(FORTUNE Magazine) – No turtles were harmed in the production of this magazine. The turtle on this page, you can plainly see, retains its dignity, even in the face of an impossibly tall obstacle. That's because it is mute and cannot make rash boasts about what it's going to do.

But executives can, and it keeps getting them into trouble. Again and again, they loudly tell the world about revenue and earnings goals they've set up--practically always the kind that consultant Ram Charan (see "Managing for the Slowdown") calls "big, hairy, audacious goals." What follows, so much of the time, are big, hairy, ignominious failures. It's as if Babe Ruth had walked to the plate during that legendary World Series game in 1932, pointed his bat at the center-field wall to predict a home run--and struck out.

Actually, it's worse: It's as if the Babe, having embarrassed himself that way once, had gone on to do the same thing in game after game. The Babe may have been crude, but he wasn't that stupid. And yet serial grandstanding about goals and targets has somehow become accepted executive behavior.

Of all the goals articulated, the most common one among good-sized companies is annual growth in earnings per share of 15%--the equivalent, you might say, of making the all-star team. With 15% growth, a company will roughly double its earnings in five years. It will almost inevitably star in the stock market, and its CEO will be given, so to speak, ticker-tape parades.

And, hey, why settle for a measly 15%? Home Depot, for example, has long aimed for EPS growth of 23% to 25%. It was sailing along at that excellent speed until October, when it announced that earnings for its third and fourth fiscal quarters would be--dreaded words--below expectations. Its stock fell by nearly 29% in one day, taking more than $32 billion of market value with it.

Stomach-turning drops like that are a reason to ask why corporations let themselves in for this punishment. Every executive breathing knows that the course of business is not smooth, not even in the best of times, much less in a rough economy like the one settling in. Why, then, do CEOs presume to state a goal? Out of hubris, of course. Out of a yen to run with the gang. Out of naivete about what's possible. And, certainly, out of some sort of feeling that it's good for their stocks. But that last point just won't stand up under examination. Talk may matter briefly, but over the long run, nothing succeeds but walking the walk. In the end, the only thing that counts is earnings.

So why don't executives just run their businesses as well as they can, report periodically on how they're doing, and return promptly to pulling the very best from their operations? Yes, even then surprises would pop. But they would not be framed in the harsh, hurtful spotlight of some overreaching expectation. And they would not as automatically damage the credibility of the players--as extravagantly set and then unmet targets are now ripping the reputation of Hewlett-Packard's new boss, Carly Fiorina.

Beyond that, the record shows that ambitious goals, often combined with incentive compensation plans that encourage unwise behavior, have time and again led corporations to "manage" earnings in unfortunate ways. Sometimes their behavior is simply uneconomical, as when they aggressively peddle cut-price merchandise at the end of a quarter, thereby stealing from full-price business down the road. Or they may use capital poorly, buying themselves earnings in the short run but creaming their returns on capital.

Worse, they may cook the books, letting their zeal for making their numbers push them over the legal line. The U.S. has had a run of high-visibility cases like that in recent years, with two big ones--Cendant and Bankers Trust--pinioning executives who pleaded guilty to false recordkeeping and reporting. (See "Lies, Damned Lies, and Managed Earnings" in the fortune.com archive.) And today we have Lucent on the grill: a case of lofty goals, thudding disappointments, a fired CEO, and a December admission by new management that revenues for the fourth quarter of 2000 had been overstated by nearly $700 million. Next came a lawsuit filed by a former Lucent executive who charges that unreachable goals induced the company to mislead the public. Lucent has said it is confident it can defend itself against the charges in court. But it has not escaped the attention of regulators. One, turning talkative recently, marveled at the scale of Lucent's overstatement. "Wow, $700 million. I think you could see some people ending up in striped suits in this one." (See "Lessons From the Lucent Debacle.")

The ultimate, pragmatic reason for not aiming at targets like 15% is the sheer difficulty--indubitable for companies of size--of growing that fast over an extended period. Take the macro environment in the U.S. as one piece of strong evidence. During a 40-year period, from 1960 to 2000, after-tax corporate profits grew at an annual rate of just over 8%. That means, obviously, that any company galloping at 15% had to do almost twice as well as the general population of companies.

There is a counterargument, though, and that has to do with the great surge in profits over the past few years. Between 1995 and 2000 (with the last year estimated) profits of the FORTUNE 500 grew at a stunning rate of about 14%. In that kind of a world, you might argue that 15% growth in profits is perfectly reasonable for the kind of superior, competition-throttling company that almost every bigtime CEO thinks of himself or herself as running. But where is that argument left when a rough year like 2001 appears and threatens to prove that economic cycles have not been wiped from the face of the earth?

Furthermore, we have been talking up to now about the difficulty of increasing earnings, not earnings per share, which is a stricter measurement by far because it wraps in the dilution that springs from the issuance of new shares. Take a growth exemplar like Cisco Systems. In the past five years its dollar profits (after all special charges and credits) have multiplied more than six times. But because it has issued a huge number of shares for options and acquisitions, its per-share earnings have grown by only four times.

That's still a five-year growth rate for per-share earnings of 34%, which is sensational. But Cisco is just now getting to be a really big company--its $19 billion in 2000 revenues will probably edge it into the top 100 in this year's soon-to-be-published FORTUNE 500 list--and its EPS growth rate is inexorably slowing. In Cisco's fiscal 2000 (ended last July), the rate was down to 16%.

That's the problem for big companies: The growing gets hard, and we have two studies to prove it. The first was done a few years ago by Wharton School professor Jeremy Siegel for his book Stocks for the Long Run. Siegel's primary purpose was to examine how the Nifty Fifty of 1972 would have treated investors who paid the sky-high prices then being asked for them and held on for 25 years--and the answer was "not badly." But a secondary part of the study looked at the group's annual growth rates in earnings per share. And only three companies out of the 50 beat 15%. They were Philip Morris, at 17.9%; McDonald's, at 17.5%; and Merck, at 15.1%.

The second study is one FORTUNE, working with Value Line, did for this article. For three different periods--1960-80, 1970-90, and 1980-99--we examined earnings-per-share growth for 150 large companies. In our sample were the 150 publicly owned companies that (a) at the start of each period were the biggest in the FORTUNE 500 or were in the very top of the "Fifties" lists that we used to do for certain industries, such as commercial banks; and (b) were still independent beings at the end of the period being studied. The fact that we threw out any company that did not last the period (because it was acquired, perhaps, or subjected to a leveraged buyout) gives the results an upward, "survivorship" bias. Beyond that, we know retrospectively that there was no shortage of business opportunity in the years we studied: Though the companies looked big to the world as each period began, they still had plenty of room to grow.

And yet the number that managed to increase their earnings per share over the periods by 15% annually was very small, even when you include the companies that hit the mark because of an oddball situation. For example, Boeing beat 15% in two periods (1960-80 and 1970-90) because it moved from hard times in the base years to prosperity in the later years. Similarly, Fannie Mae had an extraordinary 32% growth rate for the 1980-99 years because it began the period in a near-bankrupt condition, brought on by sky-high interest rates, and later got rich.

You'll see the aberrations in the chart on the facing page--and you'll also see the really strong, can't-keep-them-down earners, most especially Philip Morris, but also Merck and Abbott Labs. If you next wonder why some well-known grower like Wal-Mart isn't in the 15% group, it's because the company was too small, even in 1980, to make our list of big companies.

The uncontested leader in all this, Philip Morris, does not publicly state goals for earnings. That seems understandable: Given that the company's main product is tobacco, it is doubtful that the world would react well to its saying that it plans in the future to sell loads more of the stuff at ever higher prices and therefore crank its earnings up by more than 15% a year. But that's what it does, without articulating the goal.

Though history shows how difficult it is for large companies to hit a 15% target in earnings over any extended period, there remains no scarcity of executives willing to assume they can do it. Sometimes, of course, they tailor the standards by which they wish to be judged. Among these are two close to home: Steve Case and Gerald Levin, chairman and CEO, respectively, of the new kid on the block, AOL Time Warner, which owns FORTUNE. The company has publicized three goals: annual growth of 12% to 15% for revenues; 25% for Ebitda (earnings before interest, taxes, depreciation, and amortization); and 50% for free cash flow. But it has announced no goals for net profit or cash earnings (net profit plus amortization of certain items, such as goodwill). At the least, that omission nicely diverts attention from the small size of these bottom-line items compared with the company's market value, which was recently about $250 billion. For the first three quarters of 2000, AOL and Time Warner together had net income of only $723 million. Change the standard to cash earnings, and the figure becomes about $1.8 billion--still hardly a blockbuster amount compared with the market value, but up the ladder in respectability.

In the tailoring of goals and records, no practice is more prevalent among corporations than the ignoring of special items, most especially restructuring charges. These, a company implicitly says, are not relevant to what we've done, to the glories we've accomplished. Look at ConAgra's annual report for fiscal 2000. The first chart you will see is 20 years of what's described as "diluted earnings per share." The figures march steadily and handsomely upward, forging a 14.6% growth record that elegantly satisfies the company's goal of "double digit" growth.

Well, Terry Smith, CEO of British brokerage firm Collins Stewart and no fan of fudged data, recently reminded the readers of his newsletters that digits are fingers and that "raising two fingers," in Britain at least, has a very special meaning. ConAgra's record seems to deserve that kind of salute: If you get to the very small print beneath the chart, you find that the figures simply reflect operating earnings and do not include restructuring charges in 1996, 1999, and 2000. These totaled $1.3 billion, a huge figure in ConAgra's financial picture--and yet the amount is ignored in the company's double-digit chart. And what does ConAgra say in defense of the presentation? A spokeswoman says its approach "is reflective of the way financial analysts track and report on our company."

That is unfortunately quite accurate: Most Wall Street analysts love anything that puts a sheen on earnings. Many companies also pander generally to analysts' wishes, spending an oppressive amount of time worrying about "what the Street wants." Gillette has done that. During the 1990s, to make its goals of 15% to 20%, it also periodically engaged in "trade loading"--that is, it stuffed its distribution channels with goods at the ends of quarters to push up sales. But that game eventually stopped working, and the company began missing its targets. It threw restructuring charges into the mix, trying to clear the decks for new assaults on 15%. It also went through an episode of killing the messenger. Upon badly missing quarterly expectations on one occasion a couple of years ago, it fired its director of investor relations.

Even one of the big honchos of corporate growth, Sanford Weill, chairman and CEO of Citigroup, ignores special charges (which in his history have occurred mainly when he has acquired companies) when talking about his long-term goal of doubling earnings every five years. That kind of doubling equates to an average growth rate of just under 15%, and by Weill's reckoning the companies he has headed have substantially exceeded that mark since 1990. At Citi itself, Weill is close to pulling off another doubling act. When Citi was formed in 1998, he and his former co-chairman, John Reed, committed the company to doubling its earnings per share in its first five years, and Citi ended 2000 way ahead of schedule in meeting the goal. That's even true when all special charges are taken into account.

Weill's problem now, if it can be described as such, is that Citi's profits in 2000 came to a colossal $13.5 billion (which tops the $12.7 billion reported by the usual profit leader, General Electric). What to state now about goals? Keep talking, is Weill's first answer: "I think you have to articulate something," he said in January. "Analysts want to know what you are going to do with the company, and investors want to have a feel." Besides, he said, goals are "self-imposed pressure."

But from here--from this size--do you really want the pressure of promising to double per-share earnings in five years? Weill is indicating no. What may be "real-istic" for the future, he says, is double-digit gains, delivered consistently. Realistic that may be. But to dwell on total profits, even 10% gains over the next few years would put Citi at $21.7 billion in 2005, and that's a whale of a figure.

So Citi is one camp cutting back in goals--and there are others, though less for reasons of hugeness than humility. Under Durk Jager, Procter & Gamble had earnings growth goals of 13% to 15%; with Jager ejected, new CEO Alan Lafley is down to goals for this year of 7% to 10%. Newell used to state in every annual report that it aimed for earnings-per-share growth averaging 15% a year; then the company choked on its acquisition of Rubbermaid, and the line disappeared from print. Tiffany once claimed to be recession-proof and said it could increase profits by 15% to 25% "for the foreseeable future"; it is now hoping for 10% to 15% in 2001, and even then adds a qualifier about needing a decent economy to meet the goal.

An oddity in this environment of pulling back is that some CEOs who might logically have seized the opportunity to renounce goals have chosen not to do so. A classic example is Coca-Cola. The late Roberto Goizueta adored goals of 18% to 20% and kept on liking them even when, to make them, he had to add in capital gains from selling bottlers. The next CEO, Doug Ivester, settled on goals of 15% to 20% and didn't make them. Then came Douglas Daft, who you might think would ditch goals altogether and indeed bury any still around in that vault where they keep the Coke formula. Instead, amazingly, he announced that he would aim for 15%. With unit sales of Coke growing only by midrange single digits, that mark looks very hard to hit. In fact, the FORTUNE/Value Line study of earnings growth rates over the past 40 years produced an interesting statistic about Coke: In all three periods examined, its earnings grew at an annual rate of 12%. Mr. Daft, if you must have a goal and really want to reach it, may we suggest...

There is indeed a sense in the land that many executives just feel compelled to state goals. Carly Fiorina, CEO of Hewlett-Packard since mid-1999, gives every indication of being a charter member of that club. She, of course, came from Lucent, which came out of the womb spouting goals. She could, though, have kicked the habit at Hewlett-Packard ("legendary in its humility," she says), but didn't. Instead, only months into her job, she began announcing targets.

First, for 2000, they were revenue growth of 12% to 15% and, as she quantified it, "earnings faster than that." Then, as things went spectacularly well for HP in the early months of 2000, she went to a straight-out 15% for revenues, again with earnings "faster." Soon she added "guidance" for 2001, and it was still higher: 15% to 17% for revenues, with no target for earnings stated, though certainly "faster" was implied. Moreover, in her conversations with Wall Streeters, Fiorina was highly confident. To at least one who questioned the necessity of publicly advertising 15%, she said, "I wouldn't be promising that if I wasn't sure that it was in the bag."

And then came Nov. 13, when Fiorina was obliged to tell a shocked world that even HP's earnings for its fiscal fourth quarter (ended just two weeks before) weren't in the bag. Instead, she said, they would fall short of expectations by a few cents. The company lost $23 billion in market value in three days. The drop was almost certainly not the result of the pennies missed: Despite the fourth-quarter shortfall, HP's earnings per share for the year exceeded 1999's by 16%. No, the drop had to do with Fiorina's credibility, which in one terrible moment took a body blow. "If you go back ten or 15 years," says Samuel B. Jones Jr., chief investment officer of Trillium Asset Management in Boston, "investors might have excused a CEO for being that out of touch with the numbers. But not in today's world, when everybody knows that the technology exists for managements to keep totally posted. Carly Fiorina really looked bad on that one."

And, unbelievably, worse was still to come--though not before a dose of cheer. That was administered on Dec. 6, when Fiorina, apologizing once again for the miss on fourth-quarter data, nevertheless reaffirmed that she was still expecting 2001 revenues to grow by 15% to 17%, assuming that the economy made no worse than a "soft landing." That was the last of the optimism. Just over a month later, on Jan. 11, Fiorina lowered her guidance for the first quarter of 2001 to single digits--"low to mid"--and said uncertainties would prevent HP from updating its guidance for the full year.

In the midst of these zigs and zags--transpiring in just over two months--Fiorina came to FORTUNE, on Nov. 28, to talk to the editorial staff. To a skeptical question about why she needed to articulate goals, she said, "Well, I think there are very few companies in this day and age that get away with not providing guidance." She thought the SEC's new Regulation FD (for "fair disclosure"), which bars companies from providing important information to some investors and not to others, put a special obligation on companies to be "open and public" with whatever information they choose to provide. She didn't think they could clam up about a quarter, just saying, "Trust us."

But what if you trot out those marvelous goals, saying "Trust us," and you don't make them? What's the trust equation there? Nasty, right? Besides that, Fiorina's comments about Reg FD suggest it is something it is not. Nothing in the rule requires companies to publicly declare goals, issue "guidance," or for that matter talk to analysts at all. The rule simply says that if you're going to be wising up analysts and major shareholders, you'd better be wising up the general run of investors as well.

It may be time for a Martha Stewart lesson. Around 30 years ago she was a Wall Street broker (can you imagine being a client of Martha's? I can't), and when she took her company public in 1999 and did road shows to sell her stock, she called on her broker experience whenever she could. In one road show, an attendee reports, a questioner from the audience wanted to know whether she was going to be one of those CEOs who set big goals and then miss them. "I know where you're coming from with that question," answered Martha sympathetically. "I've been where you are. And what you want, I'm sure, is for companies to underpromise and overdeliver." Now, that would be refreshing. "A good thing," as Martha would say.

REPORTER ASSOCIATES Julie Schlosser, Jessica Sung, Matthew Boyle, Patricia Neering

FEEDBACK: cloomis@fortunemail.com