The Price of Too Big to Fail

Academics figure out the cost of the government’s implicit backing of big banks
Illustration by Erik T. Johnson

In June, Jamie Dimon told a Senate committee that no taxpayer money was “impacted” by this spring’s trading losses at JPMorgan Chase. Dimon, the bank’s chief executive, meant that no one at the Department of the Treasury had to write a check to save the bank. He’s right, and may continue to be right, even after the revelation that the bank’s trading losses might run as high as $9 billion. But checks are not the only thing of value the government can offer a bank.

On June 28, Richard Fisher, president of the Dallas Federal Reserve, said that the markets assume larger banks are too big to let fail. That much we knew. Five banks—JPMorgan, Bank of America, Citigroup, Wells Fargo, and Goldman Sachs—held more than $8.5 trillion in assets at the end of 2011, equal to 56 percent of the U.S. economy, according to the Federal Reserve.

Graphic by Bloomberg Businessweek

Fisher also pointed out that this assumption lowers borrowing costs for those banks, which he called an “unfair subsidy.” He didn’t name names, but we can. Even better, we can give you an idea of the size of this subsidy. By one estimate, between 2007 and 2010, simply being too big to fail saved America’s biggest banks a combined $120 billion in lower borrowing costs. Citigroup saved some $50 billion. And even with its fortress balance sheet, JPMorgan saved $10 billion thanks to its size and importance.

The estimate comes from a paper Frederic Schweikhard and Zoe Tsesmelidakis presented in late May to a New York University Stern School of Business conference on credit risk. The pair, both Ph.D. candidates at Goethe University Frankfurt, relied on a well-tested tool of finance—the Merton model of pricing corporate debt—to come up with their estimate. Neither the banks nor Treasury would comment.

The Merton model is tricky to understand. Luckily Robert Merton, a Nobel laureate now at Massachusetts Institute of Technology’s Sloan School of Management, answered the phone to help explain it. “Pick up two pieces of paper,” he said, “are you doing it?” Indeed. “Now,” he continued, “label one ‘corporate bond’ and the other ‘full faith and credit of the United States Treasury.’” Put them together and it looks like two slips of paper stacked together. “That’s risk-free debt,” he said, “you’re gonna get paid no matter what.”

The price of every loan, he explained, consists of two pieces: the inconvenience of parting with your money for as long as the borrower needs it, and the risk that the borrower will default before repaying you—which, in the case of the big banks, is eliminated by the Treasury’s implicit guarantee. “Those are different activities. One is the time value of money and the other is insurance,” he said.

In 1974, Merton published a model to price the guarantee part of corporate debt. The model treats a loan like an insurance policy on the borrowers’ assets. The lender, Merton reasoned, gets extra money for the guarantee from the borrower—like an insurance premium. If a catastrophe happens, the lender acts like an insurer paying a death benefit: He loses some or all of his principal. (If you’re a financier, you also know this as a put option, an insight that Merton used to do his math. But if you are a financier, you can also skip ahead.) Merton’s model wasn’t widely used until a rash of corporate defaults in the late 1990s. To have a theoretical method adopted for practical use, says Merton, “I would say the three most important factors are need, need, and need.” Investors suddenly needed to understand how to protect themselves from these defaults.

Schweikhard and Tsesmelidakis have applied Merton’s insight to the risk of bank default. A credit-default swap (CDS), they reasoned, is also an insurance policy. Their paper takes stock prices of financial firms from 2002 to 2010 and uses Merton’s model to derive the theoretical prices of credit-default swaps on those companies—excluding any effect of the government’s backing—for a fictional swap holder. They then compare that price to actual market prices, and discover a varying difference between the two, which they call “the wedge.”

The wedge shows that, in times of crisis, actual CDS prices are lower than theoretical CDS prices derived from the bank’s stock values. Stockholders of large banks know that after a default, the values of their holdings will decline no matter what. But debt holders of those same companies know that a bailout might protect their assets. And they seem to expect one. For American financial firms the wedge—the difference in confidence between stockholders and debt holders—begins to spread in May 2007. Right after the collapse of Lehman Brothers, it briefly disappears. For a moment, no one expects a bailout. Then the wedge yawns again, undisturbed by the passage of the Dodd-Frank Act, which promises an end to bailouts. The wedge does not close again until the relative stability of fall 2011 and spring 2012. And it begins to widen again in early May of this year, with more bad news from Greece.

Dodd-Frank might make a pledge of no more bailouts, but the market seems to know that you can wake up a Congressman in the middle of the night, tell him the world is about to end, and expect him to help. And this knowledge is extremely valuable, even to a company like JPMorgan, which never needed a bailout. If you accept Merton’s logic and Schweikhard-Tsesmelidakis’s data, since 2007 the United States has given away an insurance policy that would have cost JPMorgan dearly on the market. And with that free insurance policy, the bank saved a lot of money.

Schweikhard and Tsesmelidakis look at bank bond issues between 2007 and 2010 and estimate the amount each bank saved in lower financing costs, given the market’s assumption of a credit guarantee from the United States government. This is how they arrive at their savings tallies, bank by bank—again, a combined $120 billion. And this was only through 2010.

Even if one of these fortunate banks doesn’t need a government bailout, the arrangement does cost taxpayers. The implicit guarantee makes government bonds ever so slightly riskier and therefore raises borrowing costs. Sometimes, of course, big banks get punished: London-based Barclays was just fined 290 billion pounds ($455 million) for attempting to rig the London interbank offered rate.

But a government’s implicit guarantee can be worth more than any fine. Jamie Dimon is right. No one wrote him a check, either during the financial crisis or after this spring’s trading losses. Somehow taxpayers managed to give him $10 billion anyway.