This piece from liberal arch-nemesis Matt Steinglass is fairly typical of the responses to my post on financial regulation:
The point of regulators is that they are different from investors because they approach markets as neutral arbiters, who don’t stand to profit from any proposed instrument. They counterbalance banks not because they are smarter than banks, but because they don’t stand to personally or organizationally make any money if an instrument turns out to generate money, so their assessments are not colored by the tint of hypothetical lucre. Of course banks’ desires (to roll around in piles of freshly minted bills) are partially counterposed by their fears of risk (ending up standing on the corner wearing a pickle barrel selling pencils), but that’s still not the same as being impartial and financially uninterested. Judges are not smarter than lawyers, and basketball referees are not smarter than players or coaches. They’re necessary not because they’re smarter but because they are unbiased.
It is my understanding that judges do, in fact, tend to be picked from the top of their profession, but leave that aside. This somewhat misunderstood what I was saying. The point is not really that regulators are dumber than bankers--though in fact, government salaries simply will not allow financial regulatory agencies to get top talent. The people there tend to be young, getting experience for the next job (often in the regulated industry); people who cannot get jobs in the private sector; and a handful of extremely committed idealists (or ideologues, as you prefer).
But assuming arguendo that the regulators are every bit as smart and well-trained as the analysts they regulate. This is adequate only for certain kinds of regulation: the kind where the goals of the regulators are fundamentally different from those of the regulated.
If they could get away with it, some companies would lie in their advertising or sell adulterated goods; we want regulators to catch them at it. Companies have an incentive to present an inaccurate picture of their financial well being to investors; that's what auditors are for. Depositors in an FDIC-insured bank have no incentive to check on whether the bank is sound, so we put regulators in charge of doing it for us.
But what happened in the markets was not a case of fraud. It was a case of the systemic mispricing of credit risk. More importantly, it was a case of the systemic mispricing of credit risk on the buy side: Bear Stearns didn't fail because it had originated too many dodgy securities, but because it had bought too many. The banks have just as much incentive to price risk correctly as the regulators do--probably more, actually, because the regulators are unlikely to get fired if they miss one. It's hard to make a clear case for managerial moral hazard as a result of the Bear Stearns bailout--they all lost their jobs.
The FDIC does a pretty good job at what it does--ensuring capital adequacy and providing for rapid and orderly wind-up of the affairs of insolvent banks. But commercial banking is relatively uncomplicated. Consider something like a proprietary derivative pricing model--how will a regulator deal with this? And how do you walk an institution with an active trading book through insolvency? The Fed basically dodged these questions by selling Bear to Morgan, which has the ability to maintain trading operations. You can't run a trading desk if every order has to go through the receiver.
So if we say "Give a regulatory body broad powers", we are inherently stipulating that the regulator will have a better risk pricing model than the banks do. As of now, however, no one has a good pricing model for these risks. The regulator will probably be more conservative than the banks--but what reason do we have to think that the regulator's conservatism will be closer to the ideal balance than whatever incentive the banks have to substitute beta for alpha? The socially optimal level of credit risk--even systemic credit risk--is very far from zero.
I guarantee that merely by writing this post, I will get at least one angry blogger or commenter ranting that I am a libertarian moron who doesn't understand the difference between PROFIT and POLITICS. Au contraire. Both are incentives that work well in some contexts, not in others. Political incentives are not a good way to organize, say, one's agricultural output. They are a very fine way to organize one's wars--or at least, better than the alternative.
You cannot simply assume a priori that regulatory incentives will be more socially optimal than the profit motive. Profit, in this case, a pretty strong motive for doing what we want them to do: avoiding catastrophic failures. That's why I think that a powerful regulatory body is only an unquestionable win if you have some reason to think that it will be smarter than the banks.
I guarantee that merely by writing this post, I will get at least one angry blogger or commenter ranting that I am a libertarian moron who doesn't understand the difference between PROFIT and POLITICS.
This will probably be one of the same bloggers or commenters who argued, to the death of both credibility and point, that the Bear Stearns crash absolutely "was not a bank run", based on a pretzeled interpretation derived from Petty Technical Distinctions page 741, section 813, subsection 4.2(K)(a)(i)(3)(µ), as summarized by their brother-in-law, who works for the cleaning and maintenance divsion of the Paduca Savings & Loan (population: 450) and therefore knows what goes on in real banks.
I thought Dan did a good job explaining the incentive problems in his response to your earlier post:
"Regulators make their risk/reward judgments with the assumption that personally getting in trouble with their bosses is the biggest risk.
If they approve a business practice and it fails, they get in trouble. If they deny permission for a business practice that would have created thousands of new jobs and millions of dollars in tax revenue, nothing happens -- because, of course, you can't prove that such a thing would have happened, had the government not meddled in the economy. Regulators have no reason to care if the economy does well, if the business does well, or if the regulatory decisions they make actually make anybody better off. They have every incentive to forbid and restrict, and no incentive to approve and allow."
hat’s still not the same as being impartial and financially uninterested.
I should hope that our bankers are not "financially uninterested," but if so, you can see where it might get them into trouble.
And no, I have no substantive point. Thanks for asking.
If a company's assets are priced on models, then the executives have tremendous leeway to value them as they wish. This results in bonuses for the executives. They also have an incentive to keep problems hidden as long as possible in the hopes that the markets will turn around. So there is still a lot of potential for fraud and wishful thinking even though the problems were on the buyside.
My suggestion for regulatory reform is for all derivatives to be recorded with the Fed and capital be placed against them. Basically, a CME or CBOT for all off-exchange derivatives. Perhaps an open source component could be added by publishing the derivative book while hiding the identity of the counterparties. With many eyes on the derivative book, less extreme outcomes would be more likely.
As I have pointed out here amoung other places, and FWIW, Larry Summers concurs with the gist of what Megan is saying.
You're right that the profit motive provides all the incentive necessary, but being humans, we need more than the appropriate incentives to get around our natural biases.
From what I can see, in a generally rising market, it's human nature to systematically understate mid- and long-term risk in the face of short-term gain.
If you accept that's the case--and I submit any of the hundreds of traders who lost their asses in the last year or so as evidence for it--some mitigation of the problems caused by this systematic mis-evaluation of risk doesn't require knowing what the correct level of risk is, only that it's likely to be higher than those who are profiting believe it to be.
It is my understanding that judges do, in fact, tend to be picked from the top of their profession, but leave that aside.
I haven't laughed so hard since Anita Hill found my pubes on her Coke can.
We regulate institutions offering insurance very differently from financial institutions.
It's easy to talk about, but probably very hard to formulate, a set of financial institution regulations that basically require them to treat financial products that are the equivalent of insurance (like certain kinds of credit default swaps) as insurance. To wit, the product is not liquid but can only be resold to entities that are equally qualified to underwrite that product, etc.
Let me second Justice Thomas's comment about the caliber of the judiciary, especially at the state level.
Megan, you're being inconsistent, I think. It seems a priori true to you that regulators cannot be drawn from the top of the talent pool because their salaries are low, but you seem to understand that judges are drawn from the top of their profession. I'm pretty confident, though, that judges often make significantly less than the lawyers whose yammering they have to sit through. Of course, there's more of a prestige premium on being a judge than for being a regulator, but, on the other hand, the lower salary might make the last two comments about the state of the judiciary a little clearer.
The idea of a "neutral regulator" is so monumentally silly that it causes one to doubt that a productive conversation could be had on the topic of optimal regulation with someone who proposed the notion. One may as well discuss celestial navigation with a person who believes the earth is flat.
It is stunningly naive to believe that the government can attract people qualified to regulate financial institutions. The best and brightest in finance either want to be academics or earn hundreds of millions or billions.
Neither of these routes is available in government.
T.W. Andrews:
From what I can see, in a generally rising market, it's human nature to systematically understate mid- and long-term risk in the face of short-term gain.
...--some mitigation of the problems caused by this systematic mis-evaluation of risk doesn't require knowing what the correct level of risk is, only that it's likely to be higher than those who are profiting believe it to be.
in a generally rising market this may be so, however as Keynes pointed out, in a generally falling or flat market the exact opposite is the case. So what we need is a bias we can apply to the system, a bias that is smarter than the existing banking a finance system and hence can work out which bias is applying at this point.
Dave,
The best and brightest in finance either want to be academics or earn hundreds of millions or billions.
Neither of these routes is available in government.
You're rather optimistic to make that statement.
n a generally rising market this may be so, however as Keynes pointed out, in a generally falling or flat market the exact opposite is the case. So what we need is a bias we can apply to the system, a bias that is smarter than the existing banking a finance system and hence can work out which bias is applying at this point.
The bias doesn't have to be that smart. We don't need to know the correct price of risk to add some additional cost to it any more than we need to know the "correct" price of carbon to put a modest tax on it.
It didn't take a genius to tell that the market for mortgage backed equities was generally rising. Similarly, it's not rocket science to say that the market is flat or falling for those same equities now.
It's not necessary for regulators to know what the correct price is in order to apply some correction, only to be fairly certain that the current price is wrong, and in which direction. If those conditions are met it seems to me that a modest "risk adjustment" cost could be added to the relevant transactions. In cases where those cases are not met, no adjustment could be applied (or even a negative one if it's clear that human nature is making investors too risk averse, as would be the case in a falling market).
To be fair, I don't know the ins and outs of financial markets and so can't offer any thoughts about how to implement such a scheme. It still seems to me though, that just because regulators don't have perfect information, we must act as if they have no information. It's been repeatedly demonstrated that given a perceived opportunity for gain, people underestimate risk. If we're going to socialize the cost of this mis-evaluated risk, it seems we could raise the cost of it without needing to know more than the buyers of various financial instruments.
TW Andrews,
The issue is relevant insofar as the hypothetical regulator has to decide how much of a risk charge to apply. As Ms. McArdle notes, the optimal level of risk is not zero. But, the incentives for the regulator will inevitably be in favor of keeping risk at zero. Hence a discretionary policy of adding cost for risk will always yield a suboptimal level of risktaking.
TW Andrews--yes, the regulator can certainly simply develop a strong bias against risk. But you're assuming that this will be more socially optimal than the level of risk we have now. This is simply availability bias--the problem we just had seems the most vivid, and it's much harder to see the problems we have when regulators prevent people from taking risks that would have paid off. You need an actual argument as to how the regulator will be better at delivering a socially optimal level of risk than the banks, not simply a statement that the regulator will deliver less risk.
Think of it this way: some people would argue that we need regulations to cut down on the amount of food that Americans eat. But a regulator who viewed its goal simply as reducing food intake, without knowing how many calories people actually need to live, would be worse than the current state of affairs. At least until the regulators starved to death.
A couple of random comments. First, it's probably true that a system of government regulation would impose a "suboptimal" level of risk.
However, as others have argued, and as the recent financial problems have shown, it's far from clear that existing markets are allocating risk optimally either. The benefits of short term profitability and, potentially, the moral hazard of being too big to fail, can tend to push financial executives towards risk levels that are also "suboptimal," only on the high risk side. Yes, the CEO of Bear Stearns lost his job, but how many million dollars did he take home before doing so? I'd guess he's not exactly hurting at this point.
I'm not convinced the analysis of an optimal risk based on some sort of simple expected profit level is the right approach, though, which is why I scare-quoted suboptimal. What is the unit of analysis you're using to optimize? For a company executive, it's obviously the company itself. But the costs of a large bank failing are not confined to that bank--they could trigger larger financial impacts across the whole sector. Which is why the feds are willing to step in. So now you've got a case where the interests of the country imply a lower level of optimal risk than the interests of any individual bank, because of the external costs. Which implies the need for a regulatory regime to force banks to take fewer risks.
In addition, I think the potentially large costs of a financial sector meltdown call for a risk-averse approach, and hence it's much better to err on the side of over-regulation than under-regulation.
Lastly, I think the issue of intelligence is pretty much irrelevant. The average government regulator doesn't need to be smart enough to craft a regulatory scheme, she just needs to enforce a regime that others have crafted. It only takes a few smart people to devise the scheme in the first place, and I'm confident that they could be found in academia or elsewhere, when needed. As you say, it's no guarantee. But I think some sort of regulatory leavening that would reduce the extremes of a natural financial boom-bust cycle is a good thing.
Judges tend to be drawn from the top of their field despite the relatively low salaries because of the power and prestige associated with being a judge. For many lawyers, the lower salary could be worth the trade off.
Not so the regulator. I don't think there's anyone on the planet who'd give up the money they'd earn in the private sector for the dirty looks they'd receive as a petty bureaucrat. And a financial regulator, come on- no one with the pure, honest-to-goodness heart of a do-gooder goes into finance.
The average government regulator doesn't need to be smart enough to craft a regulatory scheme, she just needs to enforce a regime that others have crafted.
You are under-estimating the difficulties of implementation. A hell of a lot of smart plans have been wrecked by poor quality implementation. For example, hospitals have significant problems with getting the right medicine to each patient. Also, another detail, while smart people may design the scheme, it needs to get passed into law by politicians. And I think climbing the political ladder depends more on brillance at horse-trading than at finanical regulation.
The older I get the more and more important I think implementation is.
"it's far from clear that existing markets are allocating risk optimally either. The benefits of short term "
I'm willing to state with 100% certainty that the existing markets are not allocating risk optimally.
What I don't know is, in what way are they wrong. The current problems were caused by people underestimating the risk of a certain class of assets. Tomorrows problem could well be the opposite.
Several years ago, anyway, a major complaint about banks was their unwillingness to make loans to "traditionally disadvantaged" groups. If that complaint is valid, that problem is in overestimating the risk of a certain class of assets.
Assuming that that complaint was accurate, how is a regulatory body going to neutrally deal with this? And, realistically, how is it going to be neutral, even if it was smart enough to know better?
I agree with Megan's points that regulators are unlikely to be better at analyzing the risks involved than traders and banks. Regulatory decisions would also be suspect because of their supposed bias towards zero risk, potential regulatory capture, and so on. All of these are plausible concerns, at least. But do we really need more "active" regulation to address weak points in the financial system as opposed to simply better information?
What if we just put stronger disclosure requirements on off-balance-sheet investments (and on their holders, effectively putting them partway back on the balance sheet) and on CDO's and similar securitized instruments, so that investors have more information on what they're really buying. The crux of the CDO crisis seems to be that everyone largely relied on representations that they were high-grade debt, when in fact they were junk. Shouldn't more transparency fix that pretty easily?
Maybe your idea of regulation is too complex.
If the regulations were limited to leverage and off balance sheet financing it would not take too much intelligence to implement it.
On the other hand what did the system get by Bear Sterns being 25 fold levered?
What did the system gain from allowing financial institutions to hide their loans, investment from both public regulators and private investors in off balance sheet creations?
Maybe you can argue that it allowed the US to increase the share of homeowners by a few percentage points. But maybe that was not really an improvement in the well being of our economy-society.
As a general rule you do not have to be a general to fight the last war. Everyone does it. But for a few decades after WW II not allowing financial institutions to repeat the same old mistakes worked very well. We really only had systematic
problems when the regulators , i.e., politicians
and business leaders decided it was OK to allow financial institution to figure out new ways to get around the old regulations. Essentially this meant we allowed financial institutions to figure out new ways to make the same old traditional mistakes.
The problem is not the letter of the law. It is the spirit of the law. From the depression until
around 1980s we looked at these two things as being as being the same thing. But around 1980 this changed and the regulators started only enforcing the letter of the law and ignoring the spirit of the law. If some smart lawyer could find some way to get around a regulation the regulators started letting them get away with it. It was not that the regulators got a lot stupider around 1980. Rather, we made a political decision to allow it to happen.
Think of it this way: some people would argue that we need regulations to cut down on the amount of food that Americans eat. But a regulator who viewed its goal simply as reducing food intake, without knowing how many calories people actually need to live, would be worse than the current state of affairs. At least until the regulators starved to death
Sure. I'm not saying we should institutionalize a strong bias against risk, any more than we should a strong bias against eating. But, in an where people can make short term money by ignoring middle and long-term risk, they will systematically understate that risk. In such an environment, adding some moderate cost to risk seems like it would be a good idea in the same way that adding a moderate cost to food would be a good idea in an environment of expanding waistlines, but terrible in a time of food scarcity. The regulator or regulation would need to adjust to circumstances, and be wiling
I think of it this way--if once it's obvious that prices are steadily moving upwards, (and I don't think this would have been a tough call in the most recent case, nor for tech stocks in the 90's or tulips in the 17th century) a regulator could add a modest cost to transactions, it would offset some of the people's systematic bias towards underestimating risk. Once the market has clearly turned down, this cost could be eliminated or even made negative (since in a down market people are systematically risk-averse).
I don't see that a regulator would need to be that clever to manage this.
It might be politically implausible, but I don't see wh
Some facts may be useful
The point is not really that regulators are dumber than bankers--though in fact, government salaries simply will not allow financial regulatory agencies to get top talent. The people there tend to be young, getting experience for the next job (often in the regulated industry); people who cannot get jobs in the private sector; and a handful of extremely committed idealists (or ideologues, as you prefer).
Yes, we regulators employ recent MBAs and Masters candidates; often from top schools. We also have quite a few individuals who formerly worked in the private sector: mostly they have joined the regulatory community for a change in lifestyle. We also employ significant quantities of PhDs. Finally, former collegues of mine work in fairly high level positions at virtually all of the major banks and IBs. What we may lack in narrow technical understanding we make up with a breath of observations by being able to see practices across the industry.
For Dave who states "The best and brightest in finance either want to be academics or earn hundreds of millions or billions." No, we do not generally make as much as the bankers (particularly those on the deal side); much of our staff have joined in part due to a commitment to public service. There are motivators in this world other than money (though let's be clear, we like to be adequately compensated for our jobs)
For Ann who states "Regulators have no reason to care if the economy does well, if the business does well, or if the regulatory decisions they make actually make anybody better off. They have every incentive to forbid and restrict, and no incentive to approve and allow." Please give us more credit than this. Some of us are acutely aware that the reason we seek to maintain a healthy financial system is so that capital and savings can be efficiently channeled to productive investment, and so that monetary policy can be executed effectively.
For all those who suggest a regulator will always seek a socially sub-optimal level of risk asymptotically approaching zero, I'd point out that despite regulation numerous banks and investment banks have managed to lose quite a bit of their capital base (something we are not per se proud of).
I'd love a response on this:
I find your motto "asymmetrical information" ironic in light of your current argument, for it is asymmetric information and moral hazard that is at the heart of the problem. This is the fundamental case for regulation. You take for granted that banks only maximize profits and therefore they have incentive to be smart. Are you sure that the financial and professional incentives of the executives and managers of these banks are fully inline with the interests of shareholders? Look what Merrill Lynch did to their risk managers, they reassigned them when they began to speak up about the excessive risk stemming from the CDOs on their books. Look what Moody's did to their analysts of mortgage backed securities, they caved in to issuers' demands that they should assign softer analysts. This is the pursuit of short-term profits, which often are in the financial and professional interests of executives and managers. When everyone starts doing this you get the systemic risk we are now witnessing.
Theoretically, it would be nice if there was a way to get more transparency in banks' balance sheet reporting, but the devil is clearly in the details there.
OTOH, it's waaaay too early to start thinking about how to change regulations to fix a problem we haven't even seen the half of yet. imo, there's lots more to learn about what went wrong here.