To borrow a description from Sir Winston, the American Funds family is a riddle wrapped in a mystery inside an enigma. The firm manages assets greater than the annual economic output of Australia, representing the savings of some 50 million shareholders, and its name is a touchstone for financial advisers everywhere. But despite its omnipresent role in the nation's retirement savings, the family maintains a veil of secrecy over its inner workings and the brain trust that runs the funds.

But if money equals power, it hasn't corrupted Capital Research & Management, the shy advisory company that runs these funds. Capital's people today work in the same manner that they have for nearly eight decades -- head down and nose to the grindstone. That's how American has delivered the results and garnered the cash to become the largest mutual fund family in the world, with more than $1.1 trillion in assets under management.

The American Funds' consistently superior track record is virtually unparalleled. All 11 of the firm's stock funds with ten-year histories have above-average total returns in their categories for the past decade. A majority of American's stock funds have landed in the top 25% among their peers over the past three years.

Driving that performance is a philosophy of deep research and investing for the long term. That may seem like common sense, but Capital Research's innovation has been to let multiple managers of each fund run portions of assets independently -- mimicking the nimbleness and efficiency of much smaller funds while working on a supersize scale.

Please, no publicity

Few of American's millions of shareholders recognize the name Capital Research & Management, which is exactly how the firm's leadership likes it. You could drive by the company's posh Los Angeles headquarters, in a 55-story skyscraper, and never know it. The building's owners offered to put Capital Research's name on the structure when it leased the top five floors, but tru to character, the firm declined the free publicity (today the "Bank of America" logo appears on the building's exterior).

Capital's roots trace to an Alabama architect named Jonathan Bell Lovelace, who made a bundle in stocks in the 1920s and deftly exited the market before the 1929 crash. In 1931, Lovelace founded a firm to provide financial advice to California companies. His firm soon took over several exchange-traded closed-end funds, including Investment Company of America, which Capital converted to a regular mutual fund in 1934 and still runs today.

Capital's defining moment came in the late 1950s, when it developed the multiple-manager approach. Because assets are divided among managers, no one person handles more than, say, $10 billion to $20 billion, and each can pursue his or her best ideas without having to persuade the whole team. Plus, a single person's gaffes or windfalls can't dominate a fund's results. "The fund shareholder gets a much more stable ride because of the lending of different approaches," says Dale Hanks, a Capital vice-president.

Without that innovation, Capital probably couldn't manage the amount of money it does in so few funds. Measured by total mutual fund assets, Vanguard is a close second, with a hair more than $1.1 trillion, and Fidelity ranks third, with $903 billion. But Vanguard runs more than 100 funds, the largest of which are tied to indexes. Fidelity spreads its assets among nearly 200 funds, with several of the largest being closed to new investors.

By contrast, American runs just 27 funds, 15 devoted entirely or partially to stocks, plus nine target-date retirement funds, which invest in other American funds. Although American manages ten of the nation's 17 biggest stock funds, including the largest -- Growth Fund of America, with $185 billion in assets -- the firm has never closed a fund to new money and says it has no plans to do so. Instead, whenever a fund's assets swell to gargantuan levels, Capital adds another manager to the fund's roster. That has helped it avoid the problems that tend to come with outsized assets, such as managers running out of ideas for new money or being unable to invest in smaller companies.

The brain trust

A relentless focus on hiring is also critical to Capital's stellar results. The firm increases its staff of investment decision makers by about 8% each year. It now employs 105 analysts and 68 portfolio managers across offices in London, Geneva, Tokyo, Hong Kong, Singapore and five U.S. locations.

Selective doesn't begin to describe the hiring process, which can easily last a year. Dozens of one-on-one interviews gauge a candidate's intelligence and personality -- Capital abhors the "star manager" culture and courts modest types. It's a coveted employer in the industry, and employees rarely leave; on average, Capital's portfolio managers have been with the company for 24 years.

Once they make the payroll, professionals are expected to be bold and creative. Managers are encouraged to stick to their guns when stocks they like get pounded by the market. All analysts run money alongside portfolio managers, so many remain analysts and cover the same companies for decades. "I'd say we know our companies better than anyone else," says Gregg Ireland, a manager for Growth Fund of America and American New Perspective.

Why not close?

Concerns about asset growth haven't stemmed the enthusiasm of advisers and brokers, among whom the American funds are wildly popular. "American Funds is truly a world-class organization," says Mark Schlafly, president of FSC Securities. "It's a group you want to be associated with." Michael Ward, an adviser with LPL Financial, says the funds' consistency and low expenses are typically the top reasons advisers love them. On average, the Class A shares of American's diversified domestic-stock funds charge 0.61% in annual expenses -- less than half the average for all diversified U.S.-stock funds.

Criticism of the shop's refusal to close funds is nothing new. But some fund watchers think this time is different, particularly because assets have ballooned by more than 200% over the past seven years. Morningstar analyst Greg Carlson, for one, thinks that endlessly adding new managers may eventually lose its effectiveness. Already, for example, 11 managers, plus a group of analysts, pick stocks for Growth Fund of America.