Tweaking Archimedes to find tail risk

By Natalia Puzanova

Thursday, August 30, 2012

Natalia Puzanova describes a new way of constructing hierarchical Archimedean copulas which ensures that sampling is straightforward and that tail risk is not underestimated

Subscriber-only article

This article is available only to Creditflux subscribers and free trial users within 30 days of publication.

Already a subscriber? Not logged in? Click here to login.

If you have not already done so,
you may request a FREE TRIAL by clicking here

This trial will give you:
  • 4-weeks' free online access to our
    most recent subscriber-only articles
  • Daily breaking news alert sent by email
  • A print copy of Creditflux

If you currently have a free trial, you will see this message when you try to view articles older than 30 days.

Comment by: Anonymous. Posted 11 years ago [2012-09-20 23:36:17]

Research of this nature is important. The author's mathematical dexterity is impressive and I chose to take her word for many of the equations while reading this paper. As advice, though, I note that the last paragraph simply says that the Archimedean copula gives a different result than the Gaussian copula. There is no assertion that the former is more accurate than the latter. The unspoken but implicit argument appears to be that the more complex the analysis, the more reliable the result. As an industry, please let's not go down this road again!

Newsletter

more stories...