Two years after Eliot Spitzer's assault on shoddy stock research, there's evidence that Wall Street has begun to clean up its act.

In the last year, sell-side brokerages have seeped into five of the top 10 rankings on a list of the best-performing research shops complied by Investars.com, a site that tracks stock selection for the public. That's a marked improvement from two years ago, when nine out of the top 10 spots were occupied by independents.

The data helps vindicate Spitzer's global research settlement, which was designed to restore the integrity of Wall Street investment advice after years of conflicts and fraud. At the heart of the settlement was transparency: Spitzer's rules required brokerages to account for their stock picks publicly so investors could decide whose advice was best.

"The reforms that were carried out actually helped improve performance of the sell side and resulted in improved sell-side products for regular investors," says Investars Chief Executive Kei Kianpoor, who testified before Congress in July of 2001 on the need for transparency. "The whole point of the settlement was to ensure transparency, so firms would be held accountable for their ratings.

"There's been a lot of focus on independence, but the focus instead should be on performance," Kianpoor said.

The settlement, finalized in 2003, also required brokerages to spend millions of dollars to provide their customers with independent research that had no connection to investment banking, as an alternative to their own advice. It also forced the big firms to release historical ratings information to the public, so performance-trackers such as Investars.com could make the record public.

Speaking at the Investorside Research Association conference in New York last week, Spitzer said it's too early to tell whether regulatory reforms or the improved stock market deserve more credit for bringing Wall Street's track record up to snuff.

Nevertheless, sell-side analysts do appear to be more discerning. While sell ratings accounted for fewer than 2% of the ratings published on Wall Street in 2002, Investars.com reports that sell ratings now make up at least 10% of the ratings at all the major brokerages. In most cases, the figure is closer to 15%.

"Is sell-side research better than before? I'm not sure -- there are too many market variables involved in making that judgment," Spitzer said. "But my sense is that the internal analytic products at these firms have more integrity today than they did several years ago, because their customers have access to more sources of information."

If there is a gripe about the impact of the settlement, it is that independent research outfits are not held to the same high standards of transparency as their larger Wall Street brethren.

As retail investors get exposed to more independent stock research, comparing the performance of different firms is sometimes difficult. While the settlement ensured that the public could track the research performance of sell-side firms, many independent firms that serve retail investors through the settlement don't make their performance available, and they are not required to do so.

(It should be noted that the publisher of this Web site,

TheStreet.com Inc.

(TSCM)

, also owns Independent Research Group LLC, a firm that serves institutional investors only and is not involved in the global research settlement.)

Michael Mayhew, the chief executive of Integrity Research, noted that there are fewer than 30 independent firms ranked on Investars.com, while research from at least 100 firms is being made available to retail investors through the settlement.

"There's no transparency for the independents," Mayhew said. "While the settlement may have helped investors with one problem, it may have also helped create another."

Mayhew predicts that independent research will continue to grow in the years ahead, while research produced by sell-side firms declines as trading execution costs go down. Ten years ago, he said, there were probably fewer than 10 independent research firms, and there are now 300 to 400 firms serving various niches in the market.

"Some sell-siders have been discussing getting out of research altogether," he said. "We're going to be in a marketplace that is changing and changing rapidly."

That makes some of the recent trends worrisome. Major independent firms such as Alpha Equity Research and Computrade, which once garnered praise for stellar research performances, stopped making their ratings public when they joined Jaywalk, a subsidiary of Bank of New York that aggregates independent research for brokerages under the settlement.

"The fact that the public does not have access to the track record of how these firms perform is worrisome," Kianpoor said.

Lehman Brothers consultant Mark Fichtel, who selects independent research for the bank's clients as part of the settlement, said he monitors the performance of the research himself, although his data is not available to investors.

"We recognize that this is a piece of information that investors don't have," Fichtel said. "On the other hand, it isn't like this is an unmanaged program. Those of us involved with Jaywalk are monitoring performance, and I know that I change it when it appears to be warranted. And we have an obligation to report to regulators at the end of the year what performance we've seen."

Fichtel said there was no basis for requiring independent firms to disclose their ratings since they don't necessarily have any conflicts of interest.

"That's not an issue for independents," he said.

Critics reply that independent firms could run into other kinds of conflicts of interest. They might house proprietary trading desks and research analysts under the same roof, for example.

"Just because a firm has no investment banking division, does that mean it is really independent?" wondered Jonathan Boersma, senior director with the CFA Institute. "The jury is still out on that."