Very off-topic, but the sub-prime crisis is the biggest accounting story around, so I'm not spending much time thinking about tax right now.

Lehman Brothers' numbers yesterday, and Morgan Stanley's today were full of bad news about sub-prime, but not terrible news. So are we out of the woods?

I don't think so, and at least partly for accounting reasons. We have more to say about this in tomorrow's edition, but suffice to say that the models being used to describe the position of US banks are said to be as complex as the derivatives they purport to describe. Until we know precisely how badly the banks are doing, and they know how they are doing, will they be confident enough to lend to each other?

Lehman Brothers announced a $700m write-down on its sub-prime derivatives yesterday. But my sceptical side enjoyed this comment, left by 'London Banker' on FT's excellent Alphaville yesterday,

The real test will be the breakdown of Level 1 (mark to market), Level 2 (mark to model) and Level 3 (mark to make-believe) earnings. Until I see these numbers, I don’t believe anything that’s written about good news. I’ve heard the banks’ CFOs have been meeting all through the past week to agree the methods for flattering their accounts this quarter and a cease fire on sniping at each other to instill confidence. If there’s a big leap in Level 2 and Level 3 but no outcry from the analysts, that will indicate a deal’s been done.

Is that a bit too cynical? Or just common sense?

UPDATE: Here is our news story on the crisis at the US banks