Planet Money
1:45 pm
Tue November 20, 2012

Ask A Banker: Are The Banks Still Too Big To Fail?

Hi! I'm back again. I'm a former banker, now a writer at Dealbreaker and an answerer of real and imagined questions about the financial world here. You can send questions to with "ask a banker" in the subject line, or ask on Twitter (@planetmoney). The last couple of times we've talked, we've discussed things that scare people about the stability of the financial world, like derivatives and high frequency trading. Today's question is another one of those. BE AFRAID! Next week we'll discuss puppy ownership among investment bankers or something. Anyway:

Q. Do you think that the banks are still too big to fail?

Of course! Come on, man! Next question.

Q. What exactly does it mean to be too big to fail, anyway?

So: Banks are basically in the business of owing money to people.* All the money in your checking account is money the bank owes you. And banks' trading businesses are mostly conducted with borrowed money.**

Banks borrow all this money and invest it in things — mortgage loans, Treasury bonds, derivatives, whatever. Sometimes those investments lose value. If they lose enough value, then either (1) some of the people who loaned money to the bank don't get their money back, or (2) somebody outside the bank comes up with money to give to the people who loaned money to the bank.

In the U.S., banks where you can open an ordinary bank account are all insured through a federal program. Banks pay for this insurance. When a bank fails, whether it's big or small, people with ordinary bank accounts will get their money back.

Really big banks have that insurance too, but they also have something else. Or, people think they do. People have an expectation that, if a really big bank blows up, the government will step in to repay lots of people who loaned the bank money — not just people with ordinary bank accounts but bondholders and derivatives counterparties.

This is not an insurance policy that everyone agrees on in advance and the bank pays for; it's just an implicit guarantee, a sense everyone has that this would happen, because it would be too much of a mess if it didn't. If you're a giant bank, "too big to fail" is like a random blessing: You blow up, and poof, you've got insurance that covers all this stuff you didn't buy insurance for.

"Too big to fail" also means that the government will endeavor to keep you open for business: it won't just pay off your debts and sell you off for spare parts.*** The government isn't just interested in protecting your creditors from losing money; it's interested in keeping you around doing what you're doing, because you're doing something important. JPMorgan, for instance, is one of two banks that runs the plumbing of the triparty repo system; if JPMorgan went away, nobody would be triparty repo'ing any more, and you'll just have to trust me that That Would Be Bad. (But it would.)

Q. So why would a bank want to be too big to fail?

Remember: banks are in the business of owing people money. Being too big to fail makes you better at owing people money: if you somehow lose all their money, the government will step in and give you money to give back to them.

That's good for the people you owe money to. They sleep more soundly — and lend you money more cheaply — knowing (guessing!) that (1) you'll probably give them their money back and (2) if you don't, someone else probably will. Customers and counterparties and lenders and bondholders, if they're smart, want their bank to be too big to fail.

This allows too-big-to-fail banks to pay lower interest rates on the money they borrow than other banks, which can make them more profitable. That lets them make more money for their shareholders — who should also love too-big-to-fail banks — and for themselves. Bigger bonuses for everyone!

Q. So too big to fail is good, then? Everybody seems to win?

Oh, sure. Everybody but the people who'll be bailing out the banks when things go wrong. I guess those people are called "you," in some loose taxpayery way.

Q. That sounds less nice. Just how much money am I giving these banks anyway?

No one really knows. The too big to fail subsidy lives entirely in a world of uncertainty. It's not even clear which banks are too big to fail. A lot of people were genuinely surprised that Lehman Brothers, for instance, was allowed to fail, and I guess some were surprised that AIG wasn't.

There are some predictions about who is too big to fail and who isn't. Moody's, the credit rating agency, for instance, adjusts its credit ratings of the big banks to a level "reflecting Moody's assumptions about a very high likelihood of support from the US government for bondholders or other creditors in the event such support was required to prevent a default." But those are just assumptions.****

There are various clever ways to try to estimate the size of the "too-big-to-fail subsidy." The Cleveland Fed rounds up some research here, and guesstimates that the number is about $45 billion a year. So the average American is paying about $143 a year to subsidize JPMorgan and its hefty friends. Not counting overdraft fees.

But that really is just a guess.*****

Q. So is anyone doing anything about too big to fail?

Of course. Here's a nice speech from the President of the New York Fed last week, laying out some of what the government has done to reduce the likelihood that banks will be too big to fail, and also the likelihood that they'll fail. (Quite different things!)

One such effort is a thing called "G-SIFI capital surcharges." This forces banks that are really big and interconnected (the ones you might guess are too big to fail) to raise additional capital. "Capital" is a slippery beast to wrestle with but you could say that more capital makes a bank (1) safer and (2) less profitable. This has two advantages: one, it makes too-big-to-fail banks safer, and two, it makes it less attractive to be a too-big-to-fail bank, since you pay for that status with extra capital.

You can tell that this is a pretty good idea by the fact that the big banks are complaining about it. You can tell that it's not <i>that</i> good an idea by the fact that no one's actually done anything to stop it: if those surcharges actually made it way less profitable to be a too-big-to-fail bank, some of those big banks would be getting smaller so as to be more profitable. The fact that that doesn't really happen****** suggests that too-big-to-fail is still pretty attractive.

Q. Why not just break up the banks?

Sure. I mean, fine. Remember though that bigness makes banks better at their job of owing people money. That's an important job. If the banks got worse at it, some people would be sad. What would they do about it? Maybe just get by with people owing them money worse, like you do with Apple Maps.

Or maybe they'd turn to shadow banking, which is a popular way of having money owed to you without going through banks. That's been on the rise in recent years, in part because banks are not as risk-free as (people thought) they used to be.

Q. Oh. Should we worry about shadow banking?

Sure, probably. Puppies!

* Or being owed money by them but whatever. Banks are also in the business of being owed money by people but that is sort of less interesting. All businesses owe money and are owed money; the weird thing about banks is how much money they owe, not how much they're owed. Take JPMorgan; it has give or take $200bn of stockholders' equity on $2.3trn of assets, meaning that for every dollar JPMorgan has, it owes someone 91 cents.

** Pretty much the only banking business that isn't about owing people money is traditional investment banking - advising on mergers and underwriting securities deals. You can have good or bad or mixed feelings about this but for the most part it isn't exactly scary.

*** Think of Hostess, the snack food company, which has been bankrupt but running its business for months, but which recently filed to liquidate and stop running its business. Twinkies are apparently not too big to fail.

**** And they're hard to make in a vacuum. You could just about believe that if Bank of America, say, came up with some idiosyncratically moronic way to lose $100 billion, the government would let it fail. (JPMorgan maybe not so much.) But the realistic way for Bank of America to lose $100 billion would be the sort of economic disaster scenario where every bank is losing that kind of money. And <b>all the banks combined</b> will always be too big to fail.

***** Similarly you can take a guess about how much the last round of bank bailouts cost taxpayers. That's just an accounting exercise - there's no uncertainty! they already happened! - and yet is bafflingly difficult.

****** One possible counterexample is Jefferies, a small investment bank that has studiously avoided going over the $50 billion asset threshold that puts you at risk of that "G-SIFI" designation, possibly to avoid the capital surcharge. This is not much of a counterexample, though, as no one was really going to think that Jefferies was too big to fail.

Copyright 2012 National Public Radio. To see more, visit