(FORTUNE Magazine) – Last June, power-tool wunderkind Joseph Galli, 42, left the prosaic, tactile world at Black & Decker and beamed himself into the realm of limitless potential as president of Amazon.com. Among the attractions, no doubt, was the chance to change the world by selling people books and things online (so much cooler than selling them gadgets through stores!). But there were also the 3.9 million stock options that Amazon dangled before him--a grant that dwarfed anything B&D or rival suitor PepsiCo would pay. So what if his new salary would be a piddling $200,000? The options offered a vista of immense riches.

For a while Amazon's stock went on a classic dot-com bender. The value of Galli's as-yet-unvested options floated past $200 million, well on their way to the $1.3 billion that Amazon's 1999 proxy said the face value of Galli's options would be by 2019 if its stock rose a modest 10% a year.

Then came the hangover.

From their $113 peak last December, Amazon's shares crashed to $52. At that price, all of Galli's options submerged, carrying a paper value of zero. To be sure, Galli could still get to his $1.3 billion by 2019, but it's now a considerably riskier proposition than it seemed last summer. "Internet stocks are very volatile," says Galli, gamely. "This is all about the long term."

Galli is hardly the only old-economy executive to see the bait that lured him to the Internet get snapped up by the Nasdaq bear. Last September, George Shaheen left the top job at Andersen Consulting to head Webvan--and got a signing bonus of an epic 15 million options. By December, Webvan's stock had jumped so high that Shaheen's boodle towered $255 million over the grant price. Today it's gurgling, 15% underwater. Or how about Priceline.com's trophy hires, former Citigroup CFO Heidi Miller and the ex-chief of AT&T's consumer business, Dan Schulman? Both were welcomed to the new economy with fistfuls of options, all now soaking wet.

And, of course, for every Galli and Shaheen, hordes of less exalted employees--engineers, software designers, programmers, even secretaries--are circling water coolers from Boston to Palo Alto lamenting their shrunken options wealth. Of the 497 tech IPOs since the beginning of 1999, 69% are selling below their first-day closing price, according to Hoover's IPO Central. "Silicon Valley works on the romantic idea of getting very rich very fast," says Gabor Garai, a Boston lawyer who advises dot-coms on their pay plans. With tech stocks on the rocks, however, the romance is fading fast.

As that romance fades, so does the tech companies' ability to attract top talent willing to toil 18 hours a day in closet-sized cubicles and to bind that talent, once hired, to a new e-employer. "People who were recruited from traditional companies are starting to return," says Becky Stein, an executive recruiter at Russell Reynolds in San Francisco. "Once burned, they don't want to risk joining another dot-com, even for options at a new low price." Glenn Davis of Next Step Recruiting targets new-economy sales managers whose options have tanked: "The first question I ask is, 'What are your options worth?' If the answer is nothing, candidates you couldn't budge six months ago are often anxious to move." And some of them are demanding to take a piece of the company with them, according to Silicon Valley lawyer Garry Mathiason. A defense lawyer in employment lawsuits, Mathiason is seeing a rising tide of suits from newly option-poor workers who claim that management lied about the stock's outlook. "The ground is already trembling," he says. "This could become a full-scale earthquake."

Sure, it's tempting to argue that there's nothing amiss here that a big summer rally in Nasdaq couldn't fix. Today's disgruntled losers are simply the victims of a temporary run of bad luck.

But that misses the point. As a form of compensation, stock options are fundamentally flawed and always have been, precisely because their value is largely a matter of luck, especially for executives who plan to hang around for only a few years. Options are great at attracting and keeping employees in smoothly rising markets, but bear markets expose their weaknesses.

Those shortcomings were magnified in this downturn by the way tech companies have used options. Making huge, up-front option grants as a kind of signing bonus--a Silicon Valley standard--increases options' riskiness to employees. Spreading them liberally up and down the work force only multiplies the number of employees tempted to jump ship in bad times. Because options don't have to be expensed, earnings-challenged techs are encouraged to use them in lieu of salary to conserve cash--ignoring the hidden cost of shareholder dilution. To make matters worse, the enormous gains of the past few years have created a sense among optioned workers that they are entitled to become millionaires. Now that the party's over, companies are finding that to hang on to their best people, they have to either reprice existing options or supplement them with big new grants at today's lower prices. Both are simply ways to buy off would-be defectors by further diluting shareholders--without fixing the underlying problem.

This is not to say that tech companies should abandon options. Far from it. Used properly, they can be a superb way to motivate managers to pull for the shareholders. But before tech companies can design a proper options package, they first have to admit what they're doing wrong.

According to an old saying, each new generation of young people thinks it was the one that invented sex. It's the same with Silicon Valley and stock options. They had long been used sparingly by old-economy companies to give upper managers an incentive to build shareholder value--although compensation experts questioned how effective they really were at doing that. Then along came the cybercrowd with their dynamic new options strategies, and things got suddenly worse.

The fundamental weakness in all options is that, in the short term, most of the rise or fall in their value comes not from good or bad management but the normal ebb and flow of the stock market. Nowhere are those random ups and downs more pronounced than in tech, where stock values are hitched to ideas far more than to assets or earnings, and where investors are constantly reevaluating those ideas. Even established techs like Yahoo or Oracle lurch much more dramatically in market value than a GE or a Ford.

Old-economy companies mitigate options' risk by awarding them in comparatively small grants each year. That way executives receive a mixture of low- and high-priced options as the market weaves through its cycles, not unlike investors who use dollar-cost averaging to avoid buying all their shares at a peak. Tech companies, on the other hand, tend to award their options in less frequent supersized packages. To recruit a CEO, for example, it's de rigueur to wrap the equivalent of four years or more of annual options into one megapackage that's awarded on signing. When the market is soaring, as it was in late 1998 and 1999, those binge grants create dazzling gains that can help future recruiting efforts. But bunching your bets is always risky. "If you're lucky enough to get shares at the low point, you land a windfall," says Brian Hall, a professor at Harvard Business School. "If you get all your stock at the high point, you get nailed."

Another Silicon Valley ploy is to extend options to virtually every employee. A highly sought-after engineer, for instance, might get paid on the same model as a CEO--with a smaller salary than he would earn in aerospace, say, in exchange for a huge up-front options grant. "We want to minimize cash and get people focused on the stock," says Steve Schoch, CFO of eToys. Schoch can be pretty sure his employees are focused on the stock: It's down 92% from its high last year.

The problem is that deeply out-of-the-money options no longer motivate employees either to perform or to stay. Quite the opposite. And it's not as if the market's turmoil has cooled the competition for top talent. In fact, to rekindle their stocks, many downtrodden companies need to grow rapidly, and that means hiring a lot more people--and keeping the ones they have.

Confronted with the shortcomings of their options strategies, tech companies might be expected to look around for alternatives. But they're not. The few companies that are talking publicly about dealing with their employees' underwater options are desperately looking for ways to dial back the clock to the way things were before the Nasdaq bear arrived.

They have two alternatives. The first is to "reprice"--to cancel old out-of-the-money options and issue an equal number of new ones at the current price. Apple, Oracle, and others have used that strategy in the past when their stocks cratered; Iomega, CDnow, and Barnes&Noble.com have all repriced options that submerged in the recent selloff.

Institutional shareholders hate when companies pull this. "Their attitude is, 'I can't reprice my shares. Why should management reprice theirs?'" says Hall of Harvard. Still, if angering shareholders were the only obstacle, hundreds of companies would be rushing to reprice, says Frederic W. Cook, one of America's leading compensation consultants.

What is restraining them is a revolutionary rule the Financial Accounting Standards Board finalized in March. The FASB provision states that any in-the-money gains on repriced options become an expense. That threatens repricers with potentially huge hits to earnings and suggests that repricing will move from the tech mainstream to the margins, mainly as a refuge for troubled companies.

By far the more popular course is to make new, special grants at today's low prices without erasing the old award. Since there's no obvious repricing, the method escapes the new FASB sanctions. However, shareholders have no reason to love this strategy either, particularly if it's used when the stock's price has fallen only modestly by tech standards, say 30% or 40%. Remember, options usually don't expire for ten years or more. If the stock recovers and shoots past the strike price of the old grants, employees not only earn a fortune on the new options but also make money on their previous awards. Investors get stuck with a double dose of dilution.

A case in point is a huge new grant at Microsoft. In April, Microsoft made a special award of 70 million options to its 34,000 employees at $66.63 per share. The company argued that its people needed a lift, in part because the options they received last July in the low 90s sat almost 30% under water. The gesture could cost investors plenty. If Microsoft's stock rises to $133 in five years, an annual return of 14%, both tranches of options will move far into the money. In the past Microsoft has offset option grants by buying in shares. Assuming employees exercise half the options issued in the special grant, Microsoft would need to spend $3.7 billion to repurchase the shares. That's the equivalent of almost half its fiscal 1999 earnings.

These new grants at Microsoft and other tech companies are one more load on top of an already heavy options overhang. In new-economy companies, unexercised options average some 16% of shares outstanding, compared with 5% for traditional outfits. For every 100 shares outstanding, Qualcomm now has 19 options, Amazon 23. One way or another, shareholders pay. The companies either have to spend billions buying in shares, a la Microsoft, or dilute investors, as Amazon and Cisco do.

Not surprisingly, the prospect of extra, makeup grants riles investors. "We're already disturbed by the dilution levels, and they're rising fast," says Ken Bertsch, head of corporate governance for the giant pension fund TIAA-CREF. Adds Jack Ciesielski, a leading accounting expert: "These new grants will swell the number of options out there even more. This is really an end run around repricing."

Still, the options culture is so deeply ingrained in the tech industry--and the competition for labor is so relentless--that companies have little choice but to come up with new ways to hand employees bushels of ripe options. The question is how best to do it. By far the most expensive solution is to make new Microsoft-style grants when the old awards are still within reach. While new awards can make sense, employers should set two conditions. First, they should act only if the old options are so far out of the money that they don't motivate anybody. Second, they should parcel out the options over time, to avoid making the grant a gamble on today's prices.

That approach is working for FreeMarkets, an auction house for industrial equipment. CEO Glen Meakem watched FreeMarkets' stock careen from $280 the day of its IPO in December to around $50 now. "We hired lots of people in the $200s and $300s with big options grants," says Meakem. "We're getting hundreds of calls every day from headhunters." To keep people, Meakem made new awards in the $40s and $50s to employees who joined in, say, the $300s. That's not a bad idea; it's probable the highest-priced options will never be in the money. For new employees he's offering small upfront grants and the potential to earn additional yearly grants based on performance. Pay experts applaud. "I recommend making grants every quarter," says Cook. "That really smooths out the volatility."

While companies figure out how best to deploy options as recruiting and retention tools, they might also refocus their strategy on options' original goal, which is to reward superior performance. Alfred Rappaport, professor emeritus at Northwestern's Kellogg Graduate School of Management, thinks the selloff presents a great opportunity to do just that. He suggests replacing underwater awards with new packages of indexed options. In these instruments the strike price rises with either the S&P 500 or a basket of stocks in the same sector. To reap a big payoff from indexed options, managers have to do more than show up at work during a bull market; they have to beat their benchmark. If the market crashes, the executive still makes money as long as the stock falls less than that of his rivals--and the company never has to make unhappy employees whole. Says Rappaport: "This takes the tails-you-win, heads-I-lose element out of options."

Believe it nor not, one tech company does feature an indexed option plan, probably the best pay package in America. This paragon is Level 3, an operator of fiber-optic phone networks. For executives to ring the bell in Level 3's plan, the stock has to beat the S&P by a wide margin. For example, if the stock nips the index by a mere five points, CEO Jim Crowe's options go $1 million in the money. If he waxes the S&P by ten points or more, he gets three times what executives would pocket from plain-vanilla plans. Last year Level 3 beat the S&P by 70 points, 90% to 20%; Crowe's options are in the money by $121 million. Shareholders aren't complaining. They simply pay him out of their own outsized gain.

Unfortunately, it's not likely that Silicon Valley will follow suit anytime soon. The morning after the options party, tech companies remain obsessed with preserving their homegrown brand of options package. Sure, it's self-defeating. Sure, it's perverse. But they invented it. Like sex.