INVESTORS HAVE TO MAKE some big decisions before buying a mutual fund -- and picking the right flavor of fund is only one of them. They must also decide how to buy it: through a broker or directly from the fund.

Investors tend to split evenly on that question, half going with brokers, half going direct. But if a recent academic article is right, you might want to think twice before calling a broker.

"The bulk of evidence fails to identify tangible advantages of the broker channel," concludes the working paper, written by Daniel Bergstresser, an assistant finance professor at the Harvard Business School; John Chalmers, associate professor of finance at the University of Oregon's Lundquist College of Business; and Peter Tufano, a professor of financial management at Harvard.

In an interview, Chalmers emphasized that the purpose of the paper was not to discredit brokers who put their clients into mutual funds.

"We admit there are likely to be benefits to using a broker to buy mutual funds," says Chalmers. "However, with one exception, these benefits don't show up in the characteristics we're able to measure."

In fact, funds sold through brokers performed a shade worse in the study. The direct funds trailed their benchmarks by 1.070%, on average each month after expenses, while the broker-sold funds lagged by 2.282%. (And that's not counting the sales charges, or loads, customers pay for brokered funds.)

Why the performance gap? For one thing, the expenses on broker funds tend to be higher, to cover the costs of distribution and advice. But there was also evidence that the broker funds underperformed somewhat worse even before such expenses are factored in. That raises the possibility that the universe of broker-sold funds simply isn't as good.

Of course, there are cases where broker-sold funds have superior investment performance. Exhibit A would be American Funds, whose portfolios continue to rack up good returns, despite massive inflows from the broker channel this year. A prominent example of the direct channel is the Vanguard Group, which relies heavily on index funds and keeping expenses down.

Chalmers surmises that one reason retail investors turn to brokers to buy funds is "a deep-seated desire to have their hand held in making these decisions."

Or, "It may be that these investors are too busy and they outsource it to their brokers," he says.

Brian Reid, deputy chief economist at the Investment Company Institute, the fund industry's trade association, characterizes the document as "a very preliminary look at comparing those funds in those two channels."

In his view, the slightly lagging performance of broker funds in the study stems from "brokers providing an additional service -- and they have to be compensated for that."

Admittedly, distinguishing between direct funds, which can be acquired via the Web or over the phone with the help of a company sales rep, and brokered funds is not always easy. Funds in the study's direct category include some 401(k) offerings.

Following up on the research that went into their paper, titled "Assessing the Costs and Benefits of Brokers in the Mutual Fund Industry," the professors are "considering how we might measure some of these more psychological benefits, or immeasurable benefits, that brokers must be providing," Chalmers explains.

Both the broker and direct channels each had about $2 trillion under management in 2002, according to the study. Analyzing roughly 5,000 funds, the paper studies the performance of funds in both distribution channels from 1996-2002.

The study examines five possible benefits offered by the brokers, such as whether brokers can help their customers locate funds that are harder to find and analyze. Other possible broker benefits: putting their clients into funds with lower expenses, distribution costs like sales loads notwithstanding; access to better performing funds; better asset allocation; and steering clients away from "behavioral investor biases," such as chasing the latest hot fund category.

The study did identify several pluses for the brokers. "Some evidence suggests that the broker channel sells funds that broadly could be characterized as hard to find and analyze," the authors write. (That would include funds that aren't rated by Morningstar.)

What's more, the authors found some evidence that "home bias" -- that is, an emphasis on U.S. funds -- "is less pronounced in funds sold in the brokered channel."

Compared to buying funds directly, however, the broker route didn't fare so well in other areas. For example, consumers using brokers "pay extra distribution fees to buy funds with higher non-distribution fees expenses," the authors observe.

And: "The funds they buy [in the broker channel] underperform those in the direct channel even before deductions of any distribution-related expenses."

What's more, there is no evidence of better asset allocation among the broker-sold funds and the funds they put clients into "exhibit substantially greater trend-chasing behavior," the authors observe.

Still, the authors stop short of deeming one channel better than the other. "In the end, costs and benefits of using the brokered channel can only be judged by the consumers that make the decisions."

Concludes Chalmers: "Our paper shows that the service the broker provides is probably not helping you pick a mutual fund that will have better performance. We speculate that the benefit is more likely to be psychological or the time-saving benefit you would achieve from not having to make these decisions."

Fido in the Doghouse

Fidelity, the nation's largest mutual-fund company, steered clear of the recent market-timing and late-trading scandals that rocked the industry. Last week, however, the Boston firm's image of probity took a big hit.

Fidelity had come under some harsh media glare, thanks to published reports pointing out that some of its traders accepted gifts and business entertainment from brokerage firms it works with.

Fidelity reacted swiftly, announcing late last week that it had disciplined 14 individuals with penalties that included warnings and fines, according to a statement. Two other individuals are no longer working for the firm.