rams a test for sub prime, page-22

commodorecontrarian,

One more post on Basel II and I'll give it back to RHG, too.

This is from Gillian Tett's column in tonight's Financial Times, talking about the banks' off-balance sheet conduits, SIV's - call them what you will.

But if this saga is striking, what is truly shocking is that the risks posed by this funding mismatch have gone so unnoticed, for so long. Never mind the fact that even a first year economics student could see that creating ABCP vehicles in this way looked a bit odd. Until recently it appears that few policymakers, bankers or investors had ever factored the prospect of an ABCP strike into their models at all.

Perhaps that was because they were too busy worrying about those hedge funds. But I would guess it was also because the existence of these (largely off-balance sheet) vehicles has been demoted to a footnote, at best, in the published accounts of banks. Or as one policy maker admits: “Nobody had imagined a scenario where the money markets froze up like this. It just wasn’t in the stress testing models.” [emp. added]

There is a perfectly easy, real world way to determine an appropriate level of reserves - look back at past bank crises (Lord knows, there have been a few!). Look at the ones which collapsed and see what their reserves were. Look at the ones which survived, but were long-term impaired, then look at the reserves of the ones that survived the best. Then, make a judgment on which level of risk you are prepared to accept in your financial systems.

I guarantee you that, 9.9 times out of 10, past experience will be a better guide than all the sophisticated statistical and computer modelling, designed and written by maths and computer nerds, with some input from the in-house economics wonk.