(MONEY Magazine) – At General Electric, numbers are a very big deal. Take $10.7 billion. That's how much money GE made last year--more than any other public company. Then there's 20, which is how many years Jack Welch has run GE--longer than any big-company CEO. And don't forget 482%, which is the total return on GE stock the past five years.

Impressive as these figures are, they pale next to another number: 100. That's how many consecutive quarters GE has increased earnings from continuing operations. Considered the longest streak of its kind (Wal-Mart saw a 99-quarter string end in 1996), GE's run stretches back to the summer of '75, when the Dow was at 800, Gerald Ford was in the White House, and Al Gore was a cub reporter for a newspaper in Tennessee. One stock market crash, two oil crises, three recessions and four Presidents later, GE is still churning out record profits.

It is this consistency that has helped make General Electric one of America's most beloved stocks while making Welch its most admired chief executive. It's also a major reason why GE boasts the market's biggest capitalization--$590 billion. And it's why, despite Old Economy profit centers like manufacturing and finance, the company trades at a higher 2001 price-to-earnings ratio, 41, than Microsoft, Nokia and other New Economy heavyweights with projected growth rates that are much greater than GE's 15% annual clip. "There's nothing Wall Street hates more than a negative earnings surprise," explains Timothy Ghriskey, manager of Dreyfus Fund and a major GE shareholder with almost 4.2% of his portfolio in the stock. "There's just no question that GE's multiple has benefited from the consistency of its earnings."

Yet when you talk to fans like Ghriskey long enough, you learn that it's a charade--that GE has employed a number of confusing but apparently legal gimmicks to achieve its vaunted consistency. No company, not even one as well managed as GE, has 100 quarters of uninterrupted growth without resorting to some fancy accounting--or, as the mannered folks on Wall Street like to call it, earnings management. "A hundred consecutive quarters? Let's just say there's a very low probability of that ever occurring in nature," says Jay Huck, a forensic accountant with the Center for Financial Research & Analysis in Rockville, Md.

Unlike Mother Nature, GE leaves little to chance. According to the accounting experts, stock analysts, large shareholders and former executives we spoke with, GE enters every quarter with a specific profit goal in mind and then does everything in its power to hit that number--even if the company's actual performance turns out to be significantly better or worse.

HOW THE ENGINE WORKS

GE officials, including CEO Jack Welch, dispute this interpretation. In a written response to questions, Welch denies that the company makes decisions to smooth its growth. But when you read between the lines--and look closely at GE's practices--it becomes clear that performance is carefully calibrated. "With [12] major businesses, there can be significant variation in any quarter," Welch notes, before adding that GE's objective these past 20 years has been steady earnings growth "with no 'surprises' for investors."

GE's ability to do this often hinges on its immense finance unit, GE Capital. During an exceptionally good quarter, GE Capital can basically siphon off excess earnings from one or more of GE's other 11 units, usually by taking reserves against problem loans. During lean times, GE Capital can reduce these reserves, or it can sell off appreciated investments (such as venture-capital shares or traditional loans repackaged as asset-backed securities) and use the gains to pump up quarterly profits. By smoothing the peaks and valleys, GE avoids either an earnings shortfall in a current quarter or a number that will be hard to beat in a future one.

Want an example? Consider GE's results during the first half of this year. These days, much of the company's tweaking involves not enhancing performance but understating it, so as not to set the profit bar too high for future quarters. By most accounts, GE is enjoying a terrific year. Two of its highest-profile units, NBC and GE Capital, reported 17% profit gains during the first half. The appliance business (dishwashers, refrigerators and the like) hasn't fared so well, but other divisions, such as GE Power Systems, have more than picked up the slack. A maker of power-plant generators, the unit reported 70% higher profits on strong demand for gas-turbine engines.

Overall, GE's net income rose 20% during the half, but there are strong indications that reported earnings would have grown even faster had the financial managers not tapped the brakes. While earnings were up 20%, cash flow from operations (a harder-to-manipulate statistic that many believe to be more representative of a company's true performance) increased 25%. That five-percentage-point gap works out to approximately $350 million. Where did the money go? We'll get to the details a little later, but a short answer is that GE took reserves and charges that lowered 2000's first-half earnings. By doing so, the company will have an easier time showing healthy gains during the first two quarters of 2001.

EVERYBODY DOES IT

Right about now, the CEOs, CFOs and division heads reading this story are nodding in admiration. GE is exceptionally good at managing earnings, arguably better than any other company. But it is hardly alone. On the contrary, managing earnings--and expectations--has become standard operating procedure in corporate America. In a world where one quarter's results can make or break a stock, GE's methods have become the model that business follows.

Sometimes the imperative to maintain smooth growth numbers has led corporate managers to commit outright fraud; health-care-services company McKesson HBOC has recently been accused of just that, following the likes of Cendant, Livent and others. But more often the machinations that companies go through are perfectly legal. The accounting standards that govern public companies are intended to be flexible, and as long as financial maneuvers are disclosed--even when they're mumbled in footnotes at the end of quarterly filings--regulators generally can't stop companies from smoothing earnings.

Nor should they, argue Ghriskey and others, because earnings management reduces the volatility of a stock like GE, and that's good for long-term shareholders. The exclamation point here: GE's 42% compound annual return (including dividends) for the past five years. Under Welch, the company has convinced Wall Street that earnings consistency--quarter after quarter--is worth paying extra for.

But there's more to it than that, because to understand GE is to understand how difficult it is for the average investor to uncover the right lessons--about a company's past, its present or its future--from reported earnings data. One of the great ironies of 21st-century investing is that while news is being disseminated more freely than ever--via the Internet, cnbc and so on--the quality of financial information is suspect in ways investors wouldn't even imagine. And to some extent, the investing public has only itself to blame. Burned by stocks that flamed out after missing quarterly estimates, many investors demand predictability and have been moving toward big-cap names with steadier numbers. Problem is, the higher the premium investors pay for predictability, the more companies try to be predictable. "It's sort of like what we see in the public schools with standardized testing and all the cheating scandals," says Larry Bitner, an accounting professor at Pennsylvania's Shippensburg University. "What's measured gets managed."

JACK BE NIMBLE

GE's success has always been explained as a management story, with John Francis Welch Jr. in the lead role. Welch is the former junior engineer who rocketed through the ranks and at 45 became the youngest CEO in a storied company history that dates back to Thomas Edison. After taking over in 1981, Welch ripped apart GE's bureaucracy and immediately set out to prove that the operation could thrive in any economic environment. "I tried to dissuade our own people and the investment community from the long-held idea that GE was a 'GDP' company that would move in concert with the U.S. economy," he says today. Welch sold off more than 200 slower-growing businesses (everything from toasters to coal mines) and acquired higher-margin companies like medical equipment maker CGR and television pioneer RCA, which owned NBC. "Overall," he says, "the strategy has redefined the company and has given us a much higher-growth, less cyclical mix of businesses." Earnings have climbed, and so has GE's stock price. Given the impressive track record, there has been little reason for shareholders to question GE's numbers.

Welch will retire next April, however, and unless GE does the unexpected and goes outside the company for its next CEO, his successor will be a total unknown to most shareholders. (The head of GE Medical Systems, Jeffrey Immelt, is considered the front-runner.) "Investors have been willing to overlook a lot because of Jack Welch's extraordinary combination of business acumen and managerial skills," says Robert Friedman, an analyst who tracks GE for Standard & Poor's. "The question is this: Will investors take a closer look at GE's books once Welch retires?"

We think they should. GE's knack for smoothing earnings has lulled investors into thinking the company is immune to the twists of fate that hurt even the best-run operations. Conglomerates like GE do tend to have steadier earnings--much as a diversified equity fund is less volatile than the typical individual stock--but it takes work for GE to keep earnings up in good times and bad. In fact, were it not for some shrewd and complex financial transactions at the end of 1997, GE's earnings would have declined in 1997 and been flat in 1999.

It's important to know this because there comes a time when not even the best financial managers can disguise a downturn. Just ask Bill Gates. It has long been suspected that Microsoft smooths earnings, and because the company has been so dependable, most investors were shocked when Microsoft's sales and earnings growth began to decline last spring. Word of the slowdown helped fuel a 24% meltdown in the Nasdaq in April and May, wiping out $1.5 trillion in paper wealth. Should the law of averages catch up with GE, many investors may suddenly be asking themselves why they're paying 41 times earnings for GE when most companies with its growth rate have a P/E closer to 15.

"A NUMBERS GAME"

While the fact that GE manages its earnings may not be well known among individual investors, it's well understood among many of the pros. William Fiala, an analyst who follows the company for A.G. Edwards, matter-of-factly points out that under Welch, GE has never taken a restructuring charge that was not offset by a like-size gain. "To a certain extent, they do have a lot of control over earnings in the near term," says Fiala.

Of course, GE insists otherwise, saying it is the diversity and strength of the conglomerate's many businesses that account for the company's remarkable record. "Consistency," Welch says, "comes from managing businesses, not earnings." GE's sensitivity on this issue may have something to do with a Securities and Exchange Commission campaign to curb excesses in earnings management. In a 1998 speech, SEC Chairman Arthur Levitt warned that the reporting of corporate earnings has become "a numbers game" and he blamed management's "zeal to satisfy consensus earnings estimates and project a smooth earnings path." Too often, Levitt said, "trickery is employed to obscure actual financial volatility," and not just by small start-ups: "It's also happening in companies whose products we know and admire." The chairman never mentioned GE by name--and he won't discuss specific companies today--but it is hard to think of a company that better exemplifies what he was talking about.

THE MISSING $350 MILLION

GE's 12 major business units--each with billions in revenues and dozens of divisions and subdivisions--offer the company plenty of places where it can squirrel away profits during good quarters, only to release them when times are tougher. The reserves that GE takes are supposed to be a hedge against anticipated losses from bad loans or inventory. However, deciding when and how much to reserve is inherently subjective, and there's often no way to know if there are other motivations behind the timing. "The line is not always clear," says Eli Bartov, an accounting professor at New York University.

Zachary Abrams, a former GE executive who was a member of the corporate audit staff in the late '80s and early '90s, says that when he worked there, the company routinely managed earnings so they came in at or just above Wall Street estimates. "As an investor," Abrams asks, "wouldn't you prefer a company that's going to grow its earnings a certain amount every year, rather than one that has a 20% increase one year and a 10% decline the next?"

GE knows the answer, and that's probably why it chose not to max out its earnings during the first half of 2000. Remember the missing $350 million? GE's public filings provide few clues to its whereabouts, other than a vague reference to first-quarter losses "resulting from repositioning of the ERC portfolio." ERC is the Employers Reinsurance unit of GE Capital, and it turns out that GE offset some of its $400 million in venture-capital gains by taking a $150 million loss in ERC's investment portfolio and a slightly smaller restructuring charge elsewhere within GE Capital. That's what the company told analysts behind closed doors. While the numbers don't add up perfectly, analyst Lawrence Horan of Pittsburgh brokerage Parker Hunter notes that GE did similar smoothing in the second quarter, wiping out a $300 million gain from the sale of assets by taking reserves against loans with potential credit problems. "They had the gain, and they looked to take reserves in an area where there might otherwise have been an impact on earnings down the road," Horan explains, repeating what he says he was told by GE executives.

What happens to those reserves later? If the problem loans end up being repaid, the reserves can be moved back into the profit column--but they can also be returned there if GE concludes the credit quality of the problem borrowers has improved. Clearly, there's leeway here under the accounting rules. "GE Capital has the ability to be a lot more nimble than a lot of the old-line manufacturing businesses," says GE alum Andrew Campbell, now CFO of Tenneco's spun-off packaging business. "The assets are a lot more portable." Indeed, the assets are so portable that after all the first-half machinations, GE Capital ended up growing its profits 17% for the period--which just happens to be the finance unit's annual average since 1980.

"TOMORROW'S EARNINGS...TODAY"

Taking reserves is not the only way GE can smooth earnings. It can sell off real estate. Or it can employ something known as a sale leaseback, according to former GE executive Abrams, who is now a vice president with Internet security firm SonicWall. Large industrial companies often have plants or equipment worth more on the open market than the depreciated values at which they're on the books. GE can turn a profit by selling that plant or equipment to another party and then leasing it back.

Then there's securitization. GE Capital has $94 billion in loans on its books, and each of those loans has an open-market value that rises or falls based upon interest rates and the credit quality of the borrowers. Packaging $1 billion of appreciated loans and selling them to investors in the form of asset-backed or mortgage-backed securities can net GE a gain of $50 million, sometimes more. "Securitization is a way of bringing tomorrow's earnings into today," Abrams says.

GE's securitization history supports the contention that the company uses sales of asset-backed and mortgage-backed securities to provide an 11th-hour boost to quarterly earnings. Since 1995, GE has issued these securities seven times in the public markets, according to Thomson Financial Securities Data. All but one of these deals occurred during the last three weeks of a quarter. The largest recent offering was a $1 billion deal in September 1998, the end of the third quarter. It was an unspectacular time for GE, with revenue up 10%. Earnings, however, rose 15%, and in all likelihood the proceeds from that $1 billion securitization helped fill the gap. "These are not spur-of-the-moment, end-of-quarter actions," insists a GE official, who notes, for example, that "sales in the capital markets are used to adjust concentrations of risk" in the GE portfolio. GE also says our figures exclude private market deals--but by the company's own count, 40% of its securitizations since December 1994 occurred at the end of a quarter.

HOW LAYOFFS CAN HELP

Reserves, sale leasebacks and securitizations can be important tools for GE, but the company's boldest forays into earnings management have involved one-time gains from the sale of businesses and other assets, losses and restructuring charges linked to layoffs and plant closings, or some combination of the two. "If you're talking about corporate restructurings, yeah, those charges are used strategically when there are excess earnings," says SonicWall's Abrams.

While charges and offset gains are always disclosed in GE's quarterly reports, it's not apparent from the filings how important they've been to the earnings streak. Consider the company's fiscal handiwork at the end of 1997. GE ate two charges that year. One was a $200 million charge GE was forced to take when one of GE Capital's borrowers, retailer Montgomery Ward, declared bankruptcy. GE offset that loss by selling part of its stake in financial services giant PaineWebber, shares GE had acquired when it sold its Kidder Peabody stock brokerage to PaineWebber in 1994.

The second charge was much bigger, and unlike the first, its timing was completely discretionary. In the fourth quarter of '97, GE took a massive $2.3 billion hit against earnings to pay for plant closings in some of its weaker industrial units. The layoffs would have happened eventually, but they clearly were timed to coincide with a separate deal GE had struck with Lockheed Martin. This highly complex transaction--it involved GE getting two Lockheed businesses in return for Lockheed preferred stock that GE had acquired years earlier--netted GE a paper gain of $1.54 billion plus tax breaks worth another $600 million.

If GE hadn't offset charges with gains, its growth curve would have gone from smooth to jagged almost overnight. Had the $2.3 billion charge been taken on its own during the fourth quarter, the earnings streak would have ended at 89 quarters, and yearly profits for 1997 would have declined from the previous year's total. And if the Lockheed transaction had occurred at the beginning of '98 instead of the end of '97, GE's earnings for 1998 would have ballooned more than 25% to about $11 billion, creating an extremely difficult comparison for the following year. (Remember: GE ended up posting just $10.7 billion in earnings in 1999.) So a quarter's worth of earnings management might have averted disappointing profit pictures in two of the past three years. Of course, everything would've balanced out over time, and GE's earnings still would have grown at an annualized rate of 15%. But stripped of its beloved consistency, GE's stock might be trading at a lot less than today's 41 times earnings.

Welch takes issue with our analysis. "It is purely hypothetical," he says. "There are always annual events that if moved from one year to the next would create a whole different picture." That's true. But back when the Lockheed deal and the layoffs were announced, at least one GE official said the timing was no coincidence. "Offsetting the gain has been our practice," a spokesman told a newspaper in upstate New York, where 1,000 jobs were lost.

GE's earnings consistency will be tested again this winter after PaineWebber's buy-out by Swiss bank UBS becomes final. GE still owns 31 million shares of PaineWebber, and according to documents filed with the SEC, it stands to realize a pretax gain of $1.4 billion as a result of the UBS-PaineWebber merger. If history is any guide, GE may try to offset this windfall with an equivalent restructuring charge. John Hovis, president of one of the unions representing GE workers, just hopes his members aren't the ones restructured. "Of course I'm worried about layoffs," says Hovis, head of the United Electrical, Radio and Machine Workers. "Every time we meet with GE executives, they keep telling us how Wall Street likes smooth, steady growth." Welch's response? "When unfortunate layoffs are necessary, we make them," he says, "and they have nothing to do with windfall--as you call it--gains."

THE 13TH BUSINESS

The only thing that gets Hovis and other union leaders more exercised than GE's conveniently timed downsizings is the way GE manages its pension fund. At $50 billion, it is 100% overfunded (according to GE's annual report), thanks largely to the bull market. This means that theoretically GE could double the benefits it pays out to current and future retirees without it costing the company an extra penny. In GE's last round of union negotiations, how-ever, the company would agree only to a small increase in benefits.

Why? According to union leaders, the answer has everything to do with earnings management. Under the accounting rules, when a pension fund earns more than the defined benefit that it is obliged to pay, the difference is reported as either an expense reduction or pension income. Either way, the money drops to the bottom line. Last year, GE reported $1.4 billion in pension income, which means the pension fund was responsible for 10% of the company's pretax profits.

Pension income stays in the fund and cannot be used for other corporate purposes, but this distinction is lost on most investors. "There's only one reason we're not doing better on this thing, and that's because the pension contributes to GE's earnings record," says Stephen Tormey, a negotiator for the electrical workers. "It's our position that the pension has essentially become GE's 13th business."

GE doesn't like to talk about its pension income, probably because doing so highlights an uncomfortable truth: Some $2 billion of GE's pretax earnings over the past two years weren't real profits. Friedman, the Standard & Poor's analyst, is troubled by the extent to which pension income is boosting GE's profits. Getting information about the pension fund from GE is difficult for him and would be next to impossible for an individual investor. "It's in the annual report, buried in the footnotes," Friedman says. "The average investor wouldn't even know what line item to look for."

The controversy surrounding GE's pension income will only grow in coming months. At the end of 1999, GE raised what's known as the discount rate on the pension fund by a full percentage point, to 7.75% from 6.75%. The discount rate reflects prevailing interest rates and is used by actuaries to calculate the cost of meeting a pension fund's obligations. GE has little discretion when it comes to setting the rate. What's important here is that last year's move will dramatically raise the amount of pension income the fund generates in 2000. Adam Reese, a consultant for Towers Perrin and one of the nation's leading pension experts, estimates that a one-percentage-point increase in the discount rate of a $50 billion fund would generate an extra $1 billion in yearly pension income. If Reese's calculations are correct--GE claims he's wrong but won't say by how much--it would mean the change in the discount rate was responsible for half of GE's first- and second-quarter earnings gains. Unfortunately, the truth will not be known until GE files its 2000 annual report early next year. Despite the fund's huge contribution to earnings, GE does not break out pension income in its quarterly reports.

DECEPTION OR PERCEPTION?

An obvious irony here is that many of GE's 340,000 employees own GE stock, and they wouldn't be enjoying their 42% annualized returns were it not for the practices their union leaders decry. For them and other shareholders, it's perfectly reasonable to ask: As long as no laws are being broken, what's the big deal about earnings management?

The big deal, says Howard Schilit, a forensic accountant who advises money managers on stocks, is that GE is forcing investors to make buy-and-sell decisions based upon misleading information. "The integrity of the financial report requires that companies report the numbers fairly and honestly," says Schilit. GE's use of reserves, leasebacks and securitizations to manage earnings, as described to Schilit by MONEY, "is blatantly improper. They are deliberate attempts to deceive the reader of financial reports."

Plenty of mutual fund managers know that GE manages earnings and yet continue to own GE stock. The risk they're taking, Schilit says, is this: If GE ever ran into problems, it could be several quarters or even a year before those problems showed up on GE's bottom line.

Think back to Microsoft. For years it deferred a share of revenue from every sale into a line item it calls unearned revenue. Fairly common in the software business, these deferrals are used to pay for maintenance or software updates. But in Microsoft's case, unearned revenue was growing so fast that Schilit became convinced that Microsoft was using the account as a rainy-day earnings fund. That's why he got suspicious when the unearned-revenue line stopped growing in the second quarter of 1999 and actually declined in the third: He thought Microsoft was tapping these funds to mask a declining growth rate.

Microsoft denies that's the case, but the fact is, its revenue growth slowed into the first quarter of 2000 and by the second quarter, its earnings were growing a mere 9%. On April 12 (nine months after Schilit warned his clients), Goldman Sachs forecast lower sales for Microsoft. The market immediately knocked $4.50 a share off Microsoft's already antitrust-depressed stock price--pushing tech stocks over the cliff. "That's the problem when a company like GE starts playing games," Schilit says. "When a company creates reserves, how are you to know when business is struggling and they are turning on the spigot?"

It's impossible to know. You may recall how many pros missed the boat on Intel's slowdown in September. Sure, there may be warning signs along the way, but if a big company wants to disguise a negative operating trend, it can usually do so in the short term. So long as investors believe two years of 15% growth are worth more than one year of 5% growth followed by another of 30%, GE and others will continue to manage earnings in good times and bad. Edison once said the greatest discovery of all is discovering what people want. It's a lesson that his heirs at GE have learned exceptionally well.