http://blogs.reuters.com/felix-salmon/2009/06/09/where-are-the-baseline-stress-test-results/ A slice of lime in the soda Thu, 23 May 2013 23:51:59 +0000 hourly 1 http://wordpress.org/?v=3.4.2 http://blogs.reuters.com/felix-salmon/2009/06/09/where-are-the-baseline-stress-test-results/comment-page-1/#comment-2568 jck Tue, 09 Jun 2009 21:08:45 +0000 http://blogs.reuters.com/felix-salmon/2009/06/09/where-are-the-baseline-stress-test-results/#comment-2568 "If we knew how much difference there was between the baseline test and the more-adverse test...." We DO know the difference "in aggregate" as Treasury released indicative loss rates for both the baseline and adverse scenarios. Ball park figures losses are about 35% lower under the baseline scenario than the adverse one, so you can extrapolate the capital needed "in aggregate" under the baseline scenario. I think it would have been unwise to release the data per bank as that would give infornmation that could be exploited against individual banks. “If we knew how much difference there was between the baseline test and the more-adverse test….”
We DO know the difference “in aggregate” as Treasury released indicative loss rates for both the baseline and adverse scenarios. Ball park figures losses are about 35% lower under the baseline scenario than the adverse one, so you can extrapolate the capital needed “in aggregate” under the baseline scenario. I think it would have been unwise to release the data per bank as that would give infornmation that could be exploited against individual banks.

]]> http://blogs.reuters.com/felix-salmon/2009/06/09/where-are-the-baseline-stress-test-results/comment-page-1/#comment-2562 q Tue, 09 Jun 2009 20:26:33 +0000 http://blogs.reuters.com/felix-salmon/2009/06/09/where-are-the-baseline-stress-test-results/#comment-2562 not quite. there are multiple dimensions to the inputs, so you would need several independent observations to do a linear extrapolation. or if i'm not correct, can you explain what how to do what you suggest? not quite. there are multiple dimensions to the inputs, so you would need several independent observations to do a linear extrapolation. or if i’m not correct, can you explain what how to do what you suggest?

]]>