51³Ô¹ÏºÚÁÏ

Analysts are more likely to issue "buy" or "strong buy" recommendations for stocks when the firms that they cover are also investment banking clients of the analyst's brokerage house, says a finance professor at 51³Ô¹ÏºÚÁÏ Davis.

of the and three co-researchers are studying the 1996 to 2000 recommendations of affiliated analysts. A preliminary analysis of recommendations issued by affiliated analysts shows they are more likely than others to issue buy or strong buy recommendations.

"Financial markets would be better off if the biases of affiliated analysts would be well documented for investors," Barber says.

He and his co-researchers -- Reuven Lehavy and Brett Trueman, both of 51³Ô¹ÏºÚÁÏ Berkeley, and Maureen McNichols of Stanford University -- have published two other studies regarding analysts.

In a paper forthcoming this spring in the Financial Analyst Journal, the researchers found that, in the year 2000, stocks least favorably recommended by analysts earned an annualized market-adjusted return of almost 49 percent while stocks most highly recommended fell 31 percent.

"While we cannot conclude that the 2000 results are necessarily driven by an increased emphasis on investment banking by analysts," they wrote, "our findings should add to the debate over the usefulness of analysts' stock recommendations to investors."

In a paper published last spring in The Journal of Finance, the researchers found that highly rated stocks performed well. However, none of several investment strategies designed to take advantage of analysts' recommendations generated significant returns after accounting for high transactions costs.

The New York State attorney general recently released details of a 10-month conflict-of-interest investigation, including e-mails in which Merrill Lynch & Co. employees privately said derogatory things about stocks they were promoting. Eliot Spitzer said the company's drive to generate fees from investment banking tainted the quality of its research and hurt investors.