From Jessica Rabe of Convergex, discussing psychological bias:
Summary: Want an edge this earnings season? Focus on traits of companies’ CEOs. Here are the characteristics to look for: gender, nationality (domestic or foreign), and political orientation. Four researchers from the University of Miami, University of Mannheim, and University of Michigan found that sell-side equity analysts “systematically underestimate earnings of firms headed by CEOs who do not belong to their own” group when forecasting earnings. Given that Wall Street analysts skew male, Republican, and US born, they tend to underestimate companies run by the opposite. Consequently, earnings surprises of firms with female, foreign, or Democratic CEOs are systematically upward biased according to the researchers. This in turn could help their share prices when these underestimated companies report their actual numbers. And, of course, there is a deeper cautionary tale: consider your own biases as you crank away on your financial model.
Note from Nick: investing is a microcosm of life itself. We go into any decision armed with our knowledge and exposed to our biases. And since humans are social – and therefore tribal – it makes sense that we evaluate corporate managements through the lens of “One of us” or “the other”. Today Jessica reviews a paper that highlights that phenomenon.
If I were to flip a coin, how much would you have to win if it lands on heads knowing that you’d lose $100 if it lands on tails? Anything greater than $100 is reasonable, but the answer for most individuals is at least $200 according to psychologist and Nobel Prize winner Daniel Kahneman. In a “Masters in Business” podcast run by Barry Ritholtz, Kahneman gave this example to explain a concept he pioneered and that we know well in finance: prospect theory.
In prospect theory, people feel greater pain from a loss than joy from a gain. In other words, we are loss averse and therefore require $200 of potential gain in the above example to compensate for $100 of potential loss as both chances are equal. Kahneman said it’s about a 2 to 1 relationship and that loss aversion comes down to evolution in which threats are more pressing than opportunities. His advice to investors: don’t look at your investment results too frequently or you may increase your chances of making a change that often results in a loss.
Here are some other investment/general biases Kahneman discusses in the interview:
- Anchoring: people tend to focus on the first piece of information given too much when making a decision. Kahneman’s example, despite what people think, those who go first in a negotiation have the advantage as “the mind tries to make sense out of whatever you put before it.”
- Availability: individuals form the best story they can based off information they have and deem unknown information as unnecessary. Kahneman’s example, if he said a national leader is intelligent and firm, you’d probably think she’s a good, not a bad leader. But what if the third word he was going to use was corrupt? Individuals tend to form judgements based on the information they have and fail to wait for information they don’t have, or how he described it: “what you see is all there is.”
- Attribute substitution: if someone is asked a tough question they can’t answer, they’ll give an answer to a related, easier question all the while not answering the more difficult question. Kahneman’s example, if he asks how happy you are these days, you’ll likely respond about your current mood as opposed to his more general question about how you are “these days”. In this case, someone substitutes an easier question.
- Theory induced blindness: we are unwilling to admit when we’re wrong, but rather try to craft the “best story possible” to make sense of what had occurred by altering “our image of what we thought earlier”. Rather than admitting a mistake, we believe we anticipated it in hindsight when we actually hadn’t. Kahneman’s example: if two football teams are evenly balanced going into a game and one “crushes” the other, you now perceive one as much stronger. Further, “that perception gives you the sense that this must have been visible in advance, that one of them was much stronger than the other”.Kahneman’s takeaway: hindsight “allows us to keep a coherent view of the world, it blinds us to surprises” and “prevents us from learning the right thing”. Even when we do admit a mistake, we say we won’t do it again when we should really learn from surprises that “the world is difficult to anticipate”.
- We strongly encourage you to listen to the podcast to hear Kahneman on more cognitive biases and heuristics, his specialty. He discusses how he started working with his late research partner, Amos Tversky, and how they came to their novel findings. He also talks about the difference between satisfaction and happiness and offers up his favorite books.Link: http://ritholtz.com/2016/08/mib-kahneman-heuristics-biases-cognition/
If we fall privy to biases as investors and in our everyday lives, what about the people who help inform our decisions on how we view companies? Turns out analysts underestimate certain companies based on the traits of their CEO. Four researchers from the University of Miami, University of Mannheim, and University of Michigan found that sell-side equity analysts exhibit group bias, whereby they “have less favorable opinion about firms that are not headed by CEOs of their own group”. Here’s their results based on gender, ethnicity, and political orientation:
- “Male analysts provide lower earnings estimates and worse stock recommendations for firms with female CEOs “. Female headed firms report earnings that surprise the market to the upside more often “because the consensus levels are lower”. Additionally, “firms with female CEOs disappoint (earnings < consensus) analysts less frequently.” When taking into consideration the forecasts of only male analysts, “the consensus is even lower for firms headed by female CEOs.”In terms of recommendations for firms with female CEOs, “there are fewer buy and strong buy recommendations and a greater proportion of sell recommendations.” In fact, “analysts issue 3.24% fewer buy and strong buy recommendations and 2.13% more sell and strong sell recommendations for female headed firms, compared with male headed firms.” When looking at recommendations from only male analysts, these disparities “become even more pronounced.”
- “Domestic analysts provide lower earnings estimates and worse stock recommendations for firms with foreign CEOs.” The researchers found that analysts are “significantly more likely to underestimate earnings of firms headed by foreign CEOs compared to firms headed by domestic CEOs.” Firms with foreign CEOs therefore have significantly higher positive earnings surprises, “while (absolute) negative earnings surprises are significantly lower for firms with foreign CEOs.”Moreover, “there are significantly less buy and strong buy recommendations for stocks of firms headed by foreign CEOs, compared to firms headed by domestic CEOs. At the same time, there are significantly more sell and strong sell recommendations for these firms.”
- “Republican analysts provide lower earnings estimates and worse stock recommendations for firms with Democratic CEOs.” Analysts who identify as Republican are “significantly more likely to underestimate earnings of firms headed by Democrat CEOs compared to firms headed by Republican CEOs.” Consequently, “positive earnings surprises based on Republican analysts’ consensus forecasts are significantly higher for firms with democrat CEOs, while (absolute) negative earnings surprises are significantly lower for firms with Democrat CEOs.”
- “The market responds more strongly to positive and negative earnings surprises of female headed companies.” The researchers concluded that investors are biased too: “stock market participants also exhibit in-group bias along the gender dimension.” They also found that “the stronger market reaction to negative earnings surprises of firms headed by female CEOs provides evidence that when there is negative news, the market punishes female CEOs more than male CEOs.”Link to study”
The upshot: earnings surprises of firms with female, foreign, or Democratic CEOs are systematically upward biased according to the researchers. Analysts underestimate the earnings potential of companies headed by these “out-groups”, which in turn could help their share prices when they report the actual numbers. These findings aren’t necessarily surprising when looking at a few stats in the report: 80% of analysts are male, 48.1% are identified as Republican (compared to 41.1% as Democrat), and only a minority is identified as foreign. If humans tend to favor those in their own groups, this shined through in the data for companies with CEOs in the “out-groups”.
As we continue to work our way through earnings season, there’s some merit in paying closer attention to the characteristics of CEOs for an edge over those just focused on the math.
Nick Colas, Chief Strategist ncolas -at- convergex . com
Jessica Rabe, Research Associate jrabe -at- convergex . com