Commentary: Let's Really Clean Up Those Numbers--Now

By John A. Byrne

Little noticed in the shock of WorldCom Inc.'s massive accounting fraud is that it came to light only after a review initiated by the company's audit committee. Yes, believe it or not, the directors in the boardroom actually asked their auditors to take another look at the numbers.

Lo and behold, that review led to the largest single restatement ever: a $3.8 billion reduction in previously reported pretax income. Together with Xerox Corp.'s (XRX ) recent concession that it overstated operating earnings by $1.4 billion, these two corporate mea culpas alone nearly match the combined total hit of $5.8 billion from all 463 restatements in 1998, 1999, and 2000.

How much worse can it get? Plenty. It's not that we'll likely see numbers as breathtaking as those at WorldCom or Xerox. But expect a tidal wave of restatements to come in what will surely amount to some sort of accounting catharsis. If ever there was a time for companies to come clean, it's now. Over the years, the quality of financial reporting has varied widely. Some companies did the bare minimum to meet accounting standards. Others were more cautious. To get beyond the current stock-market malaise, companies need to flush out the excesses so investors can regain their confidence in our financial system. It's better to take a hit now than risk dragging this out over the next few years.

Many corporate directors already are beginning to correct for their marginal oversight and governance. Today, it's a rare audit committee that isn't asking tougher questions. Even at Southwest Airlines Co. (LUV ), where Chief Executive James F. Parker boasts that the balance sheet is "so clean that it squeaks," the audit committee is showing "more interest in more details."

Directors are so fearful of their legal liability that many of them are asking auditors for a redo of the previous years' numbers, which will inevitably result in far more restatements. "Every audit committee I know is doing that," says Ralph Larson, former CEO of Johnson & Johnson. "You'd have to be asleep if you're not."

The numbers already show that is exactly what is happening. If restatements continue at the first quarter's pace, they will reach a record 240 this year alone. Until the mid-1980s, it was unusual for financial restatements to top more than half a dozen in any given year. Then, after regularly running about 50 or so a year from the early 1990s, these redos skyrocketed to a high of 204 by 1999. Even so, from 1995 through 2000, restatements accounted for only three-quarters of one percent of the average annual number of reporting companies, according to Financial Executives International.

Yet most observers now agree that in the 1990s the corporate ethos reached a new and dangerous phase: Many executives began to believe that capitalism rewarded the companies that hit their earnings targets, no matter how they did it. For every Enron or WorldCom that went over the line, dozens of other companies took full advantage of flexible accounting rules to report the highest income possible--even when those numbers failed to reflect the company's underlying operating performance.

Besides the additional oversight, there's something else driving the current cleanup. The numbers at WorldCom and Xerox are so large that most restatements--including ones that merely take a more conservative view of the numbers--could very well fall under the radar. "There is so much other bad news out there, it might be a good time to release a restatement," says Min Wu, a PhD candidate in accounting at New York University's Stern School of Business who tracks such corrections. "In good times, one piece of bad news would get you hammered more severely."

Clearly, it's in the interests of boards and management to get the bad news out now. If there is an audit committee that has yet to order another look at the numbers over the last three to five years, shame on it.

Senior Writer Byrne covers management.

    Before it's here, it's on the Bloomberg Terminal.