This Elephant Dances Too Well

Did accounting trickery help Louis Gerstner turn around IBM?

IBM’s stock was pounded Tuesday after the company disclosed that the Securities and Exchange Commission was investigating how it accounted for certain transactions in 2000 and 2001. While the probe may not amount to anything, this isn’t the first time the SEC has questioned IBM’s recent accounting practices. The accounting doubts are starting to take some of the luster off the acclaimed reign of former Chief Executive Officer Louis Gerstner Jr.

Gerstner’s nine-year tenure at IBM was wildly successful by all accounts—especially his own, the recently published Who Says Elephants Can’t Dance? Inside IBM’s Historic Turnaround. Written without a ghostwriter, Elephants briskly describes how Gerstner, an outsider in IBM’s insular and rigid culture, made tough decisions (slashing the dividend, selling off assets, ditching PC manufacturing) and transformed the company’s culture and business strategy by pushing into services. Under Gerstner, the company swung from an $8.1 billion loss in 1993 to a $7.7 billion profit in 2001. The stock price rose nearly tenfold, even as revenues rose only 37 percent. Writes Gerstner: “The best companies grow profits faster than revenue. They manage margins and expenses brilliantly.”

They also manage earnings brilliantly. The Gerstner miracle was as much about financial engineering as computer engineering—a fact he is reluctant to acknowledge in Elephants.

One of Gerstner’s shrewdest moves was selling off the global network that IBM had built to deliver data. Telecommunications companies were desperate for such capacity in the late 1990s, and IBM held an auction in 1999. “We thought we’d be doing well to get $3.5 billion, but the frenzy eventually produced a bid of $5 billion from AT&T,” he writes. IBM notched a $4.06 billion gain on the deal. But rather than account for it as an extraordinary one-time item—it’s not every day, after all, that you clear $4 billion on a deal—the company, as William Bulkeley reported in the Wall Street Journal in March 2002, instead reduced its expenses under its “Sales, General and Administrative” category by about $4 billion. By so doing, IBM made it appear it had saved $4 billion through increased efficiency, rather than scored a one-time bonanza. In late 1999, the SEC summoned IBM brass to Washington to quibble with this accounting treatment.

In his book, Gerstner also complains about “quarterly myopia”—the relentless drive by companies to meet the frequently arbitrary quarterly earnings estimates put out by Wall Street analysts. (Alex Berenson’s The Number provides the best examination of this malady.) Analysts and investors foolishly swore by the quarterly number, Gerstner laments. “Never mind that 90 days has absolutely no bearing on how well a company is doing. … Never mind that what a company can do in 90 days is manage earnings.”

Even so, Gerstner seems to have put his own executives under pressure to hit their numbers. In Power Failure, Mimi Swartz and Sherron Watkins’ describe then-IBM COO (and current IBM CEO) Sal Palmisano’s weird appearance at a 1999 Enron management conference: “IBM was going to miss earnings targets that quarter, he confessed. Suddenly, it was all too much for Palmisano. His eyes began to fill, and then his voice quavered. And then he had to stop talking. Everyone could see he was trying not to cry.”

And on the last day of the last quarter of 2001, IBM sold an operation to JDSUniphase. That fortuitously created a $280 million gain, which, as Bulkeley noted, “helped it meet analysts’ earnings expectations for that quarter.”

In Elephants, Gerstner also writes of his effort to pare back IBM’s “paternalistic benefits structure.” The company’s pension system “was geared to the company’s prior commitment to lifelong employment,” he writes. “We had to create benefits programs that were more appropriate to a modern workforce.” Yet Gerstner relied on that old-style defined-benefits program to fuel IBM’s remarkable earnings performance.

Like virtually all other similarly situated companies, IBM routinely counted the change in value of its massive pension plan toward income. The practice is acceptable under accounting rules but has little bearing on the underlying health of a business. Companies can’t use profits generated by pension funds for anything but pension funds.

According to Bulkeley in the Journal, the SEC expressed concerns as to “whether the company had told investors enough about the extent to which pension fund gains contributed to its bottom line.” In its 1999 10-K, the SEC complained, Big Blue didn’t mention pension income in the main section where it discussed results—even though data indicated that pension income accounted for nearly 6 percent of the company’s pretax income.

In its 2000 10-K, IBM did provide more disclosure. It reported that in 2000 it realized “cost and expense reductions” of $1.17 billion from its pension plans, up from $694 million in 1999. In 2000, the company’s net income was $8.1 billion. Subtract the pension gains and the asset sales, and Big Blue’s profits under Gerstner look somewhat less impressive.

The prior SEC investigations into IBM were closed without action. The same may happen with the current investigation. And one wouldn’t expect a CEO to be involved in every aspect of accounting or corporate governance. But it is clear that IBM, like so many other companies in the 1990s, went through certain contortions to meet Wall Street’s estimates. And Gerstner, who was so influential in forging a new culture at IBM in the 1990s, had to know that the company was pushing the envelope. “Show me a business executive who doesn’t completely understand the financial underpinnings of his or her business,” he writes, “and I’ll how you a company whose stock you ought to sell short.”