21st-Century Bank Run

Watching a $4 billion company fall apart in a week.

Bank run, 1931-style

If you want to know what a modern bank run looks like, consider the case of the giant commodity trading firm Refco. It went public in mid-August, but in the course of the past week it has gone from $4 billion stock-market darling to carcass. The proximate cause of the meltdown was the surprise disclosure on Monday, Oct. 10, that an entity controlled by CEO Phillip Bennett had owed $430 million to the company. A week later, trading of the stock has been halted and vultures are picking over Refco the way hyenas gnaw on the remains of wildebeest.

Refco was no boiler-room operation. It’s been around and successful for a long time. (Scandal connoisseurs will recall that Refco was Hillary Clinton’s commodities broker.) And it had been getting more and more respectable. First, Thomas H. Lee, the highly respected private equity investor, agreed to take a big stake in Refco in the summer of 2004. Then gold-plated underwriters Goldman Sachs and Credit Suisse First Boston brought it public two months ago.

Refco was a model 21st-century business—a highly digitized, high-tech services company that traded complicated financial instruments on behalf of customers all over the globe. But its meltdown shows that its real assets were not its New Economy algorithms and brainpower. Rather, this extremely modern company depended ultimately on the kind of assets that built American capitalism in the 19th century: trust, integrity, and the personal reputation of executives.

Nothing material changed in Refco’s financial situation when it announced that Bennett had secretly owed money to the company, or when it provided more details the next day about how Bennett had hid the debt. If anything, the company’s situation improved, since Bennett paid the money back and quit the same day. The company also took further proactive action, hiring Arthur Levitt, former chairman of the Securities and Exchange Commission, to help clean things up. Refco was solvent. It had tons of cash on hand. Nobody was worried that it wouldn’t be able to pay its rent, salaries, or utility bills.

But it was already too late. Refco was in the business of facilitating trades that are conducted essentially through a digital handshake. The actual exchange of cash—the settlement—takes place within a few hours or a few days. Any company operating in this environment relies on liquidity—the ability to access vast stores of credit instantaneously and cheaply—and on the willingness of other institutions to act as counterparties, to wait a day or two before receiving payment.

Once the trouble was announced, Refco’s customers wondered whether it was wise to do business with a company whose internal controls were so weak that it didn’t know its own CEO was hiding a nine-figure debt. So, the demise was swift. (Here’s the nasty five-day chart.) Within two days of the announcement of the discovered debt, Refco had to shut its nonregulated capital-markets subsidiary because it lost liquidity. In other words, people no longer trusted Refco to make good on trades. Customers began to yank funds, clients started to steer business elsewhere, and employees began furiously to look around for new gigs.

In abandoning Refco so rapidly, the market proved that creditworthiness is not an absolute attribute that can be proved by showing you have a certain amount of cash on hand, or that your equity-to-debt ratio is above a certain level. Rather, it’s relative. Companies may boast excellent credit ratings from agencies like Standard & Poor’s. But ultimately, creditworthiness is in the eye of the beholder. It’s something that people say you have, based on personal experience, reputation, and marketplace behavior.

And that’s how Refco found itself transported back 150 years. Dun & Bradstreet has been in the business of providing credit ratings since the 19th century. Its predecessor companies, including R.G. Dun & Co., employed correspondents in every major city and town who would send word to headquarters about the reliability of various businesspeople. Much of it was gossip, which is part of what makes Harvard’s collection of R.G. Dun & Co.’s massive leather ledgers such great reading. The correspondents may not have had access to merchants’ balance sheets, but they did know whether, say, a dry goods merchant in Albany, N.Y., had stiffed a supplier on a $10 bill, or which glass manufacturer in Brooklyn could be trusted for $1,000 of credit.

But Refco’s downfall isn’t simply an occasion for a history lesson, or an object lesson for people who make their living in the commodity pits. The entire global economy runs on the lubricant of easy credit extended among companies. And much of that credit depends on trust and reputation. An auto-parts company that gives its customers 30 days to pay it and has 30 days to pay its suppliers can function quite well. But if a few suppliers become worried that the company might have difficulty paying its bills, and demand to be paid in 10 days, that company could go bankrupt in a matter of days. These days, you don’t have to be a bank—or even a liquidity-dependent finance firm—to suffer a run on the bank.