Network News

X My Profile
View More Activity

Explaining FinReg: Shadow bank runs, or the problem behind the problem


On June 20, 2007, Ben Bernanke said that the subprime crisis “will not affect the economy overall.” That wasn't a one-time slip: He'd also said that "the impact on the broader economy and financial markets of the problems in the subprime markets seems likely to be contained,” and assured investors that "while rising delinquencies and foreclosures will continue to weigh heavily on the housing market this year, it will not cripple the U.S.”

Embarrassing, no? But Yale economist (and former AIG consultant) Gary Gorton is sympathetic to Bernanke's statements: Subprime shouldn't have been big enough to cause this sort of crisis. In 2005 and 2006, the market originated about $1.2 trillion in mortgages -- big, but not a vital organ of the American economy.

Subprime was the trigger for the crisis, but not the cause. What happened, rather, was that the subprime crisis set off an old-fashioned bank run in a newfangled market: the shadow banking market, which was, and is, vulnerable to runs.

Understanding this market, Gorton says, is crucial to understanding the last crisis and stopping the next one. The shadow banking market is where big banks, institutional investors, and other folks who have a lot of money do their banking -- particularly their short-term banking. So let's say I'm Ezra Bank. I've got a $100 million that I'm going to invest next month, but for now, I need to put it somewhere. I head to the "repo market," and I ask Bear Stearns to hold my money and pay me interest. They agree. But how do I know Bear Stearns won't just keep my money?

Individual depositors in the normal banking market never have that fear. The government insures our deposits. But they don't insure massive institutional deposits. So Ezra Bank would ask Bear Stearns for "collateral" -- something that is low-risk and valuable they could hold in order to make sure Bear Stearns returned their money. Something like, say, AAA mortgage-backed securities.

This manner of banking created a massive hunger for collateral. And it was this hunger, Gorton thinks, that drove the wild demand for mortgage-backed securities.

But think about the difference between the shadow banking market and your bank: The FDIC's deposit insurance exists to prevent bank runs (which happen when creditors become scared that their bank is insolvent and rush to get their money back, which in turn makes their bank insolvent). The shadow banking market doesn't have deposit insurance. So how does it deal with the problem of bank runs?

Answer: It doesn't. What we had in 2008, Gorton says, was a bank run. No one knew which banks were exposed to the subprime crisis, so everyone froze. But it didn't need to be the subprime market that experienced the shock. A lot of different types of shocks would've done the trick. The underlying problem is that the collateral is "informationally sensitive": That is to say, information can dramatically and unexpectedly change its worth (i.e.. news that subprime mortgages are defaulting makes securities based on subprime mortgages worthless), and then confidence drains out of the whole system. "It's the e coli problem," Gorton says. "When they recall 10 million pounds of burger, it brings all sales of ground meat to a halt because no one knows how much e coli there is or where it is."

This next graph comes from Gorton's paper “Slapped by the Invisible Hand” (pdf). It's a graph of repo market haircuts, which is to say, how much it cost for major investors to lend to each other.


"The figure," Gorton writes, "is a picture of the panic." That is to say, it is a picture of the bank run in the shadow banking market.

Conversely, our deposits with our banks are not informationally sensitive: Where small pieces of new information can scare the shadow-banking market, major revelations are shrugged off in the commercial banking market. News that Wachovia was going bankrupt, for instance, didn't matter because the federal government insures deposits.

To offer an analogy, consider someone with a weakened immune system who eats a bad piece of fish and gets really sick. Obviously, the first thing you want to do is deal with the illness. But when that's over, the issue you want to deal with isn't so much sushi selection (what made the patient sick this time) as the weakened immune system (what makes the patient vulnerable to dangerous illnesses). Gorton, however, thinks that we're talking about sushi rather than systems. Putting derivatives on exchanges and clearinghouses will do a lot to make sure that the system doesn't get the same illness anytime soon, but it doesn't deal with the system's vulnerability to illnesses -- that is to say, the system's vulnerability to bank runs.

Handling that would require either creating a type of safe, informationally-insensitive collateral for the shadow-banking system to use or examining and insuring the collateral the system does use. But that's a big intervention, and not one that most people seem interested in contemplating right now. Ignoring it, however, leaves the system vulnerable to shocks, even if we've done a fair amount to protect against the specific type of shock that comprised the subprime crisis.

That scares Gorton. Bank runs are not generally an isolated phenomena: As you can see in the graph atop this post, they come in waves, continuing until the country does something that prevents runs -- like deposit insurance -- rather than just preventing whatever caused the most recent run.

(For the rest of the 'Explaining FinReg' series, head here.)

By Ezra Klein  |  April 26, 2010; 3:06 PM ET
Categories:  Explaining financial regulation  
Save & Share:  Send E-mail   Facebook   Twitter   Digg   Yahoo Buzz   StumbleUpon   Technorati   Google Buzz   Previous: The big winner from immigration reform
Next: Does the Senate need collateral?


What is an example of the type of collateral a firm would use to "insure" deposits?

Posted by: rsmith34 | April 26, 2010 3:30 PM | Report abuse

Ezra writes
"[Repairing] the system's vulnerability to bank runs... would require either creating a type of safe, informationally-insensitive collateral for the shadow-banking system to use or examining and insuring the collateral the system does use. But that's a big intervention, and not one that most people seem interested in contemplating right now."

But isn't that what the $50 billion 'unwinding' fund (I prefer 'dismemberment') would do? i.e. rather than insuring the depositors/creditors or verifying the collateral, the bill would create a reserve fund that shuts the shadow bank down in an orderly fashion. Instead of directly following the FDIC example of making depositors whole, it would pay creditors cents on the dollar - an incentive that might cause them to rethink the shadow bank in the first place, yes?

Posted by: bsimon1 | April 26, 2010 3:51 PM | Report abuse

Mortgage-backed securities. You hand over your CDO full of subprime mortgages while the other bank give you their cash. That's why people freaked out so much when they suddenly thought those CDOs might be worthless.

Posted by: Ezra Klein | April 26, 2010 3:52 PM | Report abuse

Isn't this precisely what the Resolution Authority in the proposed FinReg bills is supposed to deal with?

It covers any systemically important firm, and it allows the government to step in, pay the creditors (I'm sure with haircuts involved), and then wipe everyone else out. This is the same way FDIC works with regular banks.

Am I wrong to believe that's how this works, or is there nothing in the bill to address this issue?

Posted by: cbaratta | April 26, 2010 4:00 PM | Report abuse

The haircut chart is used by Jim Hamilton as part of his "financial crisis in a nutshell" post:

Posted by: bdballard | April 26, 2010 4:11 PM | Report abuse

Resolution Authority doesn't have a direct impact here, although it could help a little.

The reserve fund is to momentarily extend the life of the shadow bank to allow for an orderly sell off, but it doesn't change the key informational problem regarding the value of complex derivatives. Deposit insurance protects against bank runs because you know that your money's safe. The money in a failed shadow bank being operated under resolution authority is not safe, it's at the whims of the market. Moreover, in the shadow bank example, you find out all of a sudden that an entire multi-trillion dollar class of assets is tainted, so there's no way to know the extent of your risk. Basically, the difference is security up to $100,000 (FDIC) vs. "some percentage of your value that's better than it would be if the bank had a fire sale."

Posted by: etdean1 | April 26, 2010 4:11 PM | Report abuse

rsmith34 -- some treasuries would do just fine as collateral -- the Treasury repo market has been pretty big for a long time. Less scary than MBSs too, so they wouldn't require the haircut

Posted by: bdballard | April 26, 2010 4:15 PM | Report abuse

I guess I don't understand the incentives for a bank to "deposit" their money in such a risky security especially in the short-term. Bare with me, I'm fairly unfamiliar with finance as an undergrad in Political Science.

Posted by: rsmith34 | April 26, 2010 4:47 PM | Report abuse

First I read David Frum's article where he states that the crisis was because of major consumer debt and China and it's wheeling and dealing with it's currency. Then, I hear on TV that it's because of bets made by the major financial institutions and now you write that the major cause was bank runs in the shadow banking industry. Which is it? Can somebody tell me the core issue behind this latest crisis?!?

Posted by: yahnik | April 26, 2010 4:47 PM | Report abuse

Interesting points by Gorton. And while this certainly has implications for financial stability, I think it's wrong to characterize $1.2 trillion in subprime as causing the economic crisis. As Dean Baker has pointed out repeatedly, there was an $8 trillion housing bubble. When it burst, it was certain to plunge the economy into recession. This is what Bernanke (and most economists) missed. The financial collapse was on top of that. Even if there wasn't a shadow bank run, we'd still be at 10% unemployment and have lost tremendous output due to the demand in the economy lost due to the collapse of the bubble.

Recall demand was artificially high during the bubble years due to soaring home prices. People used their rising home values as ATMs to fuel consumption. With home prices no longer rising, the consumption was bound to come to an end in a nasty recession, proportional to the size of the bubble. This would have happened with or without the shadow bank run by institutional investors.

Posted by: enormousturnip | April 26, 2010 6:58 PM | Report abuse

Banks SHOULD be allowed to fail. It's what prevents the money supply from expanding indefinitely in a fractional-reserve banking system. What the Fed has been doing is swapping out Bank's risky assets and turning them into marketable liquid securities, cash or M1. They’ll eventually call back these securities once the economy recovers, but can’t fully, since it would cause crippling interest rates. What the economy and taxpayers are left with is a much higher price level and lower standard of living, due to the inability to let bad investment fail, because of the large size of these banks.

Think what would happen to a mutual fund,for instance, if the original purchase price of an investment were guaranteed. The fund managers would essentially have a put-option against those securities, guaranteed not to lose money. It’s not hard to imagine that the fund managers would begin to take riskier and riskier investments, even creating instruments(derivatives) which allow for a virtally unlimited amount of risk and return. When these securities do go below the original purchase price( a loss), whoever guaranteed that investment takes the loss. When they go up, of course, the mutual fund pockets that, however.
In the case of the banks, the federal reserve and treasury has essentially guaranteed these investments. The fed and treasury don’t really take a loss, they simply create new money to cover the losses. What does happen is that the cost of living increases dramatically.

The only way out is to allow banks to take these losses. This only happens if banks are small enough to minimize cascading effects to the entire economy.

Posted by: donameche | April 26, 2010 7:15 PM | Report abuse

I certainly don't know enough about this to mount a firm counter-argument to Gorton or to you, Ezra, but wouldn't "creating a type of safe, informationally-insensitive collateral for the shadow-banking system to use" create a massive amount of moral hazard? Deposit insurance works for consumer banks because those banks are/were limited to only certain kinds of investments and are held to strict capital requirements; absent such regulation, didn't the shadow banking system use CDOs and CDSs based on the 1.2 trillion in home mortgages to leverage themselves far beyond that amount? And didn't we supposedly have S&P and Moody's "examining" and AIG "insuring the collateral the system" used?

Yes, when the worth of those securities was called into question we had the run on the shadow banking system, but that the securities were in fact worthless seems to me to be the real "problem behind the problem." What the shadow banking system was doing was chopping up bad fish to make what they were calling premium sushi. Then they were taking the sushi they couldn't sell that day, chopping it up for the next and still calling it premium fish. So let's assume we come up with some kind of diner's insurance, presumably paid for by a tax on diners and restaurants; the customer gets his money back, but what does that insurance do to prevent the selling of bad sushi? Why not chop up even more and worse fish and old sushi to leverage even further? I could be very wrong, but it doesn't seem to me that insurance in this case would make the sushi any better or kept the restaurant's customers from getting any less sick.

Posted by: KevinReeves | April 26, 2010 10:53 PM | Report abuse

Three thoughts/questions. First, do banks really need to earn interest on short term funds pending other investments? Gorton's graph suggests that the haircut (interest rate) for the initial six months covered by the graph was zero. So for that period of time (which perhaps represents normal times) the banks seemed content with no return. Second, if prudent asset management dictates that a return is necessary, why can't the banks purchase short-term Treasury obligations. Third, if a bank really wants to deposit its short-term funds with another institution, why not limit collateral to Treasury obligations owned by the deposit-taking bank?

Posted by: pfeiferharry | April 27, 2010 4:19 AM | Report abuse

The comments to this entry are closed.

RSS Feed
Subscribe to The Post

© 2010 The Washington Post Company