The study is particularly relevant as investors are struggling to weigh how to treat the new independent reports, particularly when they conflict with recommendations by a firm's own analysts. Based on the study, when there's a conflict, investors "would do better by following the recommendations of the independent research providers," says Brett Trueman, a professor of accounting at UCLA's Anderson Graduate School of Management, who co-authored the study with Brad Barber of the University of California, Davis, and Reuven Lehavy of the University of Michigan.

The independents' edge was particularly striking after the Nasdaq Stock Market peaked in March 2000. "During the bear market, the independents slaughtered the investment banks," says Prof. Trueman. The authors think the banks' performance had to do with the fact that they were reluctant to downgrade stocks because of investment-banking ties.

The study wasn't all bad news for major brokerage firms. All of the firms in the study and the independents did just about equally well during the bull market, the study found. The authors say that's not surprising because the banks were issuing "buys" at a time when shares were largely rising.

According to the study, the 10 investment banks that were part of the securities settlement turned in their worst performance between March 11, 2000, and June 2003, when stocks were performing poorly. The banks' picks underperformed those of the independents by an average of 18 percentage points a year during that period. The banks' track record was even worse for recommendations issued or outstanding after an initial public offering or follow-on stock offering during that period: Those picks underperformed by an average of 21 percentage points a year. The poor performance extended to firms with investment-banking business that weren't included in the settlement. (Under an agreement with Thomson Financial First Call, a unit of Thomson Corp. that provided the study's data, the authors agreed not to provide information about specific firms that they studied.)

The study doesn't directly answer investors' questions about how good the new research will be. That's because many of the firms that will provide research under the settlement aren't in the study. Among those that are: Buckingham Research Group, Cathay Financial and Green Street Advisors Inc.

Still, the findings suggest that it can pay to ask for a second opinion -- and to carefully consider its findings. "The research reports investors will be getting can be quite useful supplementary information," says Prof. Trueman, who along with his colleagues looked at roughly 335,000 stock recommendations made by more than 400 securities firms.

Investors can obtain the reports via their brokerage firm's Web site or toll-free number. Trade confirmations and account statements must include the ratings given the stock by the firm's own analyst and an independent firm. The firms must offer the independent reports when they solicit an order for a domestic stock and certain foreign securities covered by their own analysts.

While the settlement means that many investors will now get access to independent research that was previously unavailable to them, there has been little information about the quality of this research -- or the track records of its providers.

StarMine Corp., which rates analyst performance, says its data show uneven results on three independents providing research under the settlement: Argus Research, Buckingham Research and Fulcrum Global Partners. None "really stand out in a way that I can commend their performance overall," says David Lichtblau, StarMine vice president. While the firms have some standout analysts, he says, the overall "results are mixed, both in terms of accuracy of earnings estimates and stock-picking performance...[and] more or less in line but slightly worse than" the average brokerage firm.

John Eade, president of Argus, says that "during periods of market weakness...the independent research firms generally look great. During periods of market strength, the investment bank-brokerage research often looks good."

Robert Hoehn, director research at Fulcrum Global says that through July his firm's average recommendation is up 3.1%, "roughly twice as good as the S&P [500-stock index]." Manny Korman, director of research at Buckingham, says that without knowing the basis for StarMine's findings, it's hard to comment.

Others suggest that the independents may be losing some of their edge. "Over the past year, as the market has gone up, independents haven't performed as well," says Kei Kianpoor, chief executive of Investars.com, which tracks analyst research. One reason: Independents tend to issue more "sell" recommendations than firms with investment-banking ties, which means they do worse when share prices are rising. The settlement and media scrutiny may have also played a part.

Other studies suggest that firms with investment-banking ties to a company tend to do better on earnings estimates.