The recapitalization of Bank of America today, along with their announcement to temporarily halve dividends, resulted in a fairly substantial stock price decline. Simple-minded analysts will look at this and see their expected returns over the next few years drop, both as a result of the dividend cut and the dilution of their shares.

However, this is arguably the wrong way to think about things. Over the last year, banks have talked about their commitment to existing shareholders; but when this supposed commitment to non-dilution of existing shares results in total loss of equity value and bankruptcy, it's hard to play this out as being in the interest of shareholders. Would you rather hold diluted shares, or worthless shares because the company went under?

Many of the failed and troubled banks have attempted to raise capital over the past year (with Bear Sterns being a notable exception). But banks like Lehman Brothers raised only modest amounts of capital (less than 20% of I recall correctly). And the reluctance of banks, in general, to recapitalize following losses left them with only one other option: liquidating assets. And the assets they were liquidating were frequently some of the few profitable units. No one had any interest in buying the rest.

And it's not that no one tried to sell them. It's just that everyone recognized that they were one of the more risky portfolio components, and attempted to liquidate them in an attempt to lower weighted portfolio risk. If the banks could actually hold these securities through their maturity, they might survive. (This is one of the reasons the government won't do as badly as the headlines would have you believe with the purchase of so-called toxic assets -- this is a special case in which market value is below fundamentals.) But when you finance long-term assets with short term debt, you have a temporal risk mismatch, and you can sometimes get screwed.

The liquidation of prime assets meant that while companies might have acquired modest amounts of capital, the weights on the riskiest parts of their portfolio went way up. And there's substantial reason to believe that these weights were not properly updated, and that the volatility of the most risky portfolio components was drastically underestimated.

My opinion: If you're a bank regulator and you want to trust Value-at-Risk, go ahead. Investors beware. The entire idea of this kind of risk assessment is to ensure that banks possess enough capital to survive rare, disastrous events. But Value-at-Risk varies with the business cycle, and this primarily because the historical data used to predict portfolio risk is, in general, pretty short. So if you're using two years of data to try to estimate the likelihood of an event that's going to happen once every thirty years, this seems pretty stupid, right? And this is where derivatives are useful: we don't have decades of data. If you wanted to be careful, you could get a better idea by looking at historical fundamentals and the ways in which fundamentals would have affected derivatives at the time, had they existed. But you don't want to be careful: you want your end-of-year bonus.

Postscript: Apologies for the diversions. It's a known personality flaw.

Disclaimer: Advice is roughly worth what you pay for it. And don't mail me a check. I really wouldn't have an answer for you if you paid me!