(FORTUNE Magazine) – Not long ago Warren Buffett considered buying a big stake in a company that no doubt would have died to win over the Oracle of Omaha. But a close look made him uneasy: The company had an addiction to easy-money stock options. "Once they vested, all the options would have taken 10% of the earning power of the enterprise," he complains. And the binge showed no signs of abating. Every year the company issued another big batch of options, enriching managers and gouging shareholders. The options blizzard proved to be the terminator, the main reason America's shrewdest investor didn't add the company's name to Berkshire Hathaway's honor roll of holdings.

How did a good idea go so wrong? Standard stock options were supposed to tie pay to performance by rewarding CEOs only if they could drive up stock prices. But instead of attracting the Warren Buffetts of the world, they've become anathema to big investors. Even so, such investors aren't calling for the end of options. On the contrary, many are enthralled by a new breed of CEO that is using creative, demanding options packages that put the shareholder, not the CEO, first.

For Buffett, standard options set the bar too low, making it easy for CEOs to get rich by being average, or in this raging bull market, worse than average. The compensation of the average CEO has almost doubled since 1992, to $8.4 million last year, according to a survey of FORTUNE 200 companies by Pearl Meyer & Partners. And $4.6 million of that, a full 55% of the total, was in option grants, dwarfing salaries and bonuses. For Buffett, many of those big awards just aren't deserved. "There is no question in my mind that mediocre CEOs are getting incredibly overpaid. And the way it's being done is through stock options."

Buffett is particularly riled by a bookkeeping quirk that companies cherish. Under standard accounting rules, options, unlike cash, aren't counted as a compensation expense (though their value has to be determined and reported in a footnote). The accounting advantage creates an illusion, since shareholders bear the cost later in the form of dilution. But by using options, companies like the one Buffett declined to invest in can fatten their CEOs' pay packages without charging a dime to earnings. "I unequivocally regard the special accounting treatment given options as improper and deceptive," he says.

In addition, boards are to blame for establishing a double standard. Instead of using options as incentive pay, they claim that to stay competitive they must give out at least the same dollar value in options as their rivals, regardless of how well the CEO performs. CEOs naturally applaud the practice. The trend is driven not so much by other companies' pursuing CEOs with higher offers as by surveys. Boards hire compensation consultants--something Buffett says he wouldn't dream of doing--to gauge the average CEO grant. Following the consultants' advice, the board raises its CEO's award to at least the industry median or even higher--who wants to admit its CEO is subpar? That lofts awards into an ever-rising spiral. Performance gets overlooked in the process.

The big awards exacerbate another bad feature: incredibly easy terms. Most CEOs receive standard, "at the money" options, meaning if the market price is $50 the day they're issued, the strike price is also $50. And it stays at $50 for the entire ten-year term of the options. Getting rich doesn't require superior management at all. As Buffett points out, if the CEO buys government bonds with the company's earnings instead of paying them out in dividends, the book value over time will rise and the share price will probably bump up as well. With a million options, a caretaker CEO would make a killing. But that makes as much sense, says Buffett, as paying someone a fat commission for letting interest build up in a savings account. "These plans are really a royalty on the passage of time," he says. How about Buffett investments like Walt Disney, where Michael Eisner gets a slew of at-the-money options? Says Buffett: "With people who got huge awards and deserved huge awards, the result was right. But it doesn't mean their options were properly constructed."

In a great stock market, gains are an outrageous no-brainer. "You can be very mediocre, but when interest rates for everyone drop sharply and you don't improve your business at all, you make an enormous amount of money," says Buffett. "No one should be rewarded for that." Even scrawny ducks that can't swim or quack, he quips, rise in a swollen pond.

The rushing waters are rewarding many a leaden performance, at shareholders' expense. Example: PepsiCo's Roger Enrico has made $17 million since 1996 on a grant of 1.864 million stock options, while Pepsi has given shareholders only a 48% return, 25 percentage points less than the S&P 500. But a new breed of boss is rejecting cushy, sedan-chair plans. CEO pay is acquiring real stretch targets, the kind that inspire the troops to shrink cycle times, scale new productivity peaks, and fast-track the Viagras and Pentiums from lab to market. These CEOs are betting their paychecks on what moves investors: leaps in share price that beat the market.

Believe it or not, such action heroes are actually pushing their boards to make plans more demanding. They range from FORTUNE 500 CEOs such as Monsanto's Robert Shapiro and Transamerica's Frank Herringer to fire-eating comers led by telecom entrepreneur James Q. Crowe of Level 3. Right now, on-the-edge plans are relatively rare, though they're spreading fast--so if you're a CEO with a cushy deal, watch out. Institutional investors love the stretch option packages and want you to have one too. "These high-hurdle plans are the way to go," says Eugene Vesell, a money manager with Oppenheimer Capital, whose funds hold over $60 billion in securities. "The normal plans are almost outrageous. For CEOs, they're 'Heads I win, tails you lose.'" Adds Robert Boldt, who helps invest $140 billion for the California Public Employees' Retirement System: "When options have lots of stretch in them, companies like Monsanto will do whatever it takes to reach the target."

For Vesell and Boldt, the attraction of a tough pay plan is basic: The CEO believes so strongly in his story that he volunteers to give shareholders big gains before taking some of the gravy for himself. Powder-puff plans inspire suspicion, not confidence. Investors worry that the CEO is more interested in gaming the pay system than in outperforming the market.

The big innovation is putting teeth in options in the form of tough performance hurdles. The idea is simple: The CEO must substantially raise the stock price, in a tight time period, before he can make big money. Buffett likes these "out of the money" options as much as he despises many standard plans. He heartily approves of the one for President Alan Spoon at the Washington Post Co., where Buffett is a director and major shareholder.

Plans use all kinds of targets. Some of the best plans demand gains equal to those of the S&P or a basket of similar stocks, minus two things: dividends, and one or two percentage points, for making the CEO put most of his future wealth in a single security. A typical annual hurdle figure is 8% to 10%. Only by reaching or beating the targets does the CEO get to share in the gains.

The plans fall into two categories. Citicorp and Du Pont, for example, favor "performance vesting" options (though, unlike standard options, they have to be expensed). As with standard options, the CEO gains the entire appreciation over the market price for the life of the options. Under Citicorp's 1998 plan for CEO John Reed and over 50 other executives, that was $121 a share, and Citicorp gave Reed 300,000 options at that price. But there's a catch. Reed and the other executives get to exercise their options--meaning they vest--only when and if the stock reaches a much higher target price. In Citicorp's case, that's $200. And forget the plush ten-year term. For Citicorp, it's only five. Hence, if Reed were to raise the bank's stock price only 10% per annum, to $194 a share by 2003, he'd forfeit all his options. On the other hand, if he only barely makes it, he gets the entire windfall of $23.7 million. For Reed, it's a real nail-biter. The plan will be rolled into Citigroup stock, after Citicorp merges with Travelers.

For shareholders, the "premium priced" options championed by Monsanto, Transamerica, Ecolab, and Colgate-Palmolive are even better. At Colgate-Palmolive, CEO Reuben Mark received a megagrant of 2.6 million premium-priced options that will replace annual option awards for seven years. Again, the CEO must hit a lofty target price. But once there, he still hasn't made any money. He keeps only the gain above the target price. This year, Monsanto installed an aggressive premium-priced plan covering CEO Robert Shapiro and 31 other executives, who must increase the stock price 50%--from $50.22 on the day of the grant to $75.33 by 2003--before their options are "in the money."

But Shapiro and the other executives have to pay for their options, so they must raise the share price even higher before they can start cashing in. Monsanto allows Shapiro and his lieutenants to plow up to half their salaries over the next two years into options. All elected to participate. Shapiro is spending the maximum, $800,000, on 132,000 options at $6.06 each, half their cost to the company using the Black-Scholes model for valuing stock options. If the stock rises to only $75.33, there is no gain to cover Shapiro's $800,000 investment. He must give shareholders 10.5% returns per annum over five years, driving the share price to $81.39 (the $75.33 strike price plus his $6.06 investment), before he can start making money.

For cocky CEOs, these plans have an edge: even bigger money for super results. That's because they can get far more options, usually in huge blocks every three years or so, than yearly at-the-money grants provide. Later this year, Monsanto will give Shapiro and the other executives a second award with the same $75.33 strike price, though this time he won't have to put up his own money. (Because Monsanto's plan involves premium-priced options, not performance vesting, it doesn't have to expense them.) Since premium-priced options start way "out of the money," their present value is far below that of at-the-money options. Monsanto's premium options run $12.12 each, compared with the $20 or more it estimates for standard ones. In other words, for the same estimated cost to shareholders, Monsanto will be able to hand Shapiro perhaps 70% more options.

Shapiro is using the high-hurdle plan to create a culture that never lets up. Last summer he spun off Monsanto's chemical business to focus on life sciences: breakthrough drugs and revolutionary biotech products for agriculture. Building a whole new industry on genomics to protect crops is a far cry from turning out polyester, and it takes a different kind of manager. "The old culture rewarded longevity," says Shapiro. "I want people with a fanatical, obsessive devotion to moving new products through the pipeline." For Shapiro, at-the-money options lead to a cautious, caretaker management that's "playing defense." Says Shapiro: "I'll take my chances with the shareholders."

For small companies, it's natural to take a big gamble on pay: They can grow a lot faster than big ones, though they can collapse a lot faster too. Just a few years ago, Cooper Cos., a maker of contact lenses and gynecological devices, lay in shambles. Its CEO went to prison for insider trading, and its stock price collapsed, sending its market cap to $30 million. To the rescue came Tom Bender, now 58, an irrepressible veteran of SmithKline Beecham. Surveying the wreckage, Bender quickly decided that plain-vanilla options weren't the solution. "They're meaningless," he snips. To rouse the troops, Bender and the board put in a premium option plan of Olympics-level difficulty.

For the troops as well as himself, Bender set a series of rising stock-price bogeys with very short deadlines. Every time he hit one, a tranche of his 168,000-option grant would move into the money. The first was the toughest: raising the stock price from $11.75 to $16 in six months, a 36% jump. Bender made it. He also hit all the other goals, including the last one: raising the $11.75 price 189%, to $34, by the year 2000. With Cooper now trading at $39.50, Bender rang the bell 19 months early. "We kicked butt all the way up the line!" he boasts. Cooper now has a market cap of $600 million; its revenues are growing at 40% a year.

Premium-priced plans are fine in good or average markets. But what if stocks drop 20% and stay depressed? Most options would wind up underwater, and the premium ones, with the highest strike prices, could gurgle to the bottom. A Bender or Shapiro might perform heroically, raising his stock price 5% a year and beating the S&P, only to see his options drown. "The big danger is that good managers would leave because they aren't getting paid in a bad market," says Dan Ryterband, a consultant with Frederic W. Cook & Co.

Asolution is waiting in the wings: indexing options. The idea is that when the S&P rises or falls, the strike price moves up or down with the index. Says Steve O'Byrne, a consultant with Stern Stewart & Co.: "Indexing isolates the contribution of management from the fluctuations of the market." Right now, just one company has the guts to do it. Believe it or not, Level 3, a telecom startup, boasts the best CEO pay plan in America. Its CEO, James Crowe, already has had one brilliant experience with indexing, at MFS, a local phone company he started in 1989 as a branch of Peter Kiewit Sons', the big construction company in Omaha. Crowe and other executives made fortunes on options, but only after providing investors with returns far above the S&P. In fact, MFS had one of the greatest rides in corporate history. Just three years after it went public, in 1993, Crowe sold it to WorldCom for $14.3 billion.

Crowe is counting on an MFS-style pay plan to power Level 3. His goal: building a fiber-optic network connecting 60 cities. The pay plan can make executives fabulously rich, but only if they way outperform the S&P. The payouts start small, then explode as the stock price outpaces the index. Crowe gets nothing if the index rises 10% and Level 3's stock does the same. Even if he outperforms it by five percentage points, his 1998 options will be worth only $1.55 million in three years. Then the numbers really take off. If Crowe beats the S&P by 15 percentage points, those options rise to $11 million in value. Still, shareholders--who have already rewarded Level 3 with a market cap of $9 billion--love it, because they get to keep the lion's share of the gains.

The battle royal will come with the next long slump. Many CEOs will push to reprice their options, though they never volunteer to make their options more expensive in a bull market. "They will fall back on their situational ethics," promises Warren Buffett. But the CEOs with the grit to stick to premium options, or better still, indexing, will manage better and win shareholders' respect. To pick tomorrow's winners, follow the pay plans.

INSIDE: Becoming CEO? Call him first, page 281... Buried treasure, page 285... The Leading Edge, page 289... Not Eisner's pay stub, page 294... Ask Annie, page 296