By Steven Pearlstein
Wednesday, November 7, 2007

CHARLOTTESVILLE

Poor Chuck Prince. For the past five years, he thought his job was to clean up the ethical mess at Citigroup left by his longtime friend and patron Sandy Weill.

As Weill was taking his leave of the financial behemoth he had created, state and federal officials were moving to put the company into the equivalent of regulatory receivership. And the company was facing years of civil litigation by shareholders of Enron, WorldCom and who knows how many other bankrupt companies that claimed to have been misled by Citigroup analysts or investment bankers. As the directors pondered their choice, who better than Prince, a sober and trusted lawyer, to negotiate the settlements and make sure it would never happen again?

And what a mess it turned out to be! The subprime lending unit in Dallas that was cited for all manner of unsavory business practices. The bank's role in advising and financing Enron. The supposedly independent equity analysts who blew air into the tech and telecom bubble to curry favor with top management and help generate fees for the investment bankers. The private bank in Japan that misled customers about the risks of investments while helping other customers manipulate stock prices and hide trading losses.

As he moved from one scandal to another, Prince was forced to clean house, firing old friends and trusted allies who either knew, or should have known, about the misdeeds. To satisfy regulators and the demands of the new Sarbanes-Oxley law, he installed internal controls to make sure it didn't happen again. Employees were treated to a regular diet of ethics training.

What Prince never quite realized, however, was that Citigroup's problem wasn't that its people didn't know right from wrong.

No, the bigger and more encompassing problem was that the company and its top executives were never very good at assessing risk -- whether of cutting ethical corners and getting caught, or that certain kinds of loans would go bad, or that the market might turn this way or that.

This knack for misgauging risk dates at least to the 1980s, when the old Citibank got in trouble with its excessive lending in Latin America. It was true in the late '80s and early '90s when the bank got in too deep with commercial real estate lending and had to be rescued by a Saudi prince. It was true during the tech and telecom bubble of the late '90s. And it is still true today, as evidenced by Citigroup's headlong rush into mortgage-backed securities and off-books investment vehicles that have already resulted in write-offs of $8 billion to $10 billion.

Although it stands out among the major players, Citigroup is hardly the only organization that suffers from this inability to accurately assess risk. Merrill Lynch, the country's largest brokerage firm, and UBS, the largest bank in Switzerland, have also sacked chief executives in recent weeks for taking too many bad risks with shareholders' money.

"A lot of what we're seeing is that even very sophisticated people don't understand the risk very well," says Robert Carraway, who heads the MBA program at the Darden School of Business here at the University of Virginia. "It's not that they don't realize that things can go bad. What they don't realize is how bad they can get."

People aren't very good, Carraway says, at distinguishing patterns from random events. When fund managers, for example, deliver four or five years of fabulous returns, they invariably ascribe it to their superior skills rather than acknowledge it could simply be dumb luck.

On the other hand, when things start to go bad, the tendency on Wall Street is to discount it as a random event rather than the beginning of a self-reinforcing pattern in which everyone winds up selling or refusing to extend credit all at the same time.

In the case of Chuck Prince, it is probably unrealistic to think he would be the best judge of all the risks associated with exotic new financial instruments. But it is fair to expect a chief executive to notice when his trading desks or investment banks are earning ridiculous amounts of money for adding little value and taking what are advertised as low risks -- and to understand that those kinds of windfalls can't last.

What is also clear is that, for all his reforms, Prince never bothered to change the incentive structures that have always encouraged excessive risk-taking on Wall Street, whether financial or ethical.

Another young Darden professor, Jared Harris, has recently studied all the accounting restatements made by major corporations from 1997 to 2002 that were found to be the result of intentional misrepresentation. And what he found was that the propensity to cheat was highly correlated with executive compensation packages heavily weighted (75 percent or more) toward performance pay.

In a sense, that is the idea behind performance pay -- to give executives the incentive to take the risks that offer the big payoffs for investors, and then manage those risks. But what Harris found is that at a particular point, when performance pay tops 90 percent, the incentive becomes so strong that it encourages bad risks as well. And that would apply equally, he reasons, to financial risks as it does to the ethical ones he studied.

There is, of course, nowhere on earth where pay packages are larger or more tilted toward performance pay than on Wall Street, where it is not uncommon for top traders and unit managers to earn bonuses of $10 million or $20 million in a good year. The problem is that when those "good" bets suddenly turn "bad," there's no requirement for those outsized bonuses to be returned. And it is that heads-I-win, tails-you-lose structure that goes a long way in explaining why Citigroup, Merrill Lynch and UBS got in the trouble they are now in.

Steven Pearlstein can be reached atpearlsteins@washpost.com.