(MONEY Magazine) – Fidelity Magellan is the Coca-Cola of mutual funds. It has the name recognition, the marketing clout and the gargantuan size. Investors have long had an emotional attachment to it, and it occupies an equally sentimental place in Fidelity's history; after all, the flagship fund, founded in 1963, was first managed by Fidelity's chairman Ned Johnson and, in its heyday, was run by Peter Lynch, whose remarkable record--an average of 29% a year for 13 years--trounced the market and pretty much everyone else.

But none of those things are helping Magellan today. Since Bob Stansky, a highly regarded Fidelity manager who is now 48 years old, took over as manager in June 1996, the fund has delivered an average annual return of 7.8% (through Feb. 20), according to Morningstar, trailing the S&P 500-stock index's 8.9% annualized return. Last year the fund trailed the S&P by four percentage points--that's not horrible (total return: 24.8%), but it's disappointing for investors expecting top-of-the-charts performance.

Indeed, ever since Lynch left in 1990, Magellan has been riding on its outsize reputation while producing lackluster results. On the basis of its 10-year return, Magellan ranks 14th among the 15 U.S. equity funds with assets of $20 billion or more, according to Lipper. For both the past five years and the past three, it ranks 13th.

What happened to this American icon? Is it simply going through a rough patch, or is there a more fundamental problem? And if you're one of the 5 million or so shareholders who have entrusted your retirement to Bob Stansky and Fidelity's worldwide crew of stock researchers, should you consider bailing out?

In late February I flew to Boston to talk with Stansky--who rarely gives interviews--about the fund's middling performance and to ask why shareholders should continue to stick with him. He arrives at a conference room wearing a basic blue shirt and carrying the "sheets," printouts that contain reams of data on all the stocks he might want to own. Smart, up front and likable, he says that he still believes he can beat the market. His goal remains the same as when MONEY last met with him two years ago: to outperform the S&P 500 by three to five percentage points a year. "How do I grade myself?" he says. "Well, certainly, the past 18 to 24 months I can't give myself a good grade because the performance hasn't been what I hoped it would be."

The $68 billion question

It's easy to see where Stansky went wrong in the past year or two. Worried that tech valuations were too high and that earnings would not be as strong as others anticipated, he kept a lid on his tech holdings. At the end of 2003, the fund had 15% of its assets in tech vs. the sector's 18% weighting in the S&P 500. In particular, he had smaller stakes in bellwethers like Intel and Cisco. Those stocks outperformed the market by a wide margin in 2003. "I was clearly wrong," he says. Compounding the problem, Stansky bet on large pharmaceutical companies, including Merck, which was a top 10 holding in 2003 and fell 14%. Stansky's winners--including Citigroup, Tyco, Morgan Stanley and Best Buy--could not offset his wrong call on tech.

While that mistake seems to explain Magellan's recent weakness, it doesn't address the longer-term malaise. Since 1990, under three different managers, the fund has essentially tracked the S&P 500--giving investors indexlike returns while charging for active management. And that inevitably raises the issue of size. When Peter Lynch left Magellan in 1990, the fund had $14 billion in assets; at its peak, in the bubble days of 1999, it surpassed $100 billion and became the world's largest mutual fund, an event the press celebrated much as it did the Dow's surpassing of 10,000. (Although it closed to new investors in 1997, the fund remains open to employees of companies that offer it as a choice in their 401(k) plans.)

Steering the battleship

Today, even shrunken to $68 billion and fallen to the No. 4 spot among stock funds--behind not just Vanguard 500 Index but also two offerings from American Funds--Magellan remains so big that it's hard not to think of it as simply a proxy for the market. With so much money in the fund, if Stansky were to put $10 million in a promising small stock, the investment would have essentially no impact on the fund's return no matter what the stock did. That means Stansky must concentrate on huge companies in which he can invest millions of dollars--on average, he plows $290 million into each of his picks, and his largest stakes run above $2 billion. As a result, his fund is largely confined to big, familiar companies. His top holdings consist of blue-chip giants such as Citigroup, AIG, Viacom, General Electric and Microsoft.

Lynch, by contrast, bought whatever his research led him to believe would go up--small, obscure companies, foreign stocks, bonds, whatever--which increased his odds of finding the kinds of huge winners that help a fund outperform.

Stansky concedes that the fund's size means that he is shopping for companies largely among the S&P 200 (the largest 200 companies in the S&P 500), which range today from General Electric (with a market cap of $336 billion) down to Burlington Northern Santa Fe ($12 billion). "The size of the fund does dictate to some degree what size companies you need to buy to have significant positions," Stansky says. "Does that make it harder? I don't think so. There's plenty of action going on in those [S&P 200] stocks. You get plenty of variability there."

Stansky--and Fidelity--vehemently deny the longstanding criticism that Magellan is little more than an index fund. For example, Fidelity points out that at the end of last year, 58 of Magellan's 233 picks were not in the S&P 500, while 325 companies that were in the index were not in the fund. "I in no way feel it is an index fund," Stansky says. "I am focused on which stocks in the index can outperform." He points to media conglomerate Viacom, a clunker last year that was the fund's third-largest holding at year-end despite its position as No. 28 in the S&P 500, as an example of his decision-making. He cites his own recent underperformance as extra proof that he is not an indexer. "Do the numbers suggest an indexlike return?" Stansky muses ruefully. "Well, I wish it were the past two years, so something went wrong there." True enough. But the fund's "R-squared," a statistical measure of how closely the fund tracks the market, is, in fact, 0.99--just a smidge below Fidelity's own Spartan 500 Index fund. In plain English, when Magellan looks in the mirror, the S&P 500 index looks back.

The book on bob

Bob Stansky is a local boy who made good. The son of a furnace and air-conditioning serviceman and a housewife, Stansky grew up in working-class Worcester, Mass. He received a degree in accounting from Nichols College in Dudley, Mass. and a master's in business from New York University. He once explained that he found his calling when he bought a fiber-optics stock called Valtec Industries for $9 a share and it soared to $60. ("I said, 'This could be fun.'") Arriving at Fidelity in 1983, Stansky soon got a plum job working as an assistant to Peter Lynch on Magellan. (Was he hoping he'd someday get to run Magellan? "No," says Stansky. "I was thinking, What will help me get a few of these stocks right so I get to stay here?")

Stansky was soon running his own small sector fund, Select Defense & Aerospace. In 1987 he was handed Fidelity Growth Company, then a $206 million fund, where he could buy smaller stocks, invest in initial public offerings and trade easily. During his nine years at the helm the fund posted 16% average annual returns, beating the index by three percentage points a year, and garnered $8.5 billion in assets. Perhaps the best bet of his entire career occurred there: He invested in Cisco, beginning with its 1990 initial public offering, when it could be had for just pennies a share (adjusting for splits), and continued to trade the stock profitably.

In 1996, Stansky was tapped to lead Magellan, which has proved to be a much tougher challenge than Growth Company. But size isn't the only constraint on Stansky's performance. Stansky must also operate within Fidelity's fairly conservative guidelines for Magellan. In part, that's thanks to his predecessor, Jeff Vinik, who managed the fund from mid-1992 to mid-1996. Vinik ran the fund in a freewheeling style; he resigned after making a big, seemingly ill-timed bet on bonds.

Partly in response to that nervy call, Fidelity began imposing tighter controls on many of its big funds. The cowboy culture that had produced Peter Lynch no longer seemed appropriate for a company getting most of its new business from 401(k) investors. Managers were encouraged to seek consistent returns rather than to take big risks in pursuit of chart-topping numbers. At Magellan, the biggest and most visible Fidelity fund, the need to keep an even keel was paramount.

"It's not as if Bob Stansky does not have the right stock-picking stuff," says Jim Lowell, who edits the Fidelity Investor newsletter. "But you take someone like Stansky who had dramatic out performance at his prior fund, which was much more flexible, and place him at Magellan, where there are spoken or unspoken constraints that hobble his skills, and there have to be days when he is just knocking his head against Magellan's wall." Concludes Lowell: "If you are hobbled by the risks that you can take, eventually you are just fighting a battle you cannot win."

Paying the price

Some Magellan investors are getting antsy. Over Stansky's tenure, Magellan has had net outflows of $20.3 billion, according to Financial Research Corp. And last year Magellan experienced net outflows of $2.4 billion, even as investors were rushing to get back into equity funds.

To put those numbers in perspective, a surge of outflows came in the immediate aftermath of the Vinik episode, and the fund, as noted earlier, has been closed to new investors since 1997. But the combination of shrinking assets and sluggish performance has been costly for Fidelity. The fund carries an annual expense ratio of 0.72% (well below average). Fidelity gets up to a 0.2% bonus in periods when the fund tops the S&P 500 for 36 months; it forfeits up to that same 0.2% when it lags the index. In the fiscal year that ended March 31, 2001, Magellan earned $710 million for Fidelity. Last year that figure was down to $344 million.

With both fund holders and the company suffering from Magellan's slump, the obvious question is, Will Fidelity make any changes? Fidelity points out that over Stansky's tenure, Magellan has outperformed 69% of other large-cap core funds (according to Lipper). Stansky's boss, Abby Johnson, president of Fidelity Management & Research, gives him her full backing. "He has, and always has had, my complete confidence and support," says Johnson. "Bob and I are comfortable with the fund's size, and we have no current plans to change its structure."

At the same time, some observers believe that Fidelity may be giving Stansky and managers of other big funds a bit more leeway to take risks in search of higher returns. "Abby has loosened the reins over the past year or two," says Eric Kobren, publisher of the newsletter Fidelity Insight. Stansky "has more ability to make those bets now," he adds.

No one wants to see Magellan's numbers improve more than Stansky himself. Not only is his reputation riding on the fund, he has what he calls a "significant" amount of his own and his family's money in Magellan. "The only person I hear pressure from is myself," Stansky says--not from Magellan's board of directors, not from the Johnson family. "And no one could possibly give me as much pressure every day of saying, 'Come on, this has got to go in the right direction, let's do what we need to do.'"

Buy, hold or sell?

Roughly one in every 10 households with any money at all in mutual funds owns shares in Fidelity Magellan. "It's one of those funds that may have some sentimental value for a lot of people because of its association with Peter Lynch and its flagship status," says Roy Weitz, founder of Fund Alarm, an irreverent online guide to the fund industry.

Bob and Ruth Sessler, retirees in New Providence, N.J., have owned Magellan for more than 20 years--investing some $8,000 in an IRA, worth about $65,000 today. "It has been good to us," says Bob Sessler, who's now 73. "I can't complain about it. Maybe it's not doing as well as when Peter Lynch was running it. But that's okay, that's life."

That seems sensible, but you may feel differently. Here's our advice to current and prospective investors.

IF YOU OWN MAGELLAN IN A TAXABLE ACCOUNT. If you're one of the million or so people who has owned Magellan in a taxable account since before it closed its doors to new investors in 1997, moving your money could lead to a big capital-gains tax bill. In that case, you may want to leave your money in the fund while directing new investments elsewhere.

IF YOU OWN MAGELLAN IN A RETIREMENT ACCOUNT. If, like more than three-quarters of Magellan investors, you hold your shares in a retirement account, you needn't worry about taxes, so it makes sense to consider shifting some or all of your stake to other funds that invest in large growth companies. (Similarly, IF YOU DON'T ALREADY OWN MAGELLAN, but have access via your company's 401(k) plan, don't be tempted by the fact that you can "sneak in.") Chances are your plan has better alternatives, including many in the Fidelity family. Investors who prefer an index approach could switch to Fidelity Spartan 500 Index. Those who want active management have a number of more nimble core offerings to choose from. Among large-company funds, there's Fidelity Dividend Growth (for more conservative investors) or Stansky's old fund, Fidelity Growth Company, and Fidelity Contrafund (for more aggressive ones). Among funds that invest in a wider range of stocks, two good choices are Fidelity Capital Appreciation and Fidelity Low-Priced Stock. "Fidelity has a baker's dozen of better large-cap growth funds," says Fidelity watcher Lowell. "Why anyone would recommend Magellan is beyond me." That's just the question Fidelity needs to answer.