Using Earnings Heuristics To Boost Alpha
Fuller & Thaler asset management on the use of earnings heuristics to boost alpha.
Introduction (Via Fuller Thaler)
A critical element of our Earnings Heuristics strategy is determining whether recent earnings changes are due to permanent or temporary factors. Under normal circumstances, security analysts make these determinations well. However, in certain situations analysts are vulnerable to heuristic biases. We exploit the fact that “simple models” can beat the forecasts of experts (analysts) when the experts are vulnerable to heuristic biases. The two most important heuristic biases exploited by our EH strategy are:
Excerpts (Via Fuller Thaler)
Anchoring Heuristic: When surprised by reported earnings, analysts naturally anchor to their old earnings estimate until they are convinced the earnings change is due to permanent rather than temporary factors.
Framing Heuristic: Portfolio managers, because of their traditional investment training, frame problems in certain ways, one of which is the need to “know the story” as to why earnings have changed.
Consequently, portfolio managers tend to wait for the analysts to develop the story for them — this is normally prudent behavior. However, we do not need to know the whole story in implementing EH, as we have developed procedures for quickly determining whether the earnings change is due to permanent or temporary factors. The tendency of portfolio managers to wait for the analysts to “develop the story” allows us to establish our positions before the higher earnings expectations are reflected in the stock price.
There are other heuristic biases which underlie Earnings Heuristics. Two examples: analysts have an asymmetrical reward structure with the penalties for estimates being high exceeding the rewards for being low; risk shirking on the part of portfolio managers creates additional incentives for portfolio managers to rely on analysts’ recommendations.
By exploiting the behavioral biases which underlie Earnings Heuristics, we are able to form portfolios of stocks for which:
a) analysts will subsequently revise their estimates of earnings upward, and
b) the upward estimates will not be large enough, resulting in subsequent
positive earnings surprises.
The economic “linkage” between exploiting these behavioral biases and Earnings Heuristics’ performance is quite simple: We assume that, on average, Future stock price changes will follow the direction of future revisions of
earnings estimates and future earnings surprises.