Chapter 20. Bank Runs

“With this one piece of legislation, the fear which operated so efficiently to transmit weakness was dissolved. As a result the grievous defect of the old system, by which failure begot failure, was cured.”¹

J.K. Galbraith, on deposit insurance

Market financing of banks makes them vulnerable to “bank runs”—when depositors lose confidence in them and pull out their funds. The run on the UK’s Northern Rock showed that bank runs could still happen. The runs on Wall Street’s “shadow banks” and then on Bear Stearns were more deadly because they had investments throughout the global financial system—a run on a single bank could burst bubbles across the world.

On the night of September 13, 2007, Robert Peston, the BBC’s business correspondent,² went on national television news to announce that Northern Rock, a large mortgage lender, was receiving “emergency financial support” from the Bank of England. It was a big scoop, as the deal to help Northern Rock was not meant to be public until the following morning. Peston was careful not to exaggerate. In his second sentence, he said, “this does not mean that the bank is in danger of going bust,” and he even added that there was “no reason for people with Northern Rock savings accounts to panic.”³

The next morning, people with Northern Rock savings accounts panicked. The bank’s website went down under the weight of enquiries. Queues started to form around Northern Rock branches. As the bank emphasized online accounts, it had relatively few branches, so spectacularly long queues formed at each one. They did this even though there was generous deposit insurance. The government guaranteed the first £2,000 in every account and 90 percent up to £30,000. Only amounts in excess of this amount, rather higher than most people should keep in bank deposits, were at risk.

The incident shattered confidence in the UK’s financial regulation and in the Labour government. It demonstrated herd dynamics in action: After confidence dwindles, in a crowd setting, it disappears. The UK responded by raising the insurance on Northern Rock accounts without limit, meaning that its depositors could treat themselves as having lent to the UK government. It was then taken over by the government and when no buyer came forward, it was nationalized.

This was the biggest bank run in a developed country since the Depression. Deposit insurance, by removing depositors’ fear that they could lose their savings, had made runs a thing of the past—just as extra deposit insurance ended the run on the Rock. It did not stop banks from making stupid loans, but it did protect them from suddenly losing deposits—at least until Northern Rock—and this made the system more secure. In return, the premiums they paid and the extra scrutiny they received tended to limit growth. In the 1930s, and again now, this seems a good trade-off, but the concept had detractors.

Before the Depression, bank runs and failures were common. In theory, this is because depositors’ ability to take their money out is a kind of monitoring system. A run is a brutal form of market discipline—and markets are capable of unsentimental judgments in a way that human regulators never are. Deposit insurance was a kind of moral hazard. Thus bankers were against it in the 1930s, with one senior executive complaining that “the competence of bankers is not an insurable risk.”⁴

Market discipline’s problem is that it can be unfair and irrational. Regulators can avoid the irrational panic that followed Peston’s report on the BBC. Because markets involve group behavior, they cannot. Further, bank runs can mete out their justice on healthy banks. Banks rely on short-term funding from deposits while lending over longer terms. This is an inherently unstable model, so it is always possible for a bank to become illiquid—to run out of the ready cash provided to it by depositors—without necessarily being insolvent (which means that its assets are worth less than its liabilities, so it cannot settle all its outstanding loans and debts). Thus relying on market discipline is dangerous. While in theory it is appealing to require investors to do all the due diligence on savings accounts, experience suggests that this will not work. In practice, deposit insurance is one state intervention worth making.

The problem is that commercial banks found ways around deposit insurance while investment banks, which were never covered by deposit insurance in the first place, voluntarily chose to finance themselves in a way that made them vulnerable to a run. A critical example was the structured investment vehicle (SIV) that started as a ruse to help banks generate higher returns for their clients while staying within the regulatory rules. SIVs, like mini-banks, lend long-term by buying mortgage-backed securities and borrow in the short-term by issuing commercial paper backed by those mortgage-backed securities. The profits flow back to the bank, but the risks do not appear on the balance sheet.

As trust collapsed in the money markets, SIVs lost the ability to borrow, giving regulators a dilemma. Citigroup, which used the model more than any other bank, at one point ran SIVs worth $80 billion.⁵ As with Northern Rock, its creditors (in this case buyers of commercial paper rather than depositors), went on strike. The logical next step, if the SIVs’ creditors refused to lend them more money, was to embark on “fire sales” of their mortgage-backed bonds and force the market lower.

In Europe, banks started to bail out their SIVs and put them on their balance sheets. In the United States, the Treasury tried to organize a “Super SIV,” which would have raised loans from the commercial paper market and then bought up securities from other distressed SIVs. But the bank run carried on apace. Many believed that the Super SIV was merely a rescue operation to spare Citigroup from taking losses and the fund never got off the ground.⁶

It was this invisible bank run that brought world markets down from their all-time high. Both the MSCI World and Emerging Markets Indices set their records on the ominous date of October 31, 2007—Halloween. The next morning brought a report by Meredith Whitney, then an analyst for CIBC World Markets in New York, that Citi would need to raise $30 billion in new capital to cover the costs of the assets it would have to write off.⁷ That implied issuing new shares (and diluting the share price) and cutting the dividend.

Citi’s shares dropped 7 percent the next day. UBS, the biggest Swiss bank, fell more than 5 percent on similar concerns, while the S&P 500 financial index, covering the shares of big banks, dropped 4.6 percent, its worst fall in four years on the worst day’s trading since the Shanghai Surprise eight months earlier.

How could worries about Citi’s SIVs end the rally in emerging markets? If there were a run on Citigroup, the world’s most interconnected bank, the logic went, then there would soon be fewer funds available for emerging markets—and the implications for the U.S. economy itself, the world’s buyer of last resort, looked ugly. Moreover, many investors had to meet claims from their creditors, and that meant selling the emerging markets while they still showed a profit. Interconnectivity spread the crisis across the globe.

Four months later, another bank run took the crisis to a new level of severity. Bear Stearns was the fifth largest investment bank on Wall Street. Aggressively reliant on short-term funding through the money markets, it had not raised any extra capital in the autumn of 2007. Now it was too late. The essential problem, yet again, was that Bear was holding assets of dubious value. The market lost its confidence in them, so investors in the money markets refused to lend Bear the short-term funds it needed to keep operating. Meanwhile, many hedge funds used Bear Stearns as their “prime broker,” meaning it was the bank where they deposited their funds. In the event of a bankruptcy, their funds might be stuck there, so they pulled their money out.

There was an added element to this bank run. Many bought protection against a Bear Stearns default, using credit default swaps. The more investors who bought credit default swaps, the higher the cost of protection would be. This raised the price at which Bear Stearns could borrow, which was bound to be higher if the insurance cost more. But this was not just about protection. To buy a Bear Stearns credit default swap was to place a bet that it would go bust. There was no need to hold any Bear Stearns debt. Instead, it was like buying an insurance policy on a house you do not own—you were not at risk in the first place, but you profit if it burns down.

Hence, traders bought protection as a form of speculative attack. As they piled in, they made it harder for Bear to raise new money. That damaged its share price. Speculators then sold short the stock, making money both as the stock went down and the cost of insurance rose. A falling share price made it more expensive for Bear to raise funds by issuing new equity.

By the night of Thursday, March 13, Bear had only $2 billion in cash left, compared with the $18 billion with which it had started the week, and told its regulators that it would file for bankruptcy the next morning. Instead, in a flurry of action, the politicians arranged a deal that forced Bear Stearns to sell to JP Morgan, a much stronger bank, at a price of $10 per share. Barely a year earlier they had traded for $170 per share. $30 billion in guarantees from the government persuaded JP Morgan to do the deal.⁸

Bear Stearns was not a commercial bank, but the way it funded itself made it prone to an old-fashioned bank run, made more aggressive by short-selling and default swaps. That run removed all control from Bear’s managers. Because of Bear’s roles lending to hedge funds, holding their accounts, and standing as a counterparty in many credit default swaps, a disorderly demise could have turned into a run on many world markets. Rather than being too big to be allowed to fail, it was too interconnected to fail.

Rescuing Bear, however, created its own problems. At the worst point of the Bear crisis, the S&P 500 was down more than 20 percent from its peak—a “bear market” in more ways than one. But the hope took hold that Bear had marked the end of the crisis. The response was to buy oil in a bet that contained the seeds of its own destruction.

In Summary

• Deposit insurance almost eliminated bank runs, but money markets found ways around it.

• Investment banks chose to finance themselves in a way—through the repo market—that made them vulnerable to such runs.

• These banks became so interconnected that a run on them could crash numerous other markets—and the run on Bear Stearns showed the dangers.