(FORTUNE Magazine) – This past month has been a busy one for a couple of revered Boston institutions. In early May the Boston Celtics, one of the most storied franchises in professional sports, proudly announced the hiring of Rick Pitino as its new head coach and president. To clear the way for Pitino, former coach M.L. Carr was kicked upstairs, the assistant coaches were all fired, and a half-dozen members of the front office were given their walking papers. The moves spoke to a harsh but easily understood reality: The team had fallen on hard times, and only by making major changes could the franchise ever hope to recapture its past glory.

A short walk from the Celtics offices, a second Boston institution, Fidelity Investments, was also completing a round of major changes. Fidelity, the mutual fund behemoth run by Edward (Ned) Johnson III, had during the course of this 15-year bull market become perhaps the dominant financial services institution in the world, but it too had fallen on hard times. An old-line, family-owned company, Fidelity had grown spectacularly in the 1980s on the dazzling returns of its brand-name mutual funds, starting with Peter Lynch's Magellan fund. But Fidelity has watched those same funds lag the market in recent years. From 1994 through 1996 only four of 34 diversified Fidelity U.S. stock funds beat the S&P 500, according to Morningstar; so far in 1997, only one is ahead. In addition, at a company where fund managers rarely left, they have suddenly been leaving in droves--more than 20 departures in the past 16 months. The portfolio shop has been adrift and in turmoil, something Fidelity has long tried to hide and refuses to acknowledge to this day. This past winter, as it became clear that investors were losing faith in Fidelity, the company held a series of "road shows" in an effort to prove that all was well at the good ship Fidelity. But that just served to spark a new round of publicity about the company's woes.

So it was hardly a surprise when, in the space of two weeks in late April, Ned Johnson finally made some decisive moves. He first removed a high-ranking executive named Gary Burkhead as the head of the investment management arm of Fidelity and replaced him with Robert Pozen, Fidelity's long-time top lawyer and a Johnson confidant. Then Pozen named three deputies, including Johnson's 35-year-old daughter, Abigail--cementing the widely held view that she's being groomed to take over the company. And finally, Johnson named another close aide, James Curvey, to assume the position of chief operating officer, a job that Johnson hadn't filled in over a decade.

Unlike the Celtics, however, Fidelity did not hold a press conference to trumpet these announcements. Instead, in typically secretive Fidelity style, everything was assiduously downplayed. The Burkhead move was spun as a promotion, since he now would have "senior responsibility"--whatever that means--for Fidelity's growing institutional business. And in a circumspect interview with FORTUNE, Johnson brushed away suggestions that the changes were aimed at stanching the blood before things got any worse. On the contrary: He claimed the changes at the top were merely "evolutionary"--in fact, he said, the shakeup reflected nothing more than his desire to give three of his top executives, all of whom had held the same jobs for a while, something new to do. In other words, no big deal.

But it is a big deal, for three reasons. First, although Fidelity is a privately held company, it has become something very much like a public trust. It now holds more than $500 billion in assets from rich and poor, old and young, sophisticates and neophytes. Some 11 million people invest in one or more of Fidelity Investments' 250 mutual funds. And many of them have no choice in the matter: Increasingly, people have to invest there because their employer has chosen a Fidelity-sponsored 401(k) retirement plan. In addition, with those hundreds of billions in assets at its disposal, Fidelity has become the single most important player on Wall Street and a stockholder of immense power, with its tentacles in thousands of companies.

Second, it's a big deal because when a secretive company like Fidelity makes a major move, it has a way of setting off alarms. "If the performance deteriorates more," warned a late-April release by David O'Leary of Alpha Equity Research, Fidelity's most unrelenting critic, "Fidelity could become the Drexel Burnham of the 1990s."

Which bring us to the third and most important reason. The announcements are a grudging, long-overdue acknowledgement that for one of the few times in its charmed life, Fidelity has basic problems it can't ignore. It is not just bad luck or a tough market that has caused Fidelity's chronic underperformance these past few years. Rather, it's been a combination of factors, many of them self-inflicted: funds that grew too large, fund managers who left the firm in frustration, and the corrosion of the culture that made Fidelity's portfolio shop so special for so long. Indeed, as FORTUNE has learned, for a time in the early 1990s some Fidelity fund managers seemed to care almost as much about their personal trading accounts as about the multibillion dollar portfolios entrusted to their care.

What's more, while this was happening, Fidelity's core customer was changing--something the company took far too long to recognize. In the old days--the days of Peter Lynch, glittering fund performance, and pedal-to-the-metal marketing--the prototypical Fidelity customer was an individual investor reaching for the kind of breathtaking returns that only Fidelity seemed consistently able to deliver. As a former fund manager puts it, "Fidelity used to be 1-800-MAKE-ME-SOME-MONEY." Today, Fidelity's most important customer is a corporate administrator charged with choosing a 401(k) retirement plan for company employees. This is a person far less interested in having Fidelity take the kinds of outsize risks that were famously part of the Fidelity style. Johnson's moves are partly in response to this new reality: Fidelity simply can't run money the way it used to.

Fidelity remains an extremely powerful franchise, of course. Contrary to O'Leary's doomsaying, the company is not on the brink of collapse, or anything close to it. Virtually every competitor FORTUNE spoke to said they still feared Fidelity more than any other fund company. Fidelity, they said, still had its share of advantages: a well-known brand name and an absolute willingness to forgo profits to gain market share. It is still, by far, the largest mutual fund company in the world, with nearly twice as many assets as perennial No. 2 Vanguard. It has immense resources at its disposal, with annual revenues of over $5 billion and reported earnings of $423 million. Most competitors predicted that Johnson would get the company back on track. "They will turn this around," says a top executive at one of Fidelity's archrivals.

In one critical way, however, Fidelity is still not ready to change. It remains an extraordinarily arrogant institution, seemingly oblivious to the idea that managing a half-trillion dollars of America's money carries with it any special need for candor. To a man, the top Fidelity executives interviewed for this article denied that the company was facing any problems or even that it was changing in any real way. And they made little effort to hide their annoyance at having to answer questions. As Ned Johnson put it upon first being introduced to a FORTUNE reporter: "Are you people so f--ing bored that you have nothing better to write about than us?"

The 66-year-old chairman of Fidelity is scrunched in a chair at the head of the table in his private dining room. (One insider jokingly suggests "it's the best restaurant in Boston"; he's not far off.) Reportedly worth more than $1 billion, Ned Johnson wears a plain green suit and his trademark aviator-style glasses. He is flanked by Curvey, who ran Fidelity's business development arm before becoming COO, and a FORTUNE reporter, who is attempting to question Johnson about the changes at the top. More often than not it is Curvey, a forceful 61-year-old executive, who answers for Johnson. ("Would you mind, Ned?" "No, go ahead.") But no matter which man is talking, most of their responses are pat ones. In answering a question about why so many fund managers have left, for instance, Johnson first claims that the number of departures is minuscule given the 450 or so professionals who work in the portfolio area. Then he says that the real problem is that hedge funds are paying astronomical sums, which Fidelity can't hope to match. "We have the best group of fund managers that we've ever had," he insists.

At one point Johnson is asked about the reason behind Fidelity's recent move into some new lines of business--selling proprietary customer service software to big companies, for instance, and breaking into the health care benefits-processing and payroll record-keeping businesses. "The prime reason," Johnson replies, "is that customers need to have somebody who really cares about their self-interest in processing their business. And who cares more than we do?"

It is perhaps the most revealing answer of the entire lunch, for it shows that Johnson knows very well who his real customer is these days. It's not the retail investor. It's the hundreds of businesses that use Fidelity-sponsored 401(k) plans.

Since the early 1990s, when Fidelity began its full-frontal assault on the 401(k) market, that new customer has been the key to the company's asset growth. Fidelity now commands more than 15% of the 401(k) market, almost double Vanguard's share. Indeed, some 54% of Fidelity's assets under management constitute retirement money (which includes IRA accounts as well as 401(k)s). Once a fund complex lands a big 401(k) account--a Ford or a General Motors (both of which use Fidelity-sponsored plans, as does Time Warner, the parent company of FORTUNE's publisher)--it gains, overnight, thousands of new individual fundholders and tens of millions of dollars in new assets. What's more, money keeps arriving automatically, month after month, like clockwork. It is the mutual fund equivalent of nirvana.

That also explains why Fidelity's asset base has continued to grow over the past few years despite its troubles. Last year, for instance--a year that saw a big reshuffling of the fund managers in March, followed by the exit of Jeffrey Vinik, the controversial manager of the flagship Magellan fund, and a stream of other prominent Fidelity hands--the company still managed to increase its assets under management by some $90 billion, from $405 billion to $496 billion. (By April of this year, assets under management had risen to $511 billion.)

But despite that rise, last year was also the first time this decade that Fidelity did not increase its market share in the 401(k) business. As retirement plan consultant Peter Starr of Cerulli Associates puts it, "The 800-pound gorilla is showing signs of losing weight." Plainly, Fidelity's troubles have made the job of landing new 401(k) accounts more difficult. Says a former executive at hometown rival Putnam: "Putnam is beating them more often than they used to." Adds a former Fidelity executive: "Their 401(k) business is strong, but it's not invulnerable. Consumers are fickle today. The portability of money--even 401(k) money--is a fact of life."

Which gets to the real crux of the matter. The 401(k) market runs on a different set of assumptions than the retail market. Corporate administrators are not nearly as interested in out-of-the-park returns as they are in predictability; after all, they have a fiduciary responsibility to company employees. Over the past few years, it has gradually become apparent that many of Fidelity's best-known funds aren't exactly without risk. And as this awareness has grown, rival fund companies have tried to use that fact to cut into Fidelity's 401(k) business--even as Fidelity has attempted to mollify its 401(k) customers by reining in its funds. Once you understand that, everything else that's been going on at Fidelity suddenly makes a whole lot of sense.

Throughout the period when Fidelity was shifting its focus away from retail investors and toward 401(k) plan sponsors, Gary Burkhead was running its investment management arm. He had moved into that position in 1986, following a successful stint as head of Fidelity's trust company. Back then, Peter Lynch, who was still running Magellan, set the tone. Lynch helped mold Fidelity's fanatical desire for high-octane performance and its belief in the power of individual stock research. Lynch was willing to listen to any Fidelity analyst, no matter how green, who thought he had a good investment idea. He believed that you beat the market by picking good stocks, one at a time, rather than by trying to time the market. And he was nothing if not flexible. "You were encouraged to be creative," recalls a former fund manager. "If you were running your standard, run-of-the-mill equity income fund or growth fund or whatever, you always looked for a bit of a wrinkle--some way to beat the competition--maybe buy some foreign stocks, that sort of thing." Regardless of what a fund's name might imply, Fidelity's fund prospectuses included wide investment parameters, and as long as you won your bet, nobody asked questions. Beating the market was what mattered.

Burkhead's influence grew after Lynch stepped down from Magellan in 1990. Burkhead was, to put it bluntly, a bureaucrat--polite, serene, noncommittal, impossible to read, a native Arkansan who had buried his accent long ago. Inside Fidelity he was known for two things: his natty, English-cut doubled-breasted suits and his fierce loyalty to Ned Johnson. The latter quality is an important one for any top Fidelity executive. You had to know how to "handle" Johnson, and Burkhead had the knack. It was not always an easy task. At times Johnson has been known to launch into profane diatribes and table-pounding outbursts. "You have to wear a thick suit of armor when you meet with him," says a former Fidelity senior executive--who, like virtually everyone else FORTUNE contacted for this story, asked that his name not be used. (Curvey replies: "Give me a break. He's a gentleman, for Christ's sake.")

Smooth as he was, Burkhead did not have extensive fund management experience--and it showed. An infamous early incident took place in the late 1980s, soon after a former Lynch assistant named Rich Fenton took over the Puritan fund, a balanced fund that held both stocks and bonds. Fenton, concerned about the fund's stake in junk bonds, wanted to sell the position. But according to several of Fenton's former colleagues, Burkhead was not anxious to see Puritan lose its attractive yield and pressured Fenton to hold on. Not long afterward, junk bonds sank, and Puritan suffered. Fenton and the other fund managers seethed. (Responds Burkhead: "Never have I told fund managers which stocks or bonds to buy or sell.")

Still, it seemed a momentary blip. Even after Lynch retired, Fidelity's portfolio shop continued to operate on the same set of principles--and with the same great results. In 1991, 67% of Fidelity's diversified U.S. stock funds beat the S&P 500, the next year it was 78% and in 1993 an amazing 83%. As the company began to make serious inroads in the 401(k) market, those results dazzled corporate administrators, just as Fidelity had earlier dazzled retail mutual fund customers. Nobody worried about Fidelity's funds being too risky for retirement money, not so long as they were beating the market. On the contrary: Employees clamored for the chance to invest. Says one rival: "Fidelity had that roaring performance, and nothing else seemed to matter." Assets under management rose $34 billion in 1992 and another $69 billion in 1993. The biggest player in the business just kept getting bigger.

Yet as Fidelity's funds grew, sustaining top performance became more difficult. This is something Fidelity has obstinately refused to concede, but it's obvious. George Vanderheiden, perhaps the most highly respected fund manager at Fidelity--and Lynch's last real contemporary among Fidelity portfolio managers--now manages some $35 billion in nine different funds. (His flagship fund is the $16 billion Advisor Growth Opportunities fund.) Beginning in late 1995, Vanderheiden says, he made a $500 million bet on semiconductor equipment stocks, which he believed, correctly, had bottomed out. Before long most of the stocks doubled--yet the impact on Vanderheiden's funds was barely discernible. "You can still hit a home run," he says. "The problem is that even when you do, it doesn't show up as much."

The size issue exacerbated another problem: In an effort to boost performance, the Fidelity portfolio shop began taking ever bigger risks--with Gary Burkhead, ironically, leading the charge. The example that practically every ex-fund manager interviewed by FORTUNE points to was Burkhead's decision to ramp up Fidelity's fixed-income department in the early 1990s. Bond funds had long been an afterthought at Fidelity mainly because Ned Johnson never had much use for bonds--"certificates of conscription," he used to call them. But Burkhead believed that Fidelity could take its traditional strengths as a stock picker--superior research, ample resources, and a flair for creative risk taking--and apply them to bonds.

At first Burkhead's strategy worked spectacularly. Under Tom Steffanci, whom Burkhead hired to lead the revolution, Fidelity's bond shop racked up soaring returns in 1992 and 1993, attracting a flood of new assets. Other fund managers, relying on the research team Steffanci had assembled, began investing in such exotic fixed-income vehicles as South American Brady bonds and Mexican Tesobonos. The quintessential example was Fidelity Asset Manager, a fairly new fund marketed to lower-risk retirement-oriented investors. Thanks to manager Robert Beckwitt's willingness to use all the leeway in the fund's prospectus, Asset Manager rode international bets and derivatives to terrific gains.

But in 1994, everything changed. First there was a sequence of interest rate hikes. Then, in December, Mexico devalued the peso. As the Tesobonos crashed, so did Fidelity's supercharged funds.

It would be hard to overstate the impact of the bond fiasco on Fidelity--or on its 401(k) competitors. For one thing, it gave Burkhead a new appreciation for risk. He had tried to be the hero and had become the goat instead. He quickly reversed course, letting Steffanci and his team depart, and reeling in other managers like Beckwitt.

The bond fund blowup also gave Fidelity's competitors their first real ammunition in the battle to land 401(k) plans. The public exposure of Fidelity's misadventures with exotic assets meant Fidelity now had a "truth in labeling" problem, and the fluctuations of a widows-and-orphans fund like Asset Manager made it easy to argue that the folks at Fidelity were just too wild for retirement investors.

Those were critical business issues for Fidelity, and Burkhead's instinct to clamp down was the right one. But the way he went about it could not have been more wrong-headed. For starters, he refused to call a duck a duck. He was putting a shorter leash on fund managers, all the while denying he was doing any such thing. Fidelity entered a state of institutional denial, and it drove the fund managers crazy. Some were constitutionally incapable of operating under the new set of assumptions; others who might have made the adjustment had they only been dealt with directly became disgusted by the lack of candor.

That Burkhead's actions tended to be reactive, and clumsy to boot, only made matters worse. Though he disputes the characterization, others say he unilaterally began removing whole asset classes and investment strategies from fund managers' tool boxes, often at exactly the wrong time. Emerging market debt, for example, was all but banned from Fidelity bond funds in early 1995--and over the next two years, it was one of the few asset classes in the world to outperform the S&P 500. After a while fund managers began treating Burkhead's directives as contrarian indicators. "Once there was a problem in the market that was pervasive enough that Burkhead actually found out about it and said, 'You guys can't do that anymore,' almost by definition it was the bottom of the market," says a former Fidelity hand.

But Burkhead's biggest mistake of all may have been his failure to rein in one fund manager in particular: Jeff Vinik.

When Vinik was named to run Magellan in 1992, there was hardly an uproar. He was the most talented fund manager in the shop--not the brightest necessarily, say his colleagues, but the one with the best feel for the market. Still, Magellan wasn't just any fund. As the company's flagship, it was the one product all employers wanted in their 401(k) roster and the one every newcomer to mutual funds wanted a piece of. It was huge--$56 billion at its peak last spring--and it was powerful, and the manager who ran it set the tone for the company, both in the public eye and internally.

Vinik was not like Peter Lynch. Though he trained under the master, Vinik never shared Lynch's absolute faith in company-by-company analysis. Instead, he was always looking to catch big market trends. He was also consumed by arcane technical factors like price momentum and trading volume. ("Nobody pays more attention to [stock] charts than Jeff," says a former colleague.) One result was that Vinik often seemed only marginally interested in what analysts had to tell him, often brushing them away before they had a chance to pitch him their latest investment idea. And because of the stature Magellan's manager holds within Fidelity, Vinik's style undermined the foundation of Fidelity's advantage as an investment house--superior research by a corps of driven, motivated analysts.

The other impact Vinik had on Fidelity's culture was in the way he bought and sold securities for his personal account. Fidelity's attitude toward personal accounts had always been quite accepting (provided of course that no laws were broken). Lynch himself liked to buy and sell stocks on the side--he thought it kept a fund manager sharp--and his example encouraged others. The presumption was that part of what made you Fidelity material was a passion for the market that went beyond what you did for a living--it was personal.

Vinik was one of a cadre of managers who inherited this Fidelity tradition and took it to the limit. Hallway discussions among these fund managers were as likely to be about the stocks in their personal portfolios as what their funds were buying. And Vinik was more obsessed than most: "Jeff was a madman," says a former fund manager. In the space of one year Vinik made more than 500 trades in his own account, according to the Washington Post. (Vinik had no comment on specific details of his trading activities. He said through a spokesman that he "always conducted his personal trading activities in compliance with Fidelity's code of ethics.") Asked whether personal trading became a distraction at Fidelity, George Vanderheiden said, "It might have, for some people." He added that he never heard Vinik talk about his personal trading. (Pozen responds: "The small percentage of people that engaged in personal trading did it primarily on their free time.")

The environment changed in 1994 after the SEC began investigating a fund manager at the Invesco fund complex for alleged improper personal trading. While that episode had nothing to do with Fidelity, it spawned an industrywide reevaluation of personal-trading rules. At Fidelity, new restrictions were imposed on managers' trading practices for their own accounts. "Since the end of 1994, there hasn't even been anything to look at," says Pozen, who as Fidelity general counsel implemented the new procedures. Discussions in the hallway about personal portfolios subsided, and the volume of trading dropped dramatically.

But even after this tightening up, Vinik remained an avid personal trader, as did a small group of cohorts. Several former analysts say that they felt career advancement sometimes depended in part on how willing they were to supply certain fund managers with ideas earmarked for their personal accounts. (Evaluations by fund managers are one component of Fidelity's analyst review process.) "I didn't like it, but they expected you to feed them tips," says one former Fidelity analyst who left within the past year. ("That's absolutely untrue," counters Pozen. As for Vinik, he adds, "he followed the rules.")

And then there was the fallout from Vinik's decision to publicly praise Micron Technology in the fall of 1995 at the same time that he was selling the stock at Magellan. It burst into a full-fledged scandal when the fund's trading records were leaked to the Washington Post. All at once, Fidelity recognized that there was yet another threat to the fund's vaunted reputation. The company reacted by launching an internal witch-hunt to find the source of the Micron leak. Soon a sense of paranoia among managers and analysts was layered on top of the existing turmoil. Recent Fidelity departees contend phone and fax logs came under scrutiny. One tells of being quizzed by his boss about who the source of an unattributed quote in a press account might be.

But the single Vinik shortcoming that hit Fidelity's bottom line most harshly was this: As a portfolio manager, he failed to live up to his reputation. In early 1996, after staking Magellan's future on a turn in interest rates by salting away nearly 20% of the assets in bonds, Vinik watched the fund's three-year return record drop below that of the S&P 500. Magellan's management fees fluctuate based on performance, and after returns dropped below the market over three years, the fees fell--from 0.73% to 0.47%, which adds up to a $130 million decline on a $50 billion portfolio. Fidelity should have seen it coming. As one former fund manager puts it, "Do you want to have the center of your franchise run by a guy who has a history of making big, ballsy bets?"

Fidelity's 401(k) competition suddenly had a new weapon in their arsenal: They had Magellan itself. Putnam, for instance, hired a consulting firm to draw up a series of risk/reward charts contrasting Putnam's funds with Fidelity's. The Putnam funds were all clustered closely around the 45-degree angle that signals a fair tradeoff. The Fidelity funds, by contrast, were mostly north of the line, in the high-risk territory. "And the Magellan fund," says someone who's seen the Putnam presentation, "was just off the charts." The message was plain: Was it really wise for corporate pension executives to be handing over their 401(k) business to the company that offered a fund as volatile as Magellan? Magellan, which had always been Fidelity's greatest draw, had become an albatross. When, in June 1996, Vinik decided he'd had enough, there was a quiet sigh of relief inside Fidelity.

I've been in this job for ten days," declares Robert Pozen. "I know what needs to be done, and I know I'm the right man to do the job." It's early May. Pozen is sitting at a small table in his new office at One Federal Street--a corner office in the area where the portfolio managers work. Just an hour or so ago, the press release was drawn up announcing his new team of assistants, including Abby Johnson.

He is, at first blush, as unlike Gary Burkhead as a person could be: straightforward, without a bureaucratic bone in his body. "People keep focusing on the fact that Bob [Pozen] is a lawyer," says a Fidelity competitor. "But historically, he's been Ned's fixer. He goes in and solves problems." The man adds, "He's there to get that unruly animal back in the stable."

He is also there, plainly, to undo some of the damage of the past year and a half: not just l'affaire Vinik but the incredible reshuffling of managers (Burkhead changed the managers of 27 funds in March 1996, for instance), the steady stream of departures, and the relentless drumbeat of publicity over the continued underperformance. (In 1996 only five of Fidelity's 36 diversified U.S. stock funds beat the S&P 500.) So adrift was the portfolio shop that Burkhead had set up two small committees of fund managers and asked them to grapple with the questions "How to win?" and "What does it mean to win?" Not long ago those were questions you never had to ask at Fidelity.

So what, exactly, does Pozen plan to do about all this? Nothing that requires rocket science, it would seem--just simple, basic stuff. "We're going to emphasize stability and continuity," he says. "When we hire people, we're going to have an expectation that they'll stay in the same research job for three years. That will make them much more valuable. And we're not going to be shifting eight people every time someone gets a promotion. We are going to try to minimize change."

His other ideas are along the same lines: Beef up research. Eliminate administrative roles for fund managers such as Vanderheiden, allowing them to concentrate on picking stocks. Give more fund managers stock in the company to keep hedge funds at bay. Tell everybody what the goals are and what the rules are.

Already he's averted one potential catastrophe. In the days before he was offered the job, rumors were swirling that two more managers, Will Danoff and Steven Kaye, were about to leave. It would have been devastating, especially since both are considered steady, reliable performers with solid long-term track records--precisely the kind of managers who appeal to corporate administrators. After Pozen's appointment, the rumors stopped. "I've had long talks with both of them," Pozen says. "They've both told me they're staying."

Finally, Pozen is saying up front that fund managers can't stray from the "style box" of their particular fund. "The emphasis will still be on performance here. We are performance oriented and always will be. But the objectives will reflect how the fund is doing within its style parameters." In typical Fidelity fashion, he denies that this constitutes a big change, but it does. It means that fund managers can no longer "goose" performance by straying into any asset class that suits their fancy. What's really going on here is that Fidelity is trying to put the "truth-in-labeling" bogeyman to rest--and trying to get its risk/reward ratios back into the comfort zone of its most prized customer, the corporate administrator.

The big question, of course, is, Will it work? Can Fidelity play this new game as well as it used to be able to play the old game? Certainly, Pozen has some advantages. For one thing, he's got Robert Stansky running Magellan. (To give credit where it's due, Stansky was a Burkhead choice.) Stansky--who was Peter Lynch's chief protege his last few years at Magellan--is a first-rate stock picker, with a much steadier hand than Vinik's. And the culture of personal trading is long gone; most of the big traders have left the firm. Undoubtedly, there will be further departures, as old Fidelity hands decide the new game is not for them. But there's no reason Pozen can't build a new and healthy culture in the portfolio shop--one that reflects the company's new realities.

Still, it won't be easy. Fidelity has problems it still hasn't addressed. One, many of the funds are still too big. Top management refuses to concede what every fund manager knows in his bones: that size is the enemy of performance; that you can't manage tens of billions the same way you manage tens of millions and expect to have the same success. As assets continue to accumulate, that problem only gets worse.

Two, Fidelity remains a house divided. Pozen has no more natural affinity with the fund managers--and even less investment experience--than Burkhead did. Like Burkhead, Pozen insists that's irrelevant. "This job requires a manager," he says, "not an investment professional." Maybe, but that kind of talk makes fund managers deeply uneasy, even suspicious.

Finally, Fidelity, as always, would have us believe that it can do it all: deliver great returns no matter how many assets it brings in; reassure cautious 401(k) investors while also exciting performance junkies; and trade forever on the legacy of Magellan, even as it leaves that legacy behind. That's asking a lot.

Additional reporting by Nelson D. Schwartz

REPORTER ASSOCIATES Maria Atanasov, Amy R. Kover, and Jeanne C. Lee