I wrote last week from memory about the use by the Basel II standard of a method, additively combining non-linear and correlated risk events
I see a post on the R-SIG-Fin mailing list from conference organizer Jeff Ryan, that the presentations from R/Finance 2009 are up
Page 4 on the PDF of the Klaus Rheinberger, et al. presentation nicely states the executive summary that this is 'problematic'. The work then shows a worked example
Let's call Basel II what it is -- a top down pronouncement on meaningless rules, written in a fashion that is willfully ignorant of the lessons from the US S & L 'hot money' and actuarially un-sound deposit insurance debacle 20 years ago, of LTCM as to correlated risks and 'being the market', and of the recent Credit Default Swap insurance blowup of AIG
