Setting analysts straight

May 11, 2005 -- It's been two years since New York Attorney General Elliot Spitzer's crackdown on the securities industry spawned the "Global Analyst Research Settlement." Spitzer targeted analysts and bankers from the same company for getting a little too cozy with each other. The allegation was that such a close relationship motivated analysts to publish seemingly biased reports in favor of companies that did business with the investment bank. Recently, professors from Washington University in St. Louis Olin School of Business released a report indicating that the Global Settlement is successfully eliminating the apparent conflict of interest.

In one of the first detailed analyses of the effects of the settlement, assistant professor of accounting Tzachi Zach, assistant professor of finance Ohad Kadan and Ph.D. candidate Rong Wong found that the level of bias in analysts' reports has decreased significantly — so much so that by most measures, the professors say, they found virtually no evidence of significant bias.

Before the settlement, there was concern that analysts were being pressured by their companies to produce overly optimistic reports. Zach says that analysts would sometimes get immediate monetary gain from generating positive recommendations that could potentially lure new business. It was frequently expected that as a condition of a company doing business with a certain investment bank, the analysts would then issue positive recommendations.

"The conflicts of interest attracted a lot of attention," Zach says. "People said the analysts are not issuing unbiased opinions. Global Settlement imposed a strict separation between research and investment banking divisions."

Professors Zach and Kadan identified a set of firms for which those conflicts of interest were most likely to influence analysts' reports. The researchers then compared the recommendations issued by affiliated analysts to those of unaffiliated analysts. As other studies have proven, prior to the settlement affiliated analysts had generated more optimistic reports than their unaffiliated counterparts. Then the researchers looked at the reports issued by the same companies post-Global Settlement.

"The gap in bias between affiliated and unaffiliated reports decreased significantly after the implementation of Global Settlement," Zach says. "In fact, any observable favoritism that was detected before the settlement is no longer there."

The researchers were aware that many variables could explain the reduction of biased reports. Zach and Kadan say they tried to control for things like changes in the economy or in stock market. Even taking into consideration those influences, Kadan says that nearly every measure they used to detect bias had the same results: the Global Settlement seems to have effectively eliminated bias in analysts' reports.