When even the most honest stockbrokers call with a hot tip, you should be skeptical of the accompanying earnings forecasts.
According to new research by a professor of accounting at Brigham Young University's Marriott School of Management, analysts are unconsciously influenced to distort their forecast of a company's earnings to be consistent with previous recommendations to "buy" or "sell" that company's stock.
What does that mean for investors, who routinely look to a company's earnings as a measure of its performance?
"It is important for investors to be aware that broker analysts' earnings forecasts are biased in the direction of the stock recommendation," said BYU's Steve Glover. "Most analysts aren't producing these biased forecasts on purpose; there is simply an inherent bias in the system. Keeping that in mind can help people better evaluate potential investments."
In an earlier study published in The Journal of Accounting Research, Glover first observed a pattern of forecast errors that was consistent with an unintentional bias in actual stock market data. However, that study could not rule out intentional biases.
Glover's new follow-up study, published in the latest issue of the journal Behavioral Research in Accounting, was conducted in an environment free from incentives to intentionally bias earnings forecasts. More than 150 MBA students were provided with identical company background, previous earnings information and no information about current or past stock price. Participants were motivated to develop an accurate short-term earnings forecast.
Even though participants did not consider the previous stock recommendation relevant, the results showed that the previous recommendation did influence their forecasts.
"The bias is not intentional, but we did see optimism in earnings forecasts when the previous recommendation is a 'buy' and pessimism when the previous recommendation is a 'sell,'" Glover said.
Douglas Prawitt, a BYU accounting professor unaffiliated with Glover's study, said the study shows that recent new market regulation targeting stock analysts will not entirely fix the problem of biased earnings forecasts.
"Although the regulation may successfully reduce intentional biases, no amount of regulation can eliminate unintentional cognitive bias," Prawitt said. "Those types of biases are just hard-wired into how we think and process information."
Some academics and regulators have accused broker-analysts of issuing intentionally optimistic earnings forecasts to curry favor with management in a game of "winks and nods," Glover said. According to his two studies, which are both co-authored by Michael J. Eames of Santa Clara University and Jane Jollineau Kennedy of the University of Washington, those accusations are misplaced.
"That's nonsense," Glover said. "When a company misses the consensus analyst earnings expectation, the market typically punishes the company with a drop in stock price. Issuing intentionally optimistic forecasts would certainly not 'curry favor' with management, and therefore, this explanation cannot explain the pattern of forecast errors observed in market data. This study shows that at least some of the bias isn't deliberate."
According to Prawitt, one of the reasons Glover's studies stand out is because they involve two completely different but complementary research methodologies: "The archival market study provides strong evidence that a bias exists in the marketplace and the behavioral study indicates that the bias is caused, at least in part, by an unintentional bias."
Writer: Anthony Strike