulation to determine stressed market risks for a given portfolio. The VAR with the inputs of the fat-tail volatilities and the stressed correlations is the Stressed VAR. Bloomberg LP is implementing Stressed VAR in the PORT portfolio system.
  • Keywords
    Gaussian processes; Monte Carlo methods; marketing; risk management; Bloomberg LP; Gaussian probability formalism; PORT portfolio system; fat-tail volatilities; market risk; mathematical framework; participant behavior; risk factors; scenario analysis; standard Monte Carlo simulation; stressed VAR regulatory version; stressed correlations; stressed market environments; stressed value-at-risk; tail risk;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computational Intelligence for Financial Engineering & Economics (CIFEr), 2012 IEEE Conference on
  • Conference_Location
    New York, NY
  • ISSN
    PENDING
  • Print_ISBN
    978-1-4673-1802-0
  • Electronic_ISBN
    PENDING
  • Type

    conf

  • DOI
    10.1109/CIFEr.2012.6327832
  • Filename
    6327832