Picking Stocks Based On “Consistent” Earnings Surprises

Analysts play an important role in the market, setting expectations and giving investors with limited resources research either to base their investments on or to use as a starting point for ideas, so systemic bias among analysts could result in systemic over- or under-buying among investors.

Picking Stocks Based On "Consistent" Earnings Surprises

“We have also been concerned with this topic and concluded that when potential biases are removed the information content of analysts’ forecasts becomes much more significant,” says a recent Citi report from Chris Montagu and Liz Dinh introducing new academic research on the topic.

Performance forecasts against stocks recommendation

Economists Byoung-Hyoun Hwang at Purdue University and Dong Lou at London School of Economics have found evidence that analysts adjust performance forecasts against their stock recommendations to reduce the odds of a contradictory earnings surprise. “An analyst with a bullish recommendation has an incentive to report downward biased earnings forecasts, so that the company is less likely to experience a negative earnings surprise,” they write.

This doesn’t mean that analysts are consciously adjusting their predictions, only that they are responding to real career incentives. Even if an analyst’s stock recommendations are sound, a low performance forecast with a positive surprise leaves a better impression than a high performance forecast with a negative surprise (and vice versa for short recommendations). Hwang and Lou find evidence that these contradictory earnings surprises are more significant predictors of whether someone will “leave the analyst sample” than inaccurate forecasts.

Investment strategies

As interesting as it is to speculate on how incentives affect perception, Hwang and Lou use this information to construct an investment strategy that buys the top third of stocks with outstanding recommendations three days before earnings announcements (and shorts stocks following the same strategy), and then sells them three days afterward. Since everyone involved in setting the consensus is mentally adjusting that consensus, this stock picking strategy is trying to capitalize on a market anomaly.  As simple as it is, this strategy would have earned an impressive 1.6% monthly risk-adjusted return over the last few years.

“The bottom line for stock pickers is that the outstanding recommendation can on its own be a successful predictor of earnings surprises and earnings announcement returns,” write Montagu and Dinh.

Get our newsletter and our in-depth investor case studies all for free!