Too Big To Be Allowed to Exist?
My generic view of regulations, is that we ought to try to have fewer, but clearer and perhaps harsher. You want regulations that can’t be gamed by the regulated and don’t depend on hubristic assumptions about what the regulators are actually going to do in practice.
To that end, what’s the deal with banks that are too big to fail?
If we can identify such banks, why not try to make a rule preventing banks from becoming that big? As a tradeoff, banks that rested in the small-enough-to-fail category could be allowed to operate with much, much laxer oversight and regulation since everyone would understand that if they fail they’re going to sink. Presumably, there are some efficiency gains associated with the economies of scale involved in big financial institutions. But there would also be efficiency gains associated with relaxing the regulations on financial institutions. And the only reasonable way to seriously relax those regulations would be to commit to a no-bailouts scenario. But to do that, we need to make sure the banks aren’t too big to fail. So why not focus the regulatory effort on that — on making sure that institutions don’t get so big that they need bailing out?
63 Responses to “Too Big To Be Allowed to Exist?”
Register
RSSLogin
Topic Cloud
ArchivesBlog RollAbout Matt YglesiasContact Me
Donate
November 25th, 2008 at 4:03 pm
I hear they have something called “anti-trust” for this.
November 25th, 2008 at 4:04 pm
This is probably the most important idea you’ve raised–maybe ever. As the 9th most important person in your cubicle, you should use all of your influence to make something concrete out of it.
November 25th, 2008 at 4:14 pm
This is a Matt Yglesias all-star post. This post just reminded me why I read you.
I second the demand for something more concrete.
November 25th, 2008 at 4:15 pm
You can blame Greenspan for this one, too. He felt that US banks were too small, and was a persistent advocate for consolidation.
November 25th, 2008 at 4:17 pm
Would this be a variation of Clarke’s Law: Any institution sufficiently large enough is indistinguishable from government?
November 25th, 2008 at 4:17 pm
Wow. This is a really compelling, elegant not to mention obvious argument that I have not heard anyone making through this entire time. I probably agree with Rich, in fact — maybe Matt’s best thought ever. Hence it would of course be laughed out of the room anywhere in Washington or on Wall Street. But that’s no reason not to explore it further.
November 25th, 2008 at 4:18 pm
I sometimes read libertarians claiming that if the government was smaller and regulated less, the economic big boys wouldn’t be buying it to get what they want.
At this point, I beging to giggle.
You can shrink the government all you want; if it’s in the interest of someone the size of Citibank to have the government do something, they’ll just blow it back up to where it’s big enough to do the job.
Of course the government is going to come to the rescue when an institution that large might tank. From a Marxist perspective, of course they’re going to get what they want from the state. From a pragmatic perspective, no responsible person left right or center is going to sit back as 2 million jobs and a good chunk of GDP vanish.
The only way to stop this from happening is to have lots of small and medium sized businesses, rather than a few monsters.
November 25th, 2008 at 4:20 pm
In 1969 Clorox was forced out of Proctor and Gamble because it was thought too strong and stifled competition.
When financial institutions get too big, make it worth their while to spin off pieces.
November 25th, 2008 at 4:24 pm
This is an incredible post.
November 25th, 2008 at 4:24 pm
Hey, Matt, try doing a little research first.
Start with the following:
– Banking Act of 1933 (Glass-Steagall Act)
– Bank Holding Company Act of 1956
– Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994
– Gramm-Leach-Bliley Act of 1999
Once you understand the history of banking regulation and deregulation in this country, you might start to understand why there aren’t effective restrictions on banks becoming “too big to fail.”
November 25th, 2008 at 4:25 pm
the economies of scale involved in big financial institutions
Really? Name one. All this “everything under one roof” claptrap courtesy of Sandy Weill ca. 2000 has been proven wrong. Investment banking is a commission business with a business model of a law firm that, every couple of weeks (or days) takes out an enormous loan in the morning which it repays in the afternoon.
Investment banks are not investors. They do not take bets except that their underwriters are sufficiently prepared so that one-day market moves don’t spell disaster. Any position that an investment bank takes extending past sundown should have no expectation of taxpayer backing.
November 25th, 2008 at 4:25 pm
I especially like the two-tiered approach Matt suggests. Basically, if you’re too big to fail, you’re big enough to regulate the shit out of. I’m not sure a flat-out limitation on bank size will work; also being too big to fail isn’t just that simple — the extent of the connections of the bank’s investments to the rest of the world economy matters as well. But the concept of a regulatory environment that takes into account bailout risk to taxpayers is key.
November 25th, 2008 at 4:25 pm
Counterpoints: 1. Not sure the mechanism by which it would be made impossible ex ante to have bailouts ex post. Laws can be changed, there’s a time consistency problem.
2. My understanding, based on an article or two, is that Canadian banks have been much less subject to collapse, both now and in the actual great depression, in part because they’re diversified across the provinces (i.e., large). Also in part due to greater regulations on what they’re allowed to do with their money and higher capital requirements.
November 25th, 2008 at 4:27 pm
I thought this was sort of the point of Glass-Steagall. They separated commercial banks from investment banks to prevent enormous combo financial institutions from being created. Thus you avoided a scenario in which a huge combo bank (with both investment and commercial branches) lost a lot of money on investments and went bankrupt, thus requiring the FDIC to spend a ton bailing out depositors.
I guess Matt is making a slightly different point, but the overall idea seems much the same.
November 25th, 2008 at 4:29 pm
I believe that Black Swan author Nasim Nicholas Taleb and fractal-math guy Mandelbrot are saying something essentially similar in this video linked on BoingBoing… they may also be arguing against the globalization of banks: Makes me imagine giant dominoes hitting each other and causing a chain reaction:
Black Swan author says present economy worse than Depression
November 25th, 2008 at 4:30 pm
Scaling the severity of regulations with the size of the bank is an excellent idea. Huge banks can survive stricter regulations because of their enormous advantage from the economies of scale. Smaller banks would then be on a more level playing field, without stifling regulations.
November 25th, 2008 at 4:33 pm
Sounds like you are talking about that crazy concept called Rule of Law – you know, where a governing body is guided by a very strong set of Rules like say, oh I don’t know, a Constitution.
I think the simplistic and opposing views that the government should “do something” or “do nothing” is what a lot of fighting & frustration is about. Inaction could result in as much harm (like letting fraud go unpunished) as much as action could (like doling out money to undeserving entities).
There should be less emphasis on the “doing” than whether the doing or not doing is consistent and predictable (and dare I request, principled, such that any Rules end up not harming or favoring a particular group, class, or industry?)
But that’s just me.
November 25th, 2008 at 4:34 pm
Also, didn’t easing regulations on a smaller category of banks lead to the S&L crisis? A lack of regulation on smaller banks may run into a “too many to fail” problem.
November 25th, 2008 at 4:39 pm
Isn’t this why we used to have Anti-Trust laws on the books? And haven’t these been flaunted for years? Ignored by our government, which is supposed to enforce them?
To the normal American, no company should ever be able to get “too big to fail.” If we’re going to have regulations, that should be the number one goal. Because too big to fail leads to too much moral hazard.
November 25th, 2008 at 4:39 pm
One would think that a CEO making millions would know not to take on too much risk, without guvment bureaucrats telling him what’s too risky and what’s okay. Obviously, one can be wrong.
November 25th, 2008 at 4:41 pm
Actually, blah, a lot of S & Ls were regulated to death by being stuck with a portfolio of fixed rate loans mandated by the regulations; then there was an overreaction and it all went the other way. But a lot of banks got too big to fail becuase they took over everyone else who was failing.
November 25th, 2008 at 4:45 pm
What if there are only little banks, but they all make the same mistake that bankrupts them all? Then you’ve got a systemic failure that needs a whole lot of bailing out. To prevent that you’d have to regulate them all to make sure they don’t make that mistake. Which means you’re back to square one, no?
This happens, right? If not all banks, some critical mass of them messes up in basically the same way and gov’t steps in to prevent a calamity.
November 25th, 2008 at 4:45 pm
If Matt expects big reforms on the banking front he won’t find this reassuring:
FROM: http://www.thenation.com/doc/20081208/greider_web
“On Monday, Geithner was busy executing the government’s massive rescue of Citicorp–the very banking behemoth that Geithner and Summers helped to create back in the Clinton years, along with Federal Reserve chairman Alan Greenspan and Robert Rubin, Clinton’s economics guru. Now Rubin is himself a Citicorp executive and his bank is now being saved by his old protégé (Geithner) with the taxpayers’ money.
The connections go way beyond irony. They raise very serious questions about where the new president intends to lead and whether he has the nerve to break from the weak and haphazard strategy of the Bush administration. It has dumped piles of public money on the largest financial institutions and demanded little or nothing in return, hoping for the best. Geithner has been a central player in the deal-making, from Bear Stearns to AIG to Citi. The strategy has not only failed, it has arguably made things worse as savvy market players saw through the contradictions and rushed out to dump more bank stocks.”
November 25th, 2008 at 4:47 pm
Anti-trust laws are supposed to protect the consumer from monopolies and cartels that keep competition out of the marketplace. Nothing at all to do with “too big to fail.”
November 25th, 2008 at 4:49 pm
Truly an A+ for the day, Matt. Cutting thru the BS!
I vaguely remember the argument being made that we needed big-swinging-dick banks as big as those in Germany, Japan, and the other leading economies. I guess that didn’t work out so good….
One counterargument I can imagine: That a thousand small banks going tits-up are even harder to save than six behemoths. And the current legitimation crisis is such a tsunami that not just a few of a thousand small banks would be likely to go under. So you might end up with the same essential problem.
November 25th, 2008 at 4:57 pm
I think we need to be very cautious about limiting the growth of Standard Oil, since it plays such a large role in our economy.
November 25th, 2008 at 5:02 pm
I agree. I also would go further and require that if any financial institution is too big to fail and ends up needing a bailout, that the bailout should necessarily result in the breakup of the company into pieces that are not too big to fail.
Also, in order to prevent the moral hazard, bailouts need to be similar to bankruptcy in that the end result should be a complete or almost complete wipe out of shareholder equity. In chapter 11, shareholders are wiped out and debt holders normally become the new equity holders. Likewise any bailout should be considered a type of bankruptcy, where the new shareholders are a combination of the old debt holders and the “bailers” (the government). Naturally, this should in most cases lead to a radical change in the board of directors and senior management since the old leadership failed. The government, naturally would want to sell its interest in the company once the company is back on its feet and the crisis is over. After all, we don’t want the government to own banks; it’s simply the only way to recapitalize a bank without just giving them money.
November 25th, 2008 at 5:11 pm
Anti-trust laws work about as well as eating lo mein with a spork.
Matt is right and needs to flesh this out a bit. Just pick a number and that’s how big your enterprise can get. If get approach the limit on size, split the company into two entities.
November 25th, 2008 at 5:14 pm
When regulatory agencies allow a corporation to get so big that avoiding its failure becomes necessary and also costs the taxpayers hundreds of billions of dollars, they are not protecting the consumer very well, are they? And wouldn’t an institution that is so big that its failure threatens the economy get that big by stifling competition in the marketplace? I seem to recall there used to be a lot more banks, for example, but then for decades we had a loose regulatory environment that allowed massive leverage and approval of every merger, regardless of its effect on the consumer.
November 25th, 2008 at 5:17 pm
Yeah, I was going to say the same as blah, I thought a smaller but less-regulated bank is basically what a S&L was. Flo says there’s more to it than that, but either way I think the point is the same – to the extent that “too big to fail” means anything, it applies to sectors of the economy, not to individual companies. A hundred relatively small businesses can inflate economic bubbles and create fake paper wealth as easily as half a dozen large ones if they all have the same incentives to follow.
And as for the first paragraph of Matt’s post, what does it actually mean? Who’s out there arguing for gameable regulations or assumptions? Even Grover Norquist wouldn’t say he wants that, and yet regulations like that just sort of appear, so what does that tell us?
November 25th, 2008 at 5:19 pm
This will not work. To illustrate take this thought experiment:
Let’s say that this idea is enacted with the force of law and all large banks are split up into tiny bite size pieces. Consequently, the nation is left with tons of small banks that are not regulated with regard to say… credit default swaps. We’d still have tons of small banks that followed the incredibly profitable trend of the time and sold tons of these dubious financial products to each other and to their investors. These small banks would also be following another very profitable trend of the time in giving out risky ARM mortgages and breaking them up into obfuscated financial products. All completely unregulated.
Now, explain to me how this situation is any better than what we currently have??!! When the mortgages inevitably start to default and investors and banks begin cashing in their credit default swaps… we’d still soon discover that we have tons and tons banks that are way over-leveraged. The small banks would begin to fail as the credit market started to seize up and banks refused to lend to each other.
Correct me if I’m wrong, but the fact that it would be 10 small unregulated banks dealing in these credit default swaps and becoming hugely over leveraged versus 1 large bank doing so would not change the basic equation of what is happening.
November 25th, 2008 at 5:22 pm
YES! I’ve been thinking this same thing.
Large mergers have to go through anti-trust approval first. Is “the potential to become too big to fail” one of the metrics used to grant approval in this process? Probably not, but it should be.
Sure, banks could get too big through organic growth, but that’s not usually how it happens. It almost always happens because they merge with the competition at some point in their corporate life.
So, given that you already have a process in place to stop mergers on anti-competitive grounds, you could probably just add on one more question: “will the resulting combined entity be too big to fail?”
November 25th, 2008 at 5:22 pm
Some one else put it much more succinctly.
This would just lead from “Too Big To Fail” to “Too Many To Fail”!
November 25th, 2008 at 5:25 pm
My understanding, based on an article or two, is that Canadian banks have been much less subject to collapse, both now and in the actual great depression, in part because they’re diversified across the provinces (i.e., large). Also in part due to greater regulations on what they’re allowed to do with their money and higher capital requirements.
Traditionally banks were kept small in the US by reducing the geographic area in which they could operate. This made them susceptible to local economic shocks. Conceivably you could have small banks (at least in comparison to Citibank) that are geographically diversified.
However, through the FDIC tax payers are implicitly backing the risks of all banks big and small. Because of this the pressure depositors place on banks is not take less risks its give me a higher return. Nor do shareholders necessarily want banks to avoid risk, because they can diversify they’re own portfolio. In this environment we need regulators to come in and put a ceiling on the degree of risk banks can take on.
It is true that without FDIC protection banks would likely me much more risk averse but its easy to see in a panic even banks that acted prudently could be subject to bank runs.
November 25th, 2008 at 5:28 pm
When regulatory agencies allow a corporation to get so big that avoiding its failure becomes necessary and also costs the taxpayers hundreds of billions of dollars, they are not protecting the consumer very well, are they?
Well, if a company is allowed to poison America’s water supply, that certainly means that regulatory agencies have failed to protect the consumer. That doesn’t mean that anti-trust regulation is the answer.
As has been pointed out a few times in this thread, the problem is the financial sector’s entanglement with the rest of the economy, not the size of the institutions within the financial sector. If we cut Citigroup into 1,000 smaller pieces that are each small enough to fail, what have we accomplished exactly?
November 25th, 2008 at 5:38 pm
“You want regulations that can’t be gamed by the regulated”
Yeah, we all want that. In this fallen world, though, it’ll never happen. Tax law is a useful example: it’s a body of really formal law, full of what look like bright-line rules. But arbitrage and sham transactions are consistent problems, and there’s no sign they’re ever going away.
The most you can do is respond to particular types of evasion when they become widespread.
Perhaps you can imagine a tax code simple enough to be unambiguous; but it would impose other, enormous costs, by not being a very good tax code. (For example, it’s hard to imagine an income tax that doesn’t let businesses offset their costs against their revenue; but that offsetting is the source of an enormous amount of complexity and arbitrage.)
So I think this is kind of a fantasy—it’s not that there’s not room for incremental improvement; it’s just that the world is unavoidably complicated, so our regulations will be, too.
November 25th, 2008 at 6:06 pm
I hear they have something called “anti-trust” for this.
Unfortunately, in recent years the Supreme Court has rewritten antitrust law so it no longer enshrines the concept “big is bad.” Interestingly, many of the so-called “liberal” justices are very pro-business. Their “liberalism” extends only to hot button social issues.
Likewise, the media ignores the Supreme Court’s business rulings. So we hear endless discussions over “important” decision about sonar and whales or ten commandment postings in public parks, but nothing about antitrust or how the heck is it that state usury laws cannot constrain credit cards.
Apparently it is vitally “important” whether or not gays can marry; but whether gays – as well as everyone else – loose everything in an economic collapse is not so “important.”
November 25th, 2008 at 6:24 pm
It’s exciting to see so many people love this idea.
I was talking about it, um, 5 weeks ago.
November 25th, 2008 at 6:37 pm
Did they slip you something over there in Switzerland, MY? Maybe a tab of something with Ayn Rand’s face on it?
This post begins with an absurd assumption, and moves to an even more absurd conclusion. The absurd assumption is that “My generic view of regulations, is that we ought to try to have fewer, but clearer and perhaps harsher.” That is like saying, we ought to have fewer shoe sizes. Why should we have fewer? Is there a particular reason? Have you been going through some regulatory book? In essence, this libertarian canard is that there is some ideal norm against which we measure the regulation of enterprises, but this idea seems to have no empirical validity whatsoever. It doesn’t even make much sense, since enterprises and the contexts they operate in change. So, an enterprise could have a regulation that is out of date, or one that you believe is unnecessary or pernicious – but that doesn’t exclude that it doesn’t have a regulation that is needed. There’s no way of telling if the one side is larger or smaller than the other, a priori. It is always an empirical test. So your generic view is, to put it mildly, silly. Or less mildly, it is consistent with the pernicious discourse of faux libertarianism that George Bush likes to patch out his speeches with.
As for the view that small banks should have lax regulations, why? What good would it do? Unless, of course, the positive is that it would create more Randian freedom. It would create, along with that, a lot of fraud, much chaos in the communities the banks serve, and ultimately lead to things like the S and L collapses in the late 80s, which followed just that suggestion.
Seriously, MY, you do need to hang around with a group that isn’t so heavily into worshipping Uncle Milton F. Because they are getting to you.
November 25th, 2008 at 6:47 pm
I think limited liability might still be a problem even for small banks. It means there is no downside risk. They can put their investors money on the craps table and if they loose bankruptcy and if they win they can take their cut.
November 25th, 2008 at 7:04 pm
Has MY ever heard of a run on the banks?
having many small banks failing — even, say 1%, will cause investors — and depositors, geesh — to lose confidence in the banking system — thus causing the very problem we have today.
Will MY pls stop commenting on finance. It’s embarassing, really.
November 25th, 2008 at 7:08 pm
Will MY pls stop commenting on finance. It’s embarassing, really.
If people couldn’t comment on stuff they know nothing about, the internet might cease to exist.
November 26th, 2008 at 1:40 am
The trick is drawing a line between efficiency of scale and cartelization. Would the Wall Street meltdown have occurred if there were several dozen large financial firms (as there were not so long ago) instead of a half-dozen behemoths? Would Detroit be on the brink if the auto industry were a not-quite-as-big six?
November 26th, 2008 at 10:38 am
I thought we had spent the last 30 years making it possible for companies to be too big to fail. Its going to be hard to turn it around, even if its needed. A lot of people will be upset when we break up companies like citi group, but it does need to happen to allow the free market to work. As the free market only works if companies can fail, so oddly that means the free market needs regulation to remain free.
November 26th, 2008 at 12:08 pm
Or, you could just have a strict and credible no-bailout policy. Any bank that is too big to fail will start shrinking because people are too afraid of being associated. And if any bank remains too big to fail, well, it will soon fail, not get bailed out, and everyone dumb enough to work or invest in it will be delivering pizza and we won’t have to worry about them ruining the world economy again.
November 26th, 2008 at 12:21 pm
I’ll second (or third or nth) the above comments about “this is why I read Yglesias”.
This post should be in an idiom dictionary next to “thinking outside the box”. Although, yes, it clearly needs a note crediting urbino with the same idea.
Myrtle Parker – I think the idea is that the worst managed 10% or so, not having a financial behemoth to hide their activities in, would go down before the other 90% were beyond hope. This, of course, is guesswork at this point.
November 26th, 2008 at 1:04 pm
Two thoughts.
First, there is no need to prevent banks from becoming TBTF or to split them up when it happens. A more closely tailored solution is to force them to buy insurance from the private market adequate to cover their bailout in case it should have to happen. Obviously it’ll take some regulation to make sure the insurance is truly adequate, but this seems like a better way to force banks to internalize the externality that TBTF creates.
Second, TBTF is harder to assess than you allow for. A hundred small banks that each had the same leveraged portfolio as a single gigantic Citigroup would pose the same problem; when the underlying securities collapse, all hundred banks will collapse simultaneously, thus creating the same issues as when a single gigantic Citigroup collapses. If the government is going to get into the business of assessing TBTF — and perhaps it must — then to do it right, it would need to categorize banks’ risks and correlate them to one another in their likely response to various risks. Needless to say, this is a difficult feat even for wall street wizzes, and I sincerely doubt that the government has the competency to pull it off reliably. (On the other hand, perhaps even an unreliable attempt by the government is better than the status quo.)
How about this, then? Every bank, large and small, must purchase bailout insurance. The insurance pays out not in all cases of bankruptcy but only if the government actually decides to bail that particular firm out. This will create a countervailing pressure by the industry when the question of a bailout arises; peer firms will support a bailout only when the chaos caused by a bank failure would be more costly to Wall Street than the inevitable rise in insurance premiums after a bailout. And that, assuming the efficiency of insurance markets, is exactly the calculation we want to occur.
November 26th, 2008 at 2:15 pm
How would bailout insurance be any different than the credit derivatives that brought down AIG?
November 26th, 2008 at 3:41 pm
Matt,
from what I read, they’ve been doing the opposite, with fairly decent results for the smaller-fry banks. Since they don’t have large capitalization, they were forbidden from dealing in many of these riskier instruments. As a result, many small to mid-size banks are in pretty good shape today. The “crime” was freeing these large, behemoth Corporatist banks to indulge in these gigantic schemes and scams. I think good regulation appropriate to capitalization is what matters — plus limiting capitalization. Why not have cap sizes. That should provide for more competition and less danger to the system as a whole. The dinosaurs got too big too, and failed.
March 1st, 2009 at 5:46 am
viagra
Excellent site. It was pleasant to me.
March 22nd, 2009 at 6:17 am
tramadol
Very interesting site. Hope it will always be alive!
March 22nd, 2009 at 10:29 am
buy viagra online
If you have to do it, you might as well do it right
April 3rd, 2009 at 4:14 am
Great site. Good info
cheap brand pfizer viagra