If you thought stock research was simple and transparent in the wake of Eliot Spitzer's crusade, you haven't looked closely at Wall Street's new rating systems.

While all the major investment banks now adhere to a three-tier system to classify stocks, no standardized rules have been adopted, meaning that each brokerage has been able to define buy, sell and hold in its own unique way.

"Companies have a right to rate stocks any way they want, but they also have a responsibility to provide a common denominator so we can compare them," said Thomas White, managing partner of Best Independent Research.

Double Talk

At Merrill Lynch ( MER), a buy rating on a stock with low or medium volatility suggests a return of 10% over a 12-month time horizon, while the same rating on highly volatile stocks implies a 20% return, including dividends. Stocks rated neutral are expected to gain between nothing and 20% -- depending on the level of risk that Merrill deems them to have -- while stocks labeled sell are seen producing a negative return.

But at UBS Warburg ( UBS), a buy rating suggests that a stock will gain 15% or more over the next 12 months, including dividends, while a neutral rating implies a return of 15% or less, with the potential to decline as much as 15%. If UBS analysts slap a reduce rating on a stock, they expect it to fall by 15% or more. All of the firm's recommendations are determined by the analysts' target price.

To make matters more confusing, UBS recently started rating its price forecasts separately, assigning a 1 rating to a price target that has a high probability of being met and a 2 rating to targets that could potentially be way off the mark.

Over at Lehman Brothers ( LEH), ratings are assigned in an entirely different way. A buy rating, or overweight as they like to call it, means simply that a stock should perform better than the average stock in the analysts' coverage universe, according to a spokeswoman. In that case, investors would need to find out how the average stock had performed before acting on the advice. Lehman analysts also rate the attractiveness of their respective sectors.

"It made for great P.R. to go to a three-point system, but everyone's interpreted it in a different way," said Kei Kianpoor, CEO of Investars.com, a Web site that tracks analyst research. "Investors shouldn't need a degree in deconstruction to understand what a buy means."

Both Sides of Mouth

Last year, New York State Attorney General Eliot Spitzer uncovered emails showing that analysts at Merrill Lynch and Citigroup ( C) unit Salomon Smith Barney had privately disparaged stocks that they were publicly recommending. That prompted a larger investigation into conflicts of interest on Wall Street and led to significant fines as well as changes to the way analysts are compensated.

Although not required by any settlement or new securities laws to alter the way they rate stocks, all the major brokerages have revised their systems over the past year, by reducing the number of ratings issued on stocks from four or five to just three. But because the system has not been regulated, some say it remains confusing and can still mislead investors.

To be sure, the major brokers all explain their ratings methodologies at the top of their research reports, and the system is better than it once was. Investors can at least take some comfort in knowing that a buy rating is no longer analyst-speak for hold, and a hold doesn't really mean sell.

Looking for Balance

Patricia Walters, senior vice president of professional standards at the Association for Investment Management Research, said that while the current system has some flaws, homogenization isn't the way to go. "I think we need to find a balance between information that's comparable and information that gives companies the ability to communicate their different perspectives," she said.

In the meantime, however, White said investors should go to a third party like Investars to analyze the analysts, just as investors now go to an independent research firm like Morningstar to analyze the performance of portfolio managers.

Whether or not brokers are trying to circumvent being measured won't matter in the end, he said, because third parties will continue to monitor the analysts, "and in time these companies will come to realize that providing comparable information is a valuable tool."