Financial Regulation: How would you have it work?

Yesterday myself and several other financial bloggers got the chance to meet with several senior Treasury officials, including the Secretary himself. It was a fascinating experience and I have to admit, it was just plain cool to be within the bowels of power like that.

I am also on record as saying that Geithner was a good choice for Treasury secretary. We needed continuity as the bailout process was on-going. Geithner knew exactly where the bodies were buried in a way that other choices, such as Summers or Goolsbee wouldn't have. I have since come to view Geithner as a pragmatist, which I appreciate in anyone elected from the other party. And truth be told, a lot of the Treasury department's plans are working. I can't deny that. I panned the stress tests when they happened, but I can't deny that it worked. It created confidence where there was none. Say what you want about whether or not banks are still in trouble, I'm not terribly confident, but we're sure a lot better off today than January 19.

All that being said, I don't think much of the Administration's attempt at improving bank regulations, and I told them so. I even managed to do it politely without lacing in a Star Wars quote. Here is my main beef. I will explain myself in classic Accrued Interest style: very very long form.

The Administration's new regulatory scheme seems to focus on reducing bank risks. Higher capital requirements, more disclosures, etc. That all seems fine, except that is it really fundamentally different than what we have now? That is isn't it just an expansion of the same basic regulatory scheme that currently exists?

And if it isn't fundamentally different, does increased capital really solve anything? Citigroup had a large percentage of its risks off-balance sheet. Lehman's capital ratios were nominally quite strong on Friday, bankrupt on Monday. Both firms had adequate capital, and both either did or should have failed.

When this was pointed out to certain senior Treasury officials, their response was basically that they won't allow those sorts of games in the future. They were closing the loopholes that Lehman, Citi, and countless others exploited to hide their true leverage. I respond that even if you stop the games that banks were pulling in 2008, won't the banks come up with different games in the future?

The reality is that even if we do nothing, we might not have another financial crisis for many years. Let's say its 2022. Let's say we've just gone through 6 years of tranquility in the financial markets. Let's also say that banks have come up with some new and creative security where they get to keep all the upside with 85x leverage, but by the existing banking regulations it shows as a fully cash funded position against the bank's capital. Will bank regulators be motivated to ban this new security? I doubt it. Why do I doubt it? Because current regulators around the world looked the other way at CDO^2. Regulators aren't going to have the courage to challenge the banking industry during a period when the banking industry seems to have been right.

For the same reason I reject the notion of "regulatory supervision." Not to say that I think regulators are nefarious people, but are they going to understand the risks as well as the bank itself? And if the bank has a good reputation for taking risks, will regulators challenge them? Bear Stearns was known as the best mortgage shop on the street. Let's say you gave regulators dictatorial power, they could do anything they wanted. Would an omnipotent regulator have told Bear they were taking undue risks? Or would Bear have explained their positions, the regulator not understood them and assumed Bear was smart enough to handle it?

Transparency isn't a panacea either. Do you really think you are going to fully understand the risks at Goldman Sachs? As Yves Smith said yesterday, you'll never have Goldman revealing their trading book. And even if you did, it might not tell the whole story. By the time the disclosure, is published, their positions might be different. So what do you do? Put risks into categories? Like what? Credit ratings? We saw how well that worked! More transparency is better than less, but I'm asking the reader to be realistic about how much we're really going to know.

I'm a free market guy. I'd like to see any business be allowed to take whatever risks they can get funded. I don't want to tell what risks banks can take any more than I want to tell Macy's how many stores it should open or what flavor ice cream Coldstone should be selling.

Yes, I know. Coldstone isn't J.P. Morgan. But why not? Only because J.P. Morgan's failure has major consequences for other banks. But in a perfect world, we'd let J.P. take whatever risks it thought would make them money. That is, whatever risks the market would fund by buying J.P.'s debt and equity instruments. And if J.P. failed, then those investors would get burned.

Notice that this world wouldn't require regulators to predict where the next crisis would come from. It wouldn't even require banks to hide their leverage. The relative risk of a bank would be reflected in their cost of capital. If one bank was more aggressive than another, it would have to pay more for capital. I know, it sounds so idyllic, it can't be possible, right?

I argue that the only reason why it isn't possible is because we can't deal with a large bank (or insurance or brokerage) failure in isolation. There is always contagion. But there doesn't have to be. What if government regulation was aimed at limiting contagion post failure? It might be somewhat complicated and it might not completely eliminate all moral hazard, but its doable. Say that the government set up an FDIC-style insurance pool for over-the-counter derivatives and prime brokerage. Think of how radically different the AIG and Lehman failures would have been if no one was worried about having to face a bankrupt firm in a derivatives contract!

Like deposit insurance, I'd argue that such a regime wouldn't necessarily be costly to the government. Just like deposits, its likely that any firm's derivatives book could be sold to another firm, maybe at a loss, sure. It would be all the more easy so set up such a system if the more plain-vanilla derivatives, like interest rate swaps and most CDS were exchange-traded.

I know what you are thinking. Basically I'm saying to forget about preventative medicine and treat all patients only once they are in the ER. The problem is that regulation has done an absolutely horrendous job of preventing every crisis to date. Why do we think it will work this time? In fact, Yves Smith argued with me last night that the Basel II regulations, which are heavily credit rating oriented, helped to fuel the rise of the CDO. I agree completely. Where I disagree is the notion that a different regulatory scheme will somehow produce different results. I expect banks to do what they are incented to do.

We're seeing it already. Banks are loading up on Treasury bonds anticipating that those will get more favorable regulatory treatment in the future. Are we fueling a bubble in Treasuries? Maybe not, but the point is that regulation is inherently distortive. Replacing the old regs with new regs isn't going to change that simple fact.

So yes. I'd rather the government get out of the prevention business and get better at unwinding complex and systemically important financial institutions. It was really cool that I got the chance to tell Treasury just that. I don't know that its actually going to make a difference. But it was cool anyway.