(FORTUNE Magazine) – NO ONE SEEMS to be noticing, but Jack Welch has been quietly shifting policy at General Electric in a way that suggests a seismic change in corporate strategy. For years, GE's huge financial arm, GE Capital Services, has produced an outsize share of the company's profits. And for years as well, Chairman Welch has let the men and women who earned that money reinvest it, and $2 billion more, in the exquisitely profitable ventures they proved so capable of developing. But since 1990, Welch has been pulling money out of their hands. Big money. Some $600 million last year alone. Which means he must think he has an even more profitable place to put it. Just how productive the managers of GE Capital Services have been can be seen in a new calculation by Stern Stewart & Co., a New York City consulting firm that developed the concept of ''economic value added.'' Known as EVA, economic value added measures the wealth managers have created -- or lost -- for shareholders in running a business. Owners profit when the stock market translates EVA into higher share prices. To determine EVA, Stern Stewart adds up the cost of capital employed in the business and deducts it from the after- tax profit earned on operations. What's left over is the ''value added'' for shareholders. EVA has been adopted by Coca-Cola, AT&T, and many others as the single best measure of corporate performance.

| When you compare GE's EVA with that of its GE Capital Services subsidiary, as Stern Stewart did at FORTUNE's request, the results will set your hair on fire. For the five years through 1992 (the most recent period for which detailed information is available), GE actually lost $503 million of economic value when its financial services company is excluded. By contrast, GE Capital Services standing alone created economic value of $1.7 billion, bringing the total corporation's EVA to $1.1 billion. To put GE Capital Services' $1.7 billion contribution in context, compare it with the value created over the same period by such stock market darlings as McDonald's ($1 billion) and Kellogg ($1.6 billion). Like Aeneas bearing his aged father, Anchises, out of the smoking ruins of ancient Troy, GE Capital Services has carried its parent on its back.

Viewing GE's brawny subsidiary through the EVA prism raises some fundamental questions. How in the world has this agglomeration of 24 financial service businesses, ranging from aircraft financing in its equipment management operations to annuities in its consumer services businesses, done such a spectacular job? How long can GE Capital Services keep it up? And where else can Welch go to find returns like those? GE's wunderkind includes Employers Reinsurance, a solidly profitable property and casualty company, and Kidder Peabody, an investment bank and securities broker that has thrived in today's buoyant market. But the main engine of GE Capital Services' superior performance is a finance company called GE Capital Corp., which is the biggest piece of the pie, accounting for 60% of the assets, 83% of the earnings, and 90% of the soul of this money machine. Run by the rumpled, ebullient Gary Wendt, 51, GE Capital pours wealth into the corporate coffers by doing just about everything you can do with money except print it. This complex business leases equipment, lends money, sells insurance, and much, much more. In every operation Wendt and his team look for ways to gain an edge over their more ordinary competitors. ''We want to differentiate ourselves,'' he says. ''Money is a commodity.'' GE Capital has become the largest finance company in the U.S., having surpassed General Motors Acceptance Corp. in 1992. From 1988 to 1992, its earnings grew at an annual rate of 19%, to $1.3 billion, while its assets exploded by 18% a year. If it were publicly traded, its $93 billion in assets would put it fifth on FORTUNE's list of diversified financial companies, just behind Merrill Lynch. Alchemy should be as simple as GE Capital's formula for ginning money. First of all, GE Capital has some intrinsic advantages. Its effective tax rate has been only 25%, largely because it holds over $3 billion of tax-exempt municipal bonds through its insurance affiliates. Parent GE's effective rate has been 34%. Its position in the leasing business also allowed GE Capital to defer $5.1 billion in taxes in 1992, while the parent corporation had a net deferred tax surplus of $329 million. That said, any explanation of GE Capital's extraordinary performance must quickly turn to the character of the company. With its multitude of businesses, GE Capital ought to have a personality as fragmented as Woody Allen's. Instead it is more like Clint Eastwood, scanning the horizon squint- eyed for a fistful of dollars. Says Mike Fraizer, who runs the unit that finances commercial real estate: ''When you get up in the morning, if you're not thinking of a way to make more money, and if you're not still thinking about that on the drive home, this isn't the place to work.'' The management hails from the Edith Piaf school of business: ''Je ne regrette rien.'' When the men and women at GE Capital make mistakes or a market turns against them, bosses don't waste time looking for heads to roll on the kegling greens at corporate headquarters. Instead, they hunt for innovative ways to make a buck out of a bad situation. Norm Blake, chief executive of USF&G, the Baltimore insurance company, was executive vice president of GE Capital's finance operations until 1984, and he recalls the spirit: ''There were a lot of shootouts at the O.K. Corral there. Everybody was kind of full of themselves: They were smart, and they had the toys to play with, and they basically wanted to reinvent the world.''

THAT COMBATIVENESS persists like bowlegs, and nowhere is it more evident than in the way GE Capital's people do deals. As negotiators, they are respected for being tough, well prepared, very tough, fair, and extremely tough. In his spartan Stamford, Connecticut, office Fraizer displays two gifts from folks on the other side of some of his real estate deals. One is a 12- pound sledgehammer, the other a six-pounder. Says he: ''They gave me the smaller one after the second deal because they said I was getting nicer.'' Not that he'd be confused with Mother Teresa. Say a commercial real estate borrower falls behind in paying his loan. Rather than foreclose, as a bank would, Fraizer and his crew might restructure the loan and put money into improvements so the building can attract more tenants. Nice guys, eh? Just wait. Next, they sharpen their pencils and go after costs with a vengeance. Stephen Mann of the Clifford Cos., a New York City real estate firm that works for GE Capital on such deals, knows how flinty its managers can be. ''If you send something Federal Express priority,'' he says, ''the GE guys want to know why you didn't send it second-day air.'' They find other ways to swing the sledge as well. They remind the owner that if they did decide to foreclose rather than restructure, he could be liable for a hefty capital gains tax. The tax would fall on the difference between the unpaid mortgage -- in a sense, the price at which he ''sells'' the building to GE Capital in a foreclosure -- and the property's book value. The threat of that tax, which can be substantial on depreciated real estate, gives GE Capital the leverage it needs to muscle the owner into accepting reduced fees for managing the property when it restructures the loan. Says Mann: ''GE has a lot of people working for it very cheap.'' Employees treat each other with the same politesse. For example, in an effort to stanch the bleeding from the company's real estate portfolio, Wendt ordered his troops to rent space whenever possible in buildings that GE Capital had a stake in. That led to fierce negotiations in 1992 when Financial Guaranty Insurance Co., GE Capital's municipal bond insurance business, moved into new quarters in New York City. Stephen Berger, now at the Odyssey Partners investment banking boutique in Manhattan, was chairman of FGIC at the time and recalls the none-too-brotherly bargaining. ''Okay,'' he says, ''we had to go into a GE building. But we beat the s--- out of the real estate group over who would pay for improvements.'' Did that make him happy? Not really. ''We got a market rent, but not the bottom of the market,'' he grumbles. ''I wouldn't have paid it if GE hadn't held the mortgage.'' The way the deal turned out still rankles. Tough guys can blow it too. From 1990 through 1992, GE Capital wrote off $1.1 billion in commercial real estate and loans on highly leveraged transactions. Its credit people were dazzled by the fees, running into the millions, that they could earn assisting the leveraged buyouts of such concerns as Harcourt Brace Jovanovich's trade publishing unit and Channel Home Centers. Subsequently, both slid into Chapter 11, but have since emerged. The company will be lucky to break even on a decade of lending to the overleveraged. As he looks back, Wendt says that the fiasco violated his principle of ''differentiating'' GE Capital from competitors: ''We got in the roll of the dice with everybody else, and we were doing the same thing everybody else was.'' What ultimately differentiated GE Capital, he concedes, was its ''willingness to take on risk'' -- and the company got caught. In commercial real estate, Wendt blames the vicious turn in the market that caught not only GE, which was fairly conservative, but a lot of other lenders too. When markets turn against the company, its no-surrender attitude serves it well: It has the money and the moxie to place big bets on depressed sectors of the economy, and it has the mettle to wait for a rebound. So when real estate turned to scorched earth in the early 1990s, GE Capital stepped up and became one of the biggest buyers of distressed properties in the country, picking up $1.1 billion worth from the federal government's Resolution Trust Corp. alone. Now, with the market starting to show signs of life, GE Capital has become a seller. In 1993 it actually showed a modest profit in its property business, after barely breaking even the year before. FOR STICK-TO-ITIVENESS, consider too its adventures in the much buffeted airline industry. At the end of 1992 the company, which is one of the largest aircraft lessors in the world, with a fleet of about 500 commercial aircraft, had about $6 billion in leases and loans to the airlines. But one toxic byproduct of airline deregulation has been too many planes. The value of used aircraft fell through the floor and into the root cellar. Time to write off the leasing business, right? Instead, GE Capital averaged down the cost of its assets, much as an individual investor does when she keeps buying her favorite stock as its price falls. The company waited until GPA, the Irish aircraft lessor struggling under $5.8 billion of debt, was teetering on the edge of bankruptcy. Then last October it stepped in with a rescue plan that allowed it to cherry-pick 45 of GPA's newest aircraft for 85% of their book value and to acquire -- cheap -- an option to buy up to 80% of the company in 1998. Meanwhile GPA will pay GE Capital an annual fee equal to 0.43% of GPA's assets, or about $41 million in the first year, to manage its fleet of aircraft. When the aircraft market does begin to recover, which some experts think could happen in 18 months, GE Capital will have gains from the purchase of its new planes to offset losses on its old fleet. And while it waits for the turnaround, it will have access to GPA's crack worldwide ''remarketing'' team that finds new customers for planes as the planes come off lease. GPA's remarketing will protect the value of GE Capital's fleet by helping GE find new lessees if current ones get into trouble. GE Capital will put its chips on any table where it sees the promise of big returns. Since 1986 it has spent $27 billion acquiring such operations as private-label credit cards, which carry the name of the clothing store, oil company, or other retailer that honors them. GE Capital is the world's largest supplier and manager of this kind of private-label plastic, having 70 million accounts with over 300 retailers in the U.S., Canada, and Europe. This is a dream business. Competition and consumer pressure have pushed interest rates on bank credit cards down around 17%. But in the private-label niche, there is no other provider of, say, a Macy's card to compete with GE Capital. So interest rates remain comfortably above 20%, while GE Capital pays 5.84% to finance the unpaid balances. That's some spread! And should charge card customers get tetchy about that 20%, they are likely to direct their ire at the merchant whose name is on the card, not at the finance company behind it. GE Capital Mortgage Corp. has bought the rights to service over $40 billion worth of mortgages since 1990. Wonder why? First, it earns a fee for billing homeowners and dunning them if they don't pay. That's hard-earned money. But there's a smoother road to Easy Street in this business. Monthly payments include charges for taxes, homeowner's insurance, and mortgage insurance that servicers hold in escrow until the charges come due, usually once a year. Until then, mortgage servicers collect the interest on the float. It should come as no surprise that GE Capital also owns the largest private mortgage insurer in the U.S. If you think that the company lets the mortgage servicers benefit from the float at its expense, wait. It just introduced monthly premiums for its insurance, which will keep some of that float out of the other guys' hands. The mortgage insurer specializes in covering home loans with as little as 5% down, thus tapping into a huge underserved market -- low- income householders. Congress is beating its drum for banks to lend more money in poorer neighborhoods these days, so GE Capital has a powerful, and free, marketing force working in its favor. Says Greg Barmore, president of the mortgage business: ''We love the regulators.'' He should. The mortgage operations made $196 million in 1992, up 34% from the year before. VIRTUALLY ALL of GE Capital's businesses are big and enjoy significant economies of scale. But the last thing Wendt wants to do is compete on price and thus reinforce the notion that his service is a commodity. His preferred approach is to follow a business process through all its stages and identify ways to add value -- or in common parlance, to make money. For example, the company's railcar-leasing business is the largest in the nation and includes six wheel-repair shops scattered over six states to help customers maintain what they rent. Here workers inspect the part that keeps the roll in rolling stock, machining it to tolerances as tight as one one-thousandth of an inch. Even minor flaws can cause derailments. The fleet management business, which serves corporations that lease cars, trucks, and special vehicles, doesn't compete on price, either, but on how much it can save customers on services provided by others. When a company leases a fleet of cars, the price paid for the use of the vehicles is only the first of many costs. The lessee also foots the bill for dealer preparation, delivery, maintenance, repair, and registration. The market for these services is highly fragmented. GE Capital squeezes out the inefficiencies of mom and pop service providers through its sophisticated computer system and extensive database. In a well-lit office room in Fleet Services headquarters in Eden Prairie, Minnesota, where deer wander through the parking lot, an army of auto mechanics man a battery of phones, wangling the best prices on car repairs and keeping track of routine maintenance. In another room, employees bargain with dealers for new cars by flexing their purchasing might, and then work out timely delivery. In still another area, workers arrange for vehicles to be registered quickly, reducing downtime. You pay more for GE Capital's services, but you will pay all the other service providers less. Says Teresa LeGrand, president of Fleet Services: ''For every $1 in our service fee, we return $3 to $4 in savings to our customers.'' SUCH IS its aggressiveness that this finance company juggernaut appears to roll easily over difficulties and flatten its competitors. But one problem keeps jamming the tire treads: Where to go from here? In many businesses GE Capital simply can't get much bigger. Last September the Federal Trade Commission pressured the railcar leasing unit, which owns the largest fleet of boxcars in the U.S., into withdrawing its bid to acquire Chrysler's fleet, which is the second largest. The Feds feared that the acquisition would decrease competition. In November, GE Capital was reportedly interested in buying a British container-leasing company to add to its own Genstar, the world's biggest container lessor. But the company didn't pursue it, the story goes, because of antitrust fears.

A large potential area for growth lies overseas. GE Capital is bent on becoming more international, and its efforts produced about $150 million in earnings in 1993. But progress will take time. The more service-oriented businesses like fleet management are difficult to export to the hot markets in emerging countries because there is no infrastructure to support the car repairs and other handholding. Getting the right managers will also be tough. The company seems to breed a type of hell-for-leather but parochial executive who may not be readily transportable. Says Bob O'Reilly, the ex-Marine paratrooper who runs GE Aviation Services: ''I'm not sure the answer to going international is guys like me who were born and raised in the Bronx.'' So now, rather than launching major overseas expansion initiatives, Wendt is attacking costs. By pressuring suppliers to lower prices over the past year, GE Capital has cut expenses by $90 million, or 10%. Is GE Capital coming to the end of its great growth run? Not yet. Parent GE released fourth-quarter 1993 earnings recently that showed profits up 10% for the year, and it identified GE Capital Services as a leader, with ''double digit'' increases. But the fact that Welch is pulling $600 million out of GE Capital Services suggests that he thinks the business has matured and some of its earnings can be better deployed elsewhere. But where? In a wood-paneled conference room adjoining his office at the Fairfield, Connecticut, headquarters, Jack Welch leans back in his chair and circumnavigates the question. Asked why he is putting the bite on financial services, he reaches for a stick of sugar-free gum from the bowl he keeps in the center of the table and launches into a long rumination about how GE Capital Services is finally being made to underwrite its share of the corporate dividend. True enough, but as Welch well knows, that's not the point. THE MONEY he has tapped will surely come in handy as GE pushes into China and other Asian markets that Welch feels it must be in, despite the risk. Says he: ''I go to Mount Sinai Hospital ((in New York City)), and I have got to get down on my hands and knees to sell one CAT scan. I go to China, a guy comes to lunch and orders 100.'' Beyond that, might Welch have major acquisitions in mind? The signals are mixed. Officially the company says it has made no changes in strategy. But at least two senior executives admit privately that GE plans to emphasize growth by acquisition more than in the past. Welch himself says that he expects the rest of the company to keep pace with the growth of GE Capital, a feat that will be easier to pull off with more horses in the harness. And during a discussion of the company's vaunted credit rating, he hints broadly at an urge to merge, suggesting that he might be willing to accept a downgrade if an exciting enough acquisition prospect came along. Or as he puts it, ''One would lose the AAA for the sake of a strategic acquisition that would take one to the next level.'' Whatever his course, Welch has set himself an imposing challenge, particularly when viewed in the light of EVA. Because the subsidiary earns such high returns, the cost of capital represented by that $600 million dividend is very high. In order truly to create wealth, the CEO must do more than simply reinvest that money at a profit, even a fat one. He must earn more than he could have if he had left the cash downstream. That boils down to finding an investment vehicle more powerful than the money machine he helped create. The Man of La Mancha had dreams like that.

CHART: NOT AVAILABLE CREDIT: FORTUNE CHART/SOURCE: STERN STEWART CAPTION: ECONOMIC VALUE ADDED

CHART: NOT AVAILABLE CREDIT: FORTUNE CHART CAPTION: THE JUICE IN GENERAL ELECTRIC