Thursday, May 7, 2009 at 2:31 pm
The Treasury Department will finally be releasing the already endlessly leaked, contemplated and criticized stress tests in a few hours. After that we will be subjected to another confusing month or so of watching the banks deemed to be in need of more capital find ways to meet regulators' demands. Much of this, it appears, will involve the shuffling of paper, as banks convert preferred shares into the common stock that the Obama administration sees as the most important part of the capital base. In the case of the preferred shares acquired by taxpayers last fall at all the big banks, these may be converted into"mandatory convertible preferred" shares—a sort of common-stock-in-waiting.
These plans have been criticized from many quarters (too many to bother linking to) as, well, paper shuffling. Exchanging preferred shares for common doesn't bring any more capital into the banking system, it just tweaks the capital structure.
But here's what occurred to me this morning: Maybe bank capital isn't the issue here. Capital reserves matter a lot in a banking system with no government backstops: That's why, back in the 19th century, banks held capital stashes of 30% to 40% of assets. Capital was the only insurance against a bank run back then. Now the government insures against bank runs—and last fall it dramatically increased the scope of that insurance, with the FDIC explicitly increasing the deposit insurance limit to $250,000 and beginning to guarantee issuance of bank debt, and the Fed and Treasury implicitly making clear that no creditor of any of the biggest financial institutions would be allowed to lose any money anytime soon.
With that capital became largely irrelevant as a safety and soundness issue: Citi could have capital amounting to 2% of assets, it could have capital amounting to -2% of assets, and it simply wouldn't matter as long as the government guaranteed its liabilities—and that guarantee was credible.
Bank capital—and where it comes from—does matter in determining the most equitable way of allocating the costs and benefits of such a big government bailout. It also plays a role in shaping the financial system so it's less likely to implode in the future. That's what this whole stress test thing is shaping up to be about—forcing the shareholders (but not the creditors) of the more troubled institutions to cede more to taxpayers than the shareholders of the less troubled institutions have to.
If I'm understanding him right, economist Simon Johnson's solution to this whole mess is to explicitly guarantee all bank liabilities, subject banks to much tougher stress tests than we've just been through, then force those who fail the test into full government ownership. What we're getting instead is a muddier guarantee of bank liabilities and a muddier assertion of the government's rights. Is it the optimal approach? I don't think so. Is it a horrible approach? We won't know that for years, if we ever do."