A study in CEO greed
How one intrepid academic exposed the latest stock option scandal.
by Geoffrey Colvin, FORTUNE Magazine

(FORTUNE Magazine) - One of the questions crying loudest from the developing scandal of backdating stock options is, Why now? As new companies announce daily that they're investigating their stock option practices or that they've received inquiries from the SEC or federal prosecutors - United Health Group, Caremark Rx (Research), and Juniper Networks (Research) are among those that have acknowledged ongoing probes - you have to wonder what sparked the mess. After all, we're talking about pre-2002 behavior, mostly from the '90s. The answer leads me to say, Three cheers for the digital revolution and for SEC rules on executive-pay disclosure. Without them, we might not know how some executives were robbing their shareholders - and it might still be going on.

The problem, in brief, is that executives at some companies were backdating their stock options to dates when the stock was at its low for the year or the quarter, tilting the odds of profiting on those options heavily in their favor. Backdating isn't necessarily illegal, but following the complicated rules would largely eliminate the advantages of doing it. As actually practiced, it was stealing, pure and simple. Executives looked back over several months and chose the date on which they wanted the right to buy shares of the company they were being paid to manage and - surprise! - they chose the date when the price was lowest. Wouldn't you like to decide in retrospect when to buy a stock?

We might never have heard about this slimy behavior if a researcher at the University of Iowa, Erik Lie, hadn't decided to study the behavior of stock prices before and after option grants. He wasn't the first. Other academic researchers had studied the phenomenon and found suspicious results. But none of the researchers had suggested that executives might be doing anything illicit or illegal.

Then Lie conducted the mother of all stock option studies, looking at 5,977 option grants between 1992 and 2002. In his paper, published a year ago, he found the same suspicious results as earlier researchers, only more pronounced. Further slicing and dicing the data, he discovered that unless executives possessed truly extraordinary abilities to forecast precise overall market movements, they had to be backdating the grants.

This new hypothesis was a bombshell. The SEC had previously investigated a few individual cases of backdating, but Lie's study was the first evidence that the phenomenon was widespread. He then supplied the Wall Street Journal with data identifying six companies bearing strong evidence of backdating, leading to a March article that triggered today's mushrooming scandal.

Lie's study never would have happened before the Internet Age. Gathering thousands of SEC documents, gleaning the relevant data, crunching literally millions of numbers - it was possible in theory, but it would have been impossibly expensive. Lie relied on the Standard & Poor's ExecuComp database, taken from the SEC's online documents, and a vast online trove of market data created by the University of Chicago.

But even the Internet wouldn't have enabled the uncovering of this scandal if the SEC hadn't imposed a rule requiring companies to report executive stock options in detail starting in 1992. That's why Lie's research (and virtually all other options research) begins with that year. If not for that rule, executives could still be scamming shareholders by backdating options, and we wouldn't know a thing about it.

Of course, even after the rule, some executives got away with it for years. Reason: They could legally delay reporting option grants for so long that it was virtually impossible to figure out whether any individual grant had been backdated. That's why Lie had to analyze thousands of grants and derive his conclusions statistically. Then came Sarbanes-Oxley, which required that option grants be reported within two business days. A new paper by Lie and Randall Heron of Indiana University, still unpublished, finds that evidence of backdating virtually disappears after Aug. 29, 2002, when the requirement took effect.

Were the SEC requirements burdensome government regulations? No way. Though always highly skeptical of new regulatory requirements, I have no problems with these. They tell owners what the hired hands are up to. And it's clear, as if we needed reminding, that a few of them will always be up to whatever they can get away with. The best way to control them is to make them tell their owners what they're doing - and then make that information available widely, immediately, and freely. Top of page