Here's my problem with the financial regulation package that Sen. Chris Dodd has proposed: It hands the very regulators who failed us in 2005 and 2006 and 2007 and 2008 the responsibility for saving us next time.
If you worked backward from the bill to an account of the meltdown that produced it, you'd come up with something like, "We have a wise and brave class of regulators who did not have quite as much information or power as they needed to stop the financial crisis."
Anyone who remembers Alan Greenspan and Ben Bernanke dismissing the housing bubble knows that's not quite right. But regulator failure is a fact of life, or at least of bubbles. It's pretty much in the definition of a bubble that the relevant regulators don't believe there is a bubble; otherwise, they'd pop it. The trick is building protections that work even when the people in charge don't think they're needed.
The most successful of these is probably federal deposit insurance. Before the Great Depression, bank runs were an all-too-common occurrence. There were dozens in the years leading up to 1929. FDR's response wasn't to create a Commission on Bank Runs, tasked with watching banks and stepping in to insure their deposits if they got into trouble. He insured all consumer deposits. It didn't matter whether the chairman of the Federal Reserve thought your bank was playing nice. Your money was safe, even if he got it wrong.
If you want proof of how well it worked, ask yourself this: Did you line up outside your bank to close your account after Lehman Brothers collapsed?
The corollary today is capital requirements. When Lehman went down, its leverage was about 30:1. That means it had borrowed $30 for every dollar it had in assets. Leverage that high does a few things. First, it gives the bank more money to take risks, which banks like because it means higher profit. Second, it means that the bank has less money to pay back creditors if a bunch come calling at once, making failure more likely. Third, it means that if the bank does go down, it does more damage to the system because there are more people counting on the bank to pay them back.
Keeping capital requirements at manageable levels is financial regulation's killer app. But Dodd's bill doesn't spell out manageable levels. Instead, it leaves that up to regulators (mostly at the Federal Reserve) and gives them a bit more power to impose such requirements. "That's the story the regulators want to tell, that they just needed more tools," sighs Richard Carnell, a former assistant secretary of the Treasury for the financial market. "But they can set capital requirements now. They had ample tools, and they lacked the prescience and will to use them."
Consider this from the regulator's perspective: You think a bank is in trouble and needs some harsh regulations or, worse, to be shut down. Suddenly you're inundated by shareholders, bank executives, lobbyists and legislators from whichever state the bank is based in, all of them demanding you back off and assuring you that nothing is wrong. And who's on the side of shutting the bank down? Well, no one. And if you get it wrong or you lose the internal battle at your agency, that's a huge blow to your career.
The bill that Rep. Barney Frank passed out of the House took a more promising approach, setting capital requirements at a modest 15 to 1. That gives regulators a simple ratio to enforce rather than relying on their ability to figure out the right one to impose.
Another approach relies on the market: Luigi Zingales of the University of Chicago and Oliver Hart of Harvard University say the much-maligned credit-default swaps (essentially, bets that investors placed on various market outcomes) did a better job than regulators at predicting which firms would fail, and when. Bear Stearns, for instance, was over 100 basis points — the level Zingales and Hart identify as troubling — back in August 2007. The company didn't actually fall until March 2008. If you made some changes to the CDS market (such as putting it on exchanges so it's transparent), you could use it to trigger automatic regulatory action.
However you do it, you need something more objective, and less easily influenced, than individual regulators. They do their best, but their best has not been good enough.