[This is another of a series of editorials by Executive Editor Andrew Karolyi at the Review of Financial Studies featuring recently published papers at the journal. This editorial features “Confusion of Confusions: A Test of the Disposition Effect and Momentum” by The Ohio State University’s Justin Birru, an article in Issue 28(7) for July 2015. It was selected as an Editor’s Choice article on the Oxford University Press web site for RFS.]
Readers may recognize from my title the symbolism of investing like Mr. Spock versus Homer Simpson from the popular book by Cass Sunstein and Dick Thaler, Nudge (Penguin Books, 2009). In it, they describe how investors’ frequent lapses in judgment combined with herd mentality can reveal their inner “Homers” as opposed to choices of efficient-market-style “Econs” in the image of Mr. Spock from Star Trek. An April article in Barron’s featuring Thaler (designed to draw attention to his newest book, Misbehaving, May 2015, Norton) got my attention, particularly the discussion around one of the most common Homerian “D’ohs,” the disposition phenomenon. Per prospect theory, recall how people are more risk averse in the domain of stock gains than in the domain of stock losses where they are more risk-seeking, so they tend to hold to losers too long. [Yes, yes, I know that there is hardly a consensus in our literature on whether prospect theory actually predicts the disposition effect, thanks to Barberis and Xiong, Ben-David and Hirshleifer, and others.] The Barron’s article goes on to describe investment strategies at Thaler’s investment firm, Fuller & Thaler, and at JPMorgan’s $2 billion Undiscovered Managers Behavioral Value fund, all designed to capitalize on the disposition effect and other investor cognitive biases.
For such investors, the new article by Justin Birru of The Ohio State University in the July 2015 issue may offer a sobering reality check. The study uses investor-level data to re-examine the disposition effect specifically around a stock split. It’s an intriguing stock event that can serve as an optical illusion. Consider an example. For a stock undergoing a 3-for-2 stock split, an investor with an original purchase price of $20 should now realize that a stock price of $14 actually represents a gain rather than a loss, as the new reference price should be $13.33. What Birru finds is that the disposition effect breaks down following a stock split. In the period after a stock split, investors no longer realize gains at a rate any different than their losses – the winner/loser status of the stock is no longer significant for the selling decision. He suggests a bunch of candidate explanations. Maybe investors are inattentive to the split? Maybe they are unable or unwilling to properly update their reference price? Birru admitted to me in private communication that his paper never offers up a definitive answer here.
But that is just fine. Because this temporary breakdown provides a unique laboratory in which to investigate how and whether the disposition effect induces return predictability. That is the secret code that the Mr. Spock managers are trying to crack! Birru has the findings of prior studies looking at how the disposition effect can explain the momentum anomaly in his sights. Theory predicts that prices will revert to fundamental values in the absence of the disposition effect, so this should lead to a dissipation of momentum returns. It turns out that the momentum anomaly is alive and well in the post-split sample of stocks. He operationalizes the tests using a stock-specific capital gains variable built prior to the split event, which he uses to gauge whether the stock is above or below its fundamental value and with which he predicts their returns post-split.
What are the big picture takeaways? Birru writes to me that investors’ no longer behaving according to the disposition effect following a split is interesting in and of itself. But what he views as more important is that the setting can be exploited to shed light on the relationship between momentum and the disposition effect. Good to know this cognitive bias cannot be the primary driver of momentum returns. Maybe his effort will prompt scholars to explore other corporate events or capital market settings salient for other behavioral biases and the large array of anomalies to be pursued. One thing for sure is that Birru’s work will be a useful caution for our Mr. Spocks out there who seek to capitalize on the vulnerable Homer Simpsons.