Who Do You Trust?

Only three years after New York Attorney General Eliot Spitzer cleaned up equities research on Wall Street, investors seeking advice on which stocks to buy and sell find themselves muttering a familiar question, Who do you trust? Independent analysts who are supposed to be conflict free aren't. Big-brokerage analysts, the "sell side" bunch, are doing better--but they have a long way to go. The answer, then, as ever on Wall Street: Trust yourself and, maybe, your dog.

That doesn't mean you should dismiss Wall Street research out of hand. Much of what analysts produce is helpful, especially if what you seek is broad background on an industry. Occasionally, analyst conclusions are earth moving. Recent testimony in the trial of former Enron top dogs Kenneth Lay and Jeffrey Skilling revealed it was an analyst who first questioned the company's finances, prompting Skilling to fret aloud, "They're on to us."

Never adopt an analyst's opinion as your own without doing some homework. Even professional money managers with access to the best minds on Wall Street do their own digging. "They look at research to either bolster their confidence or push them to ask more questions," says Michael Mayhew, CEO of Integrity Research Associates of Darien, Conn., which vets research for institutions.

Don't blame regulators for not going far enough in the cleanup campaign. A settlement reached in 2003 with 12 Wall Street firms forced disclosure of analyst conflicts, set aside $460 million to fund and distribute unbiased stock research to individual investors and linked analyst pay to the performance of stocks they pick. That's a lot better than the old days, when an analyst might tout stock of a so-so company purely to ingratiate himself with management.

Yet even under the new rules there is no guarantee that independent analyst advice is conflict free. Drug company Biovail has asserted that independent research firm Gradient Analytics wrote a report trashing the company, and that the report was paid for by a hedge fund that had sold Biovail short (a bet the stock would decline). Gradient has said its conclusions were its own.

The episode raises questions about the output of the 450 independent firms, most of which sprang up after the settlement. Are hedge funds paying for conclusions that would move a stock in their direction? What else don't we know about how independent research firms operate? "The Gradient story hasn't helped their image," says Mayhew.

Meanwhile, brokerage analysts are slipping back into some bad habits. Sell ratings spiked to 12% of all opinions after the settlement, a sign that analysts were showing some backbone. But that figure slipped to 9% last year, reports research firm StarMine. "There are still far too many buy ratings, and analysts' [earnings] estimates are increasingly clinging around company guidance," Morgan Stanley strategists Henry McVey and David McNellis wrote in a recent report.

Here's how you can make good use of research:

•Ignore any analyst rating that is more than six months old. That just means the analyst has dropped the ball.

•Analysts' revisions are more important than their ratings. So an upgrade from a sell to a neutral rating is more useful information than a long-standing buy rating.