Unconvincing "stress" tests

The much awaited results of the "stress" tests on top US banks is out- the idea is to see how much extra capital these banks would need under fairly realistic economic scenarios over the next couple of years. The estimate: $75 bn in order to be well capitalised by the end of 2010. The biggest requirement is at BofA: $34 bn.Total losses at the banks are estimated at $600 bn. Of this, $363 bn would be recouped through earnings.

What do we make of these figures? Two points. One, these losses come after the accounting standards were tweaked on mark-to-market losses of toxic assets. Two, the IMF had estimated US bank losses at $1.6 trillion over the next couple of years and estimated the extra requirement of capital at around $500 bn. The market is, therefore, likely to perceived the US Treasury as under-estimating capital requirements in order to hide the gravity of the problems at banks. That, in turn, means the chances of raising additional capital from the market would be zero.

Even otherwise, chances of banks raising additional capital are bleak. FT estimates total capital requirements at $55 bn, assuming some banks want to repay capital given by the US government and also some asset sales. This amount is three times total US equity raisings in the last six months. There you are.

Since banks can't raise capital from the market and the US government is unwilling or unable to pump in more, Richardson and Roubini, writing in the FT, say that insolvent banks should be closed down, with creditors taking some of the hit (depositors alone should be fully protected). What about the Lehman effect? The authors argue that market discipline would force banks to change their behaviour in ways that would instil confidence in the market, so credit would become available.

Well, well. For this process to work out would probably take years. So implementing this solution would require some nerve. Academics have the license to prescribe theoretically neat solutions but the global crisis requires a solution here and now because there is real suffering all round. I doubt that there will be any takers for the Richardon-Roubini solution where large banks are concerned.