(FORTUNE Magazine) – For a glorious three decades, American venture capitalists did a lot of good for the economy while also doing very well for themselves. Arthur Rock and Tom Perkins, among others, got famous and rich helping start and build high-tech stars like Intel, Apple Computer, and Genentech. But now the company-creation machine that has been the envy of the world is sputtering. There's even a question whether organized venture capital is the best way to finance startups.

Take a look at the photo on the opposite page: The men around the table are angels, mostly retired and semiretired Silicon Valley executives, listening to a dinnertime presentation by an entrepreneur. They are trying to fill a big gap left by prominent venture firms that have lost their appetite for startups in favor of ongoing enterprises into which they can shovel $5 million or more at a crack. According to a survey of 637 venture funds by Venture Economics, a Boston research firm, total investment in early-stage companies skidded from an all-time high of $1.1 billion in 1987 to $749 million in 1993. If persuaded, the angels will invest small amounts of their own money, plus time and expertise.

Why has the venture capital industry changed? In a word, money. Generous management fees not tied to performance are skewing the business. Last year venture funds raised a record amount of nearly $5 billion. But three-fourths of that was concentrated in fewer than a quarter of the funds. That led to the creation of megafunds of $250 million or more whose partners derive astounding profits simply from handling the money.

Consider this: A venture firm customarily collects an annual management fee of 2% or 21/2% of invested capital--usually for the duration of the fund, which can be as long as ten years. Says Antony Hoberman, senior vice president at Alliance Capital in New York City and head of that company's venture capital department: "I don't think venture capitalists even know they've lost their objectivity about what's the right size for their funds. The bigger the fund, the bigger the fee." General partners (the investors are limited partners) can make annual salaries of up to $1.5 million from fees alone. On top of fees, venture firms get 20% or, in some cases, 30% of profits, which can boost a partner's take five- to ten-fold.

The venture capitalist's job is to offer investors returns at least several points higher than those of the S&P 500, and possibly double or triple that. But what's happening now, says Pierre Lamond of Sequoia Capital in Menlo Park, California, is that earning fees by putting the maximum amount of money to work has become an end in itself, "as opposed to looking for the best possible investment." Partly because of that shift in emphasis, average annual returns, once near 50%, have fallen dramatically, recently sliding below the returns of the S&P 500 (see chart). Some venture capitalists are looking for higher returns overseas. Kevin Kinsella of Avalon Ventures of La Jolla, California, has set up a $65 million fund, Poland Partners, that recently allied with Robert and Alan Potamkin, East Coast car dealers, to invest in the first of a chain of Office Depot stores in Eastern Europe.

In a paradox of the venture capital business, as more money becomes available, less of it goes to startups. That's because new companies tend to need small amounts of capital but huge amounts of handholding. The funds--Kleiner Perkins Caufield & Byers, the Mayfield Fund, New Enterprise Associates, and Oak Investment Partners, among others--have not added enough partners to be able to deal with lots of startups. As a result, many are moving in the direction of TA Associates of Boston, whose CEO, C. Kevin Landry, admits that his firm is now doing equity capital, not venture capital. "We're a long way from startups," he says.

As they increasingly compete for surer bets, the megafunds are driving up the valuations of existing companies. According to a recent report by VentureOne, a San Francisco tracker of industry data, the stock market value of businesses that had reached the stage of shipping products was 43% higher at the end of 1994 than the year before. The report sees "an ominous sign" in the rising valuations. "Simply put," it says, "venture capitalists were paying more for deals at a time when the public markets were paying less."

To Arthur Rock, a founder of Intel, venture capitalists have become portfolio managers. "They are more interested in creating wealth than in creating companies," he says. Now 68, Rock works alone with a secretary out of a 12th-floor San Francisco office with a stunning view of the bay. As trim and sharp as ever, he stubbornly continues to support startups. He says that he automatically turns down any entrepreneur who mentions how soon his proposed company could go public: "I'm not in the stock-trading business."

Of course, no venture capitalist--not even Rock--ever operated entirely out of the goodness of his heart. But some have been more devoted to company building. On the other hand, many venture capitalists now focus on an initial public offering as the finish line--and try to force new companies across it way before they're ready. Says Ted Greene, CEO of Amylin Pharmaceuticals of San Diego and a former venture capitalist: "I advise young entrepreneurs to walk out the minute a venture capitalist mentions the words 'exit strategy.' "

It would be hard to overstate the help that yesterday's entrepreneurs got from Rock and Tom Perkins, among others. Rock, Silicon Valley's original venture capitalist, spent days and nights at Intel. Perkins, who helped found Tandem Computers, Genentech, and many other companies, could be spotted zooming along Highway 101, Silicon Valley's principal artery, from one company to another in his green Ferrari. An entrepreneur himself who had founded and sold a laser company, he was legendary for his ability to juggle multiple commitments: He once served on 16 small-company boards at once.

Perkins, 62, who has retired from the venture capital firm he co-founded with Eugene Kleiner, a Silicon Valley semiconductor pioneer, has a simple philosophy: Build a solid company, and everyone benefits. That has enabled Perkins to buy mansions in California and Europe, to assemble an antique-car collection that he sold a few years ago for $60 million, and finally to sail into the sunset aboard his 154-foot, highly computerized, $15 million Pirini-Navi-built yacht, Andromeda La Dea. It would be hard to find anyone who begrudges Perkins his perks. But, says Samuel Colella of Institutional Venture Partners (IVP) of Menlo Park, "people like Tom Perkins are largely on the sidelines today."

Startups are feeling the neglect. Explains Jack C. Carsten, former Intel senior vice president for marketing and now a Silicon Valley angel: "A venture capitalist's most important commodity is time. A partner in a big fund who serves on ten company boards is not going to have the time at the end of the day to sit down with some unknown entrepreneur. If Steve Jobs walked into a megafund today, he would have no chance of getting financed."

What many of the big venture firms do instead is bet on big names--retired or otherwise available company founders--who have made money for them in the past. When M. Kenneth Oshman, who helped start ROLM (he's the "O" in the company's name), became CEO of Echelon, a Silicon Valley company that wants to digitize control of office buildings, homes, and factories, megafunds showered $25 million on the startup. Yet there is no evidence that recycled entrepreneurs can do better than fresh ones. Though Oshman hopes to hit a second home run, if you ask him to make a list of such successes, he'll tell you he can't.

In a similar move, a group of funds led by Kleiner Perkins last year gave $15 million to Steven Gillis, a co-founder of Immunex, a biotech firm that has had only one profitable year since it began in 1981, to start a new vaccine research company in Seattle. "It's a shock to me," says Sequoia's Lamond about the amount involved. "I think that's crazy." Not surprisingly, no-name entrepreneurs are beginning to look elsewhere. "Venture capital has become the oxymoron of the 1990s," says John Peers, a British-born veteran of Silicon Valley. "There's no venture and there's no capital--at least for startups." After being turned down by a dozen California venture capital firms, Peers turned to an angel, Merle Greenstein in Portland, Oregon, who had made millions selling scrap metal. With support from him and a Chicago family that asks not to be named, Peers started his company, Unilearn Inc., in Wilsonville, a Portland suburb. It is now shipping its first product, a workstation-like training device, to Boeing, among other companies, and a community college.

Some entrepreneurs are looking for alternative sources of capital for other reasons. "Venture capital is expensive," says Alan G. Walton of Oxford Bioscience Partners, a venture capital firm in Westport, Connecticut. "So if entrepreneurs can find capital elsewhere, they are better off." Immunologist Charles A. Morgan Jr. and marketing executive Warren Wheeler knew the story only too well. In the late 1980s they had worked for NeoRx, a Seattle biotech company, and they had seen how, in exchange for very little initial capital, venture capitalists can wind up owning about 80% of a biotech startup. So Morgan and Wheeler went to Canada, where in a series of deals with local investors, they bought British Columbia's largest drug distributor. Their company, Receptagen Ltd., will report sales of $60 million this year, an unheard-of performance for a biotech company only two years old. Receptagen uses its profits to finance research into drugs aimed at treating disease by regulating the natural process of cell death. And the founders kept 65% of the company.

Still, there are some hopeful signs within the traditional venture capital business. A generational change--young venture capitalists rising to replace aging founders--is helping create new funds in the two centers of high-tech venture capital, California and the Boston area. William B. Elmore, 42, a partner in Inman & Bowman in Orinda, California, is organizing a venture fund that will back startups. "The art of creating startups is in danger of being lost," he says. His partners, James Anderson, 44, and Kathryn Gould, 45, came from Merrill Pickard of Menlo Park.

In Boston two partners in the old-line firm of Burr Egan Deleage have started a firm, Polaris Ventures, to look for promising startups. So will another new firm, North Bridge Venture Partners, which deliberately set up its office alongside Route 128 in Waltham so that entrepreneurs don't have to drive into the crammed and crooked streets of downtown Boston, where the established funds are. The generational change is also bringing in talented women venture capitalists, such as Wende Sawyer Hutton, a partner of the Mayfield Fund since 1993. Hutton, 35, has an MBA from Harvard and worked for biotech startups before joining Mayfield.

These newcomers will join some traditional firms that have continued to concentrate on startups, like Matrix Partners. The firm, which operates in both Boston and Silicon Valley, may be the most successful venture firm around at the moment. Its recent funds have reported annual returns as high as 70%.

In the venture capital "ant hill" along Sand Hill Road in Menlo Park, home to some 50 firms, others have also resisted the trend to bigness. "We could have raised $100 million to $120 million but we stopped at $80 million," says James J. Bochnowski, one of the three partners in Delphi Ventures. "With a $150 million to $200 million fund, you can no longer invest efficiently at a $1 million clip; you have to invest at least $5 million at a clip." Delphi puts much of its money into first-round medical companies and does some seed financing, the earliest stage of all, in which an entrepreneur may need, say, $200,000 to get started. "In a large fund," adds Bochnowski, "the incentive begins to swing toward the maintenance fee. We chose not to go the megafund route."

Another exception on Sand Hill Road is Sanderling. The oldest U.S. biomedical fund, Sanderling has been in business since 1979; it specializes in medical-device and drug-development companies and boasts recent returns of close to 30% a year. Robert G. McNeil and his two partners deliberately limited their last fund to $50 million. Says McNeil, a molecular biologist by training: "Some people say we're weird." The partners each handle only two or three companies at a time, compared with about ten per partner at a big fund like Kleiner Perkins, and travel constantly around the country offering help to their fledgling companies. Says McNeil: "You get the biggest returns from startups if you can keep up the pace."

Even at a megafirm like Mayfield, partners insist that they continue to do startups and can show you a company being incubated on the premises: six programmers and their CEO-to-be debugging hospital management software. Adds Alan J. Patricof, chairman of New York City-based Patricof & Co. Ventures, one of the biggest funds around: "If the opportunity is big, we'll start small."

Bands of high-tech angels have been springing up in both Silicon Valley and the Boston suburbs. The diners shown on the opening pages of this story--some 20 mostly retired or semiretired high-tech executives--have been meeting monthly since February in a Palo Alto restaurant. They invest $50,000 to $100,000 on average and have already helped start three companies. The entrepreneurs get a bonus: high-powered help for free. Such luminaries as Jack Carsten of Intel and Robert Lorenzini, founder and former CEO of a silicon wafer company, serve as unpaid consultants to the new firms. Says Hans Severiens, the Ph.D. physicist and semi-retired San Francisco financier who organized the group: "In effect, we're going back to the original type of venture capital of the 1960s and 1970s."

In Cambridge, Massachusetts, an angel network is run out of MIT. The MIT Technology Capital Network holds about eight events annually where entrepreneurs meet with investors. Some of the meetings last all day, as a dozen entrepreneurs present ideas for their companies to more than 100 investors. Among the successful companies created with the help of the MIT network: Brooktrout Technology of Needham, Massachusetts, which develops software and hardware for electronic messaging, and Pixel Vision of Acton, Massachusetts.

Neil J. Rodberg, director of operations at the MIT network, says the big difference between venture capital firms and angels is that angels are more patient and are willing to settle for a smaller return--say, 20% to 25%--compared with the venture firms' expected 35% to 40%, compounded annually. Says he: "The expectations of the venture capital funds have changed. They're backing away from the zero stage [of company formation]. Many of them have said to us that if the investment is under $2 million, not to bother them. Now I hear $5 million being tossed around."

Another way startups can get seed money is directly from fast-growing high-tech companies. Cirrus Logic, a Fremont, California, maker of specialized chips, has financed three such companies recently. The SEC also allows small companies to go public inexpensively: As long as the amount to be raised is under $5 million, a company can bypass underwriters and even distribute its prospectus via the Internet. State development agencies and even federal organizations like the National Institutes of Health (NIH) now help finance startups.

But how well these surrogates work compared with old-style nurturing by venture capitalists remains to be seen. An entrepreneur in Silicon Valley or in the Boston area may have the luxury of acquiring the free services of experienced retired executives or ex-entrepreneurs. Elsewhere he is less likely to get such help.

Is America's technological advance endangered by the shifts in strategy of the big venture firms? "One of the concerns I have," says Rodberg, "is that as the venture capital firms continue to milk the second- and middle-stage companies, they are not backfilling with additional capital to start companies. Where do they think these middle-stage companies come from? It will have a definite effect, I think, on the economy and America's ability to compete in technological products." Rodberg can already see some areas receiving insufficient funding: biotechnology, environmental fields, and technologies useful to people with disabilities, such as voice-activated communications devices. In Silicon Valley, Carsten reports, probably half the 100 entrepreneurs he sees in a year never get funded.

But what's good for the country may not be good for the venture capital megafunds. Like politicians constantly thinking of reelection, the megafund managers are always thinking of raising their next fund, preferably bigger than the one before. How this will affect the future concerns some thoughtful observers. "In the past 30 years venture capital created an environment that encouraged formation of new companies in the U.S. to an extent unseen anywhere else in the world," says Jerry Feigen, a consultant on venture capital in Columbia, Maryland. "Are we going to have that environment in the next 30 years?"