(FORTUNE Magazine) – Among the T-shirted techies who cram the dingy offices of NetGravity, a San Mateo, Calif., software startup, Stephen Recht stands out. He wears a tie. He is 45. He has a wife and a son and a house in the staid (for California) suburb of Piedmont. He is, for crying out loud, a CPA.

But when a headhunter approached him last fall about joining NetGravity as chief financial officer, Recht was as prepared to take the plunge as any twentysomething computer jock. For him, it wasn't a plunge at all--this is his third go-round in a row as CFO of a Silicon Valley startup.

Recht says he likes "the personal satisfaction of helping build a company." But he's also a financial guy, and financial guys tend to be frank about money. "I'm betting that one of these companies does well enough that I get a big chunk," he says.

How big? When Recht joined NetGravity, he was granted options to buy somewhere between 1% and 2% (he won't say exactly) of the company's stock at a low pre-IPO price. If all goes according to plan, and NetGravity goes public in about nine months, and it grows within a year or two to a respectable but not spectacular market capitalization of, say, $200 million, he stands to clear more than $2 million.

Working hard and hitting it big is the American dream, and on the flatlands bordering the southern half of San Francisco Bay, that dream has been coming true amazingly often. From Intel in the 1960s to Apple in the 1970s to Silicon Graphics in the 1980s to Netscape and a host of others in the 1990s, enough Silicon Valley companies have delivered enough stock-option jackpots to enough employees to create an entirely new business culture. It is a culture in which calculations like Recht's are common, in which taking what in traditional business circles would be seen as foolhardy risks is considered the only rational way to behave.

And why not? Look at Mark Saul. A decade ago his Wall Street-bound and McKinsey-bound classmates at the Harvard Business School "thought I had a hole in my head" when he took a $47,000-a-year job in Silicon Valley; now, after stints at four startups, he's a multimillionaire at 35. Or Heather Beach, 29, who took a pay cut four years ago to join software maker Siebel Systems as its office manager; today she's sitting on an options stash worth more than $1 million.

There's more to the options culture than just a few lucky kids striking it rich. Silicon Valley wouldn't be what it is today without stock options--nor, arguably, would the U.S. stock market be quite as supercharged or the economy be buzzing along as smoothly. Stock options, and the belief that they may someday be worth a lot of money, make it far easier than it used to be to start and build a company. They give established companies a tool to motivate employees. They give businesses with already soaring stock prices an extra boost to earnings because of how options are treated on financial statements. A uniquely American phenomenon, the options culture helps explain why more than 400 public companies have sprung up in Silicon Valley, and why another 148 startups there got venture-capital backing in the first quarter of this year alone. Says Matt Ward, whose San Francisco consulting firm, WestWard Pay Strategies, designs stock plans for tech companies: "Silicon Valley is the economic engine of the world, and options are the fuel."

Corporate executives outside the Valley are catching on. Non-tech growth companies, including Home Depot and Starbucks, rely on stock options, as do long-established corporate giants such as Warner-Lambert, Citicorp, and Merrill Lynch. At General Mills, the venerable maker of Wheaties, Hamburger Helper, and Betty Crocker cake mixes, almost 10,000 employees get options, while some 800 managers are free to exchange pay raises for even bigger options grants--and most do.

All this is relatively new. Back in 1985, according to Sanford C. Bernstein & Co., the total value of shares set aside for options grants in the U.S. was about $59 billion; by 1996 the figure had grown more than tenfold, to $600 billion. Some 90% of public companies have employee stock-option programs. The most telling sign of options' growing appeal: In May, Japan passed a law allowing companies there to use options to pay employees.

At the heart of this tumult is no motivational guru or proselytizing management consultant. What makes stock options special is, essentially, an accounting glitch. Because these options don't have a cash value when they're first given out, they don't cost the company anything: They're free. Better yet, when they're exercised, the tax code allows the company to deduct as an expense the gain realized by the option holder--even though the company hasn't spent any money.

It sounds too good to be true, which is what a handful of grumpy people have been saying for some time. Their voices have generally been drowned out by the chorus of complaints over excessive options grants to CEOs, but in many ways it is this issue that is of greater import. Warren Buffett has repeatedly railed against stock options in his Berkshire Hathaway annual reports, arguing that it only makes sense to issue them to top executives who can actually be expected to influence a company's stock price (even then he isn't crazy about them). At this year's Berkshire Hathaway annual meeting, vice chairman Charlie Munger added his own biting assessment: "We regard current accounting for stock options as weak, corrupt, and contemptible."

A VIRTUOUS CYCLE?

Here is how an employee stock option works, in its simplest and most popular form: A company gives an employee the right to buy a certain number of its shares at a set price. After a vesting period of usually between one and five years, the employee can buy the shares at the option price, sell them at the market price, and take home a profit if, over those years, the stock price has gone up. (Unlike the put options and call options traded on secondary markets like the Chicago Board Options Exchange, employee stock options are not transferable.) Let's say you're given options to buy 1,000 shares of your company's stock at the current market price of $5. If the stock goes up to, say, $25, you can buy the shares for $5,000 and sell them for $25,000, netting $20,000.

Nothing particularly magical about that, until you consider what happens if a company gives out millions and millions of the things. Investing guru Benjamin Graham--Buffett's mentor--figured that out in 1936, when he wrote a parody press release (published for the first time in Berkshire Hathaway's 1990 annual report) announcing a "sweeping modernization scheme" at U.S. Steel. Among the innovations Graham jokingly proposed was paying all wages and salaries in the form of options.

"The almost incredible advantages of this new plan," Graham wrote, "are evident from the following: A. The payroll of the corporation will be entirely eliminated, a saving of $250 million per annum, based on 1935 operations.

"B. At the same time, the effective compensation of all our employees will be increased severalfold. Because of the large earnings per share to be shown on our common stock under the new methods, it is certain that the shares will command a price in the market far above the option level of $50 per share, making the readily realizable value of these option warrants greatly in excess of the present cash wages they will replace."

Graham's prescience was uncanny, as a look at Microsoft and what its treasurer, Greg Maffei, refers to as the "virtuous cycle" of options reveals. Because Microsoft's stock price has been going up so fast for so long (an average of 60% a year since its 1986 IPO) and the company is so generous with options (it gave out 57 million last year, on a split-adjusted basis, or almost 2,500 per employee), people have been more than willing to take a pay cut to go to work there. That keeps salary expenses down, which helps keep earnings high.

Then there's the part Graham didn't know about: a 1945 Supreme Court ruling that allows a company to deduct the amount employees realize on options ($20,000 in the example above) as an expense on its tax form. This serves to pump up earnings at an options mill like Microsoft even more. Doesn't giving out lots of options put downward pressure on per-share earnings, because there are more shares out there? Yes, but Microsoft takes care of that with massive share buybacks. (See box for a more detailed account.) The scenario comes full circle because these earnings boosts and buybacks help keep the stock price rising, in turn increasing the value of the options--and their appeal.

At startups like Recht's company, NetGravity, options magic works a little differently. It has less to do with accounting than with the boundless optimism born of Silicon Valley's success. In March 1995 a 28-year-old software engineer named John Danner quit his job at Silicon Graphics to start a new company, living on money from an options stash he'd built up at his previous employer, Oracle, until the new business got off the ground. In September 1995 he got $500,000 in seed money from two software company executives he knew--one from Oracle and the other from Netscape--both of whom could thank stock options for giving them that kind of money to invest. Then, using the lure of options in his promising Internet startup, Danner was able to recruit top engineers, salespeople, and marketing types--some from established tech companies like Danner's ex-employers, others from the ever-growing ranks of failed or struggling startups. The headhunter Danner uses takes options as part of his compensation too.

In these bull market days, options have their magic even at big companies where there's no chance at all of getting rich off them. That's what Dick Kovacevich, CEO of Norwest Corp., a big banking company in Minneapolis, learned after the company granted options one year ago to each full-time employee to buy 100 shares of Norwest stock at the then market price of $33. "The impact on people is far greater than I expected," Kovacevich says. "They think there's more value in the options than there really is: It's free, and it's just sexy. You get these things only the fat cats got."

HOW IT ALL STARTED

By now, stock options have become part of popular culture. They factor into Doonesbury plots; they show up in Russell Baker humor columns; they're used to help lure Jerry Seinfeld back to NBC for another year.

Part of the credit surely goes to the Class of '86--the hugely successful high-tech companies that went public that year, including Oracle, Sun Microsystems, Silicon Graphics, Adobe Systems, Informix, and, yes, Microsoft. Within a few years everybody in Silicon Valley knew someone who had made a pile of money from options at one of them. In 1992 a barrage of news coverage made the entire nation aware of the options bonanza. What set it off was an estimate by a Wall Street analyst that 2,200 of Microsoft's 11,000 employees each held options worth at least $1 million. Gaped a New York Times writer: "Not even the height of the Wall Street takeover frenzy of the mid-1980s made as many instant millionaires as did simple employment at Microsoft for the last five years."

To really understand why options mania set in when it did, you need to know where it comes from. Which takes us back to 1957, the year eight young semiconductor whizzes decided they couldn't stand working for William Shockley anymore. Shockley had helped invent the transistor at Bell Labs in 1947--a feat for which he shared a Nobel Prize. To commercialize his invention, he moved back to his native Palo Alto and founded Shockley Semiconductor Laboratories. He hired a dozen of the country's sharpest young engineers and physicists but promptly drove most of them away with his contemptuous manner, bizarre management techniques (such as posting everyone's salaries on the office bulletin board), and iffy business judgment.

The eight decided to start their own company, where they would be free to call the shots and treat one another as equals. That was not considered a conventional business plan at the time, and it took Arthur Rock--a young investment banker from Hayden Stone in New York whom the eight enlisted in their quest--35 tries before he found a company willing to give them both the money they needed and the freedom to spend it as they saw fit. Fairchild Camera & Instrument of Syosset, N.Y., put up $1.5 million, in return for the right to buy the new company--to be called Fairchild Semiconductor--for $3 million if things worked out.

"Suddenly it became apparent to people like myself, who had always assumed they would be working for a salary for the rest of their lives, that they could get some equity in a startup company," the leader at the new company, Robert Noyce, told an interviewer in 1980. "That was a great revelation--and a great motivation." After two years Fairchild did in fact buy out the founders, leaving each with $250,000 (equivalent to $1.35 million in today's dollars) in Fairchild stock.

It was then that stock options came into play. In mainstream corporate America, options at the time were viewed chiefly as a way to get tax-sheltered income to top executives. The reason: In 1950, Congress had passed a law allowing profits from stock options to be taxed at the 25% capital gains rate, provided you held on to the stock for at least a year. With the top personal income tax rate at 91%, it was an enticing loophole--but only for those wealthy enough to afford carrying the cost of the stock.

In California's nascent startup culture, however, options were seen in a different light. The idea wasn't to shelter income but to share wealth. Options became the key way to hire and keep talented executives, engineers, and salesmen. Inspired by the riches showered on the founders, Fairchild Semiconductor employees (including some of the original eight) began leaving in the early 1960s to start more companies. "Bob [Noyce] began to realize that to retain these people, he had to give them options," recalls Rock, who by this time had moved to San Francisco and become a pioneer venture capitalist. Top management at Fairchild in Syosset balked, plagued by an "East Coast mentality," Rock says. "They couldn't see clear to giving out options to people below the executive level."

Fairchild came to regret it. By 1968 the semiconductor subsidiary had lost most of its top engineers and executives to smaller rivals that were more liberal with stock and options. Noyce finally quit, accompanied by R&D chief Gordon Moore, another of the founding eight. With Rock's backing, they decided to form a company called Intel that wouldn't be stingy about granting options to employees. "We didn't include everybody, but almost everybody," says Moore, who retired in May as Intel's chairman. (Noyce died in 1989.) Other startups followed Intel's example; by 1977 it hardly raised eyebrows when Apple Computer--another startup backed by Arthur Rock--granted options to all its employees.

Ironically, outside of Silicon Valley the stock option was becoming an endangered species. Congress, at the urging of Senator Albert Gore Sr. of Tennessee, had begun in the 1960s to whittle away at options' tax-advantaged status, which some denounced as an unfair boon for the rich. In 1976 lawmakers killed the tax-sheltered option completely. While there was nothing to stop companies from granting options that were taxed as regular income, with the top tax rate still at 70% and the stock market going nowhere, most saw no point. The accounting glitch that today helps drive the options frenzy was around back then, but no one seems to have focused on it. Executives had been conditioned to think of options as tax-avoidance vehicles, period.

Still, there were stirrings of change. After the 1976 law there was no longer a tax benefit to holding on to shares once you'd exercised your options. In fact, if you dumped the shares right away, you didn't need to worry about having your money tied up in stock. Add in the fact that most people expected up to 70% of any gain to go to the IRS, and you understand why folks at the few companies that still gave out options started asking for, and getting, a whole lot more of them. Options grants made in this period tended to be, by previous standards, huge.

The explosion came in the 1980s. Congress slashed the top personal income tax rate to 50% and then to 28%. At the same time, the stock market revived from its decade-long torpor. And the few people who had kept the options faith started getting really, really rich. When Toys "R" Us emerged from bankruptcy in 1978, fully 15% of its shares were set aside for options for executives and store managers. By 1982 the stock price had risen 20-fold. CEO Charles Lazarus cashed in some of his options for $43 million; underlings W. John Devine, Norman Ricken, and Seymore Ziv pulled in another $30 million among them; even store managers found themselves with sizable windfalls.

It was the first of the great executive stock option bonanzas--and it was impossible for other executives not to notice. At Toys "R" Us, Lazarus and his cohorts made the logical next step: Options are great, so why not give them to everybody? "We thought, 'This will be a great motivational tool,' " recalls Michael Goldstein, who joined Toys "R" Us in 1983 and is now CEO. That was still too radical an idea for most companies, but issuing options to chief executives and other top managers became standard practice. It didn't hurt, of course, that there was also a movement afoot to link executive pay more closely to the fortunes of shareholders.

As options grants grew bigger and bigger, executives and directors also couldn't help but notice the wonderful things options did for the corporate income statement. Because remember: They're free.

THE RALLY IN THE VALLEY

As six o'clock news promos go, it was definitely out of the ordinary. "Coming up," the KGO-TV anchorman promised his San Francisco Bay Area viewers: "Why thousands of Silicon Valley workers are in a snit about stocks." On that day, Mar. 25, 1994, some 3,000 people, joined by Geoffrey the giraffe of Toys "R" Us, the Milpitas High School marching band, and lots of local news crews, had come by car, bus, and light rail to the San Jose Convention Center to shout, cheer, hoist placards, sign petitions, and do whatever else they could to defend their beloved stock options.

The threat that stirred them came from outside, from a group of accounting deep thinkers who had concluded that options are, in fact, not free. This was hardly a revelation. Ever since 1938, when the Securities and Exchange Commission made clear that it would leave the setting of accounting standards to the accounting profession, the American Institute of Certified Public Accountants had been wrestling with how to value stock options. The most straightforward solution--waiting until employees exercised options and counting the profits they made as a company expense (as the IRS does)--offended accountants' notions of matching expenses and income in time. In that scheme, a company might report a compensation expense in 1997, when the option was exercised, for work that had actually been done in 1988, when the option was awarded. The method AICPA settled on instead was to value options on the day they are issued, by subtracting the exercise price from the stock's current market price. If the exercise price were set at that day's market price, as was typically the case, it meant options were free.

Anybody who had ever traded other kinds of options--puts and calls on stocks, corn futures, whatever--knew this wasn't the case. While employee stock options aren't transferable on the open market, people will often pay for an option even if its future exercise price is precisely equal to the present price of the underlying security. In 1973 finance professors Fischer Black of the University of Chicago and Myron Scholes of MIT were the first to quantify plausibly the value of call options--with a formula that uses the market price of the underlying stock, the exercise price of the option, the term of the option, the risk-free interest rate (i.e., T-bills), the volatility of the stock price, and the stock's dividend yield. The Black-Scholes model transformed global finance, helping create the modern derivatives business.

By the early 1980s the accounting profession realized it had to revisit the way stock options were being treated. Responsibility for setting accounting standards had migrated from AICPA to the Financial Accounting Standards Board, an independent panel dominated by accounting-firm types but also including former Wall Street analysts and corporate executives. As is its habit, FASB took its time developing a new options standard. In 1991, with skyrocketing executive paychecks--and by extension, stock options--a hot political issue, Senator Carl Levin, a Michigan Democrat, introduced legislation urging FASB to get a move on. Still, it wasn't until 1993 that the board issued its proposal: corporations should expense options at the time they are granted, using Black-Scholes to determine their value.

The reaction from the business world was swift and fierce. Some of it focused on shortcomings of Black-Scholes. The model is best suited to valuing options with durations of months, not years. Furthermore, it relies on a variable, a stock's historical volatility, whose usefulness in predicting future value is unreliable at best.

But the opposition was also visceral: FASB was trying to kill options, and that was a terrible thing. That's what sparked the March 1994 gathering in San Jose--the "Rally in the Valley," as organizers called it. "Give stock a chance!" shouted Kathleen Brown, then California's treasurer and a gubernatorial candidate, to cheers from the crowd. "Don't stop the engine of economic growth that has absolutely fueled this California economy!"

The fear was that if options were expensed at all, companies that used them extensively would see their reported earnings drop, and that in turn would hurt their stock price. FASB's response failed to persuade. It argued that according to academic studies, changes in accounting standards don't affect stock price. "You're trying to confuse me with logic here," retorted Home Depot Chairman Bernie Marcus at one FASB public hearing. "It's not going to work. I deal with the emotional side of the street. I deal with Wall Street."

Marcus and others in the business community applied lobbying pressure to Congress. Before long, bills were introduced in both houses to strip FASB of its powers and put accounting standards in the hands of the SEC. Faced with this threat, the accounting board simply backed down. In its final ruling, FASB concluded that while expensing options using Black-Scholes or a similar pricing model was "preferred," companies could keep doing things the same old way. They just had to disclose their estimated options expense in a footnote.

WHAT HAPPENS NEXT

The footnoted options expense estimates began showing up for the first time in this year's annual reports--"immaterial" at General Electric, but $111 million at Intel (2% of profits), $43 million at Morgan Stanley (4% of profits), $570 million at Microsoft (26% of profits), and $61 million at Netscape (291% of profits). While Black-Scholes isn't perfect, the figures it generates show clearly that fiddling with how options are accounted for can have a dramatic impact on some companies' earnings. Indeed, Sanford C. Bernstein analysts Michael L. Goldstein and Jonathan Freedman estimate that the accounting benefits of increased options use have inflated corporate profits in the 1990s about 5%.

So are stock options a flawed, dangerous, artificial stimulant whose use a company could come to regret? Or are they the perfect motivator, a way to share a company's wealth among its employees to almost socialistic effect, yet with terrific implications for the capitalistic success of the enterprise?

The answer is both. Consider the case of Silicon Graphics (SGI), a leading member of the Class of '86. "We had ten years of being the new Intel in Silicon Valley, a lot of that thanks to stock options," says Richard Kramlich, a venture capitalist who provided SGI with startup funding and still sits on its board. A maker of high-end computer workstations, SGI grew from a smart idea of Stanford professor Jim Clark in 1982 to a force in the industry with a market value of roughly $8 billion in mid-1995. The company cultivated one of Silicon Valley's most aggressively stock-oriented environments, with each day's stock price posted in the lobbies of its buildings and employees sharing lavishly in its Wall Street success. But between mid-1995 and mid-1996, the company's failure to meet its ambitious growth targets and investor worries that PC makers were encroaching on its turf combined to cut its market value in half. The timing couldn't have been worse: As SGI's troubles began, Netscape, which Jim Clark had left to found, completed a stunningly successful IPO in August 1995. "There was this belief that you could leave here and join any startup, and within 12 to 18 months be rich," recalls Kenneth Coleman, SGI's senior vice president for administration. Employees left by the hundreds; in the ensuing turmoil, SGI went from simply missing its too-high targets to posting actual profit declines.

After that kind of options-accelerated tumble, you'd think SGI would swear off the stuff. But the company isn't even considering it. In the tech industry, and especially Silicon Valley, no one would work for you if you did. Starting new companies and pushing new ideas is always going to be risky, the thinking goes, so it's only proper to let the people who help build the dreams share in the rewards. And from a larger perspective, that's not such a bad thing. It's an Austrian economist's dream: Schumpeter's cycle of creation and destruction, accelerated by options to the point where it's an accepted, everyday part of doing business. For a buy-and-hold investor, this can be a nightmare--it may even mean that momentum investing, the much-maligned technique of jumping off a hot stock the moment it stops rising, is a rational response to economic reality.

Among the Norwests and General Millses of the world, though, it's unclear whether options would appeal to management were it not for their magical accounting properties. Especially when you consider the potential dilution that options represent for existing shareholders. Intel co-founder Gordon Moore likes to point out that options granted to others have helped whittle his stake in the company from 16.5% to 5.5% since Intel went public in 1971. If the business grows fast enough, of course, this isn't a problem. ("I'm not complaining," Moore notes, since that 5.5% is now worth more than $6 billion.) But with more mature companies, such growth is harder to count on.

This potential dilution is known as the options overhang: the number of shares a company has set aside for unexercised options and future options grants, expressed as a percentage of the company's current shares outstanding. The average options overhang of U.S. companies has risen from 5.1% in 1984 to 11.8% now, according to Sanford C. Bernstein. At many companies it's substantially higher. At General Mills, for instance, the overhang is 25%--more even than at Intel.

Companies often buy back stock to reduce the dilution that occurs when options are exercised. Traditionally, a buyback was a bullish sign: A company's management felt its shares were undervalued and could think of no more efficient way to deploy cash than to buy its own cheap stock. Now, however, a vast number of buybacks involve highly valued stocks--because high prices are what prompt employees to exercise their options. In effect, companies are buying expensive stock on the open market and selling it to employees for a whole lot less.

For the moment, stock repurchases--$125 billion worth in 1996, estimates Sanford C. Bernstein, up from $37 billion in 1991--are helping to push the stock market ever higher, which in turn makes outstanding stock options that much more valuable. And that in turn increases companies' potential option liability. Smart executives at well-run companies realize this, and some have found clever ways to protect themselves. Intel, for instance, began selling put warrants on its stock, which entitle the investor to sell stock to the company at a fixed price. Microsoft has added its own innovation: a $980 million convertible preferred stock offering last December. In both cases, management is using the proceeds from selling these securities to buy back shares. That's great if the company's stock keeps rising in value. If it declines significantly, however, the company loses money on the deal--presumably when it can spare it least.

Put it all together and a pattern starts to emerge. "Any upward move is magnified by options," says Morgan Stanley market strategist Byron Wien. "And any downward move as well." Microsoft executives brace themselves for the day when their stock-option escalator stalls. Says treasurer Maffei: "We think, worry, fear, and consider inevitable that there will come a time when that won't be able to continue."

The other fear is that someone will take away the benefits afforded by accounting quirks. That would hardly end options' allure or impact for startups, but it would eliminate a layer of distortion in reported earnings. FASB isn't about to try again. While some institutional investors have challenged the ever larger options grants sought by managements at annual meetings, they're generally mollified by modest reductions in the size of the grants, and they're unlikely to do much more. No one is calling into question the very idea of options.

The only real threat is posed by Senator Levin and his Republican colleague John McCain of Arizona, who are on a quest to force companies to either expense employee stock options on their income statements or stop treating them as deductible expenses on their tax forms. "No other form of compensation is treated this inconsistently," Levin says. But the bill they have proposed offers an out: Give options to all your employees, and you don't have to follow its dictates.

The options culture, it seems, isn't going away anytime soon. Just this spring, Intel--which over the years had gradually narrowed the scope of its options plan--expanded it from one-fourth of its 50,000 employees to everyone. Says treasurer Arvind Sodhani: "Eventually, every corporation in the United States will give every employee options."