More bubble stocks exist today than at any other time over the past century—with one exception: the top of the dot-com bubble in March 2000.
That’s the conclusion I reached when applying the findings of a recent academic study to the current stock market. Entitled “The Dynamics of Disagreement,” the study was conducted by three finance professors: Kent Daniel of Columbia Business School, Alexander Klos of Germany’s Kiel University, and Simon Rottke of the University of Amsterdam. (Daniel used to be Goldman Sachs’ co-chief investment officer.)
To appreciate what the researchers found, it’s helpful to remember that in the normal give and take between buyers and sellers, bubble stocks should be few and far between. That’s because short sellers will be quick to bet against exuberant investors as they try to bid up their favorite stock to unreasonable prices. The professors hypothesized that bubble stocks will be more likely to form when short sellers face more than the usual obstacles to selling stocks short. To test this hypothesis, the researchers focused on high-flying stocks for which relatively few shares are available in the share-lending market. That lower volume makes it less likely that would-be short sellers will be able to find shares to borrow and ultimately sell short. That in turn makes it more likely that such stocks—what the professors refer to as “constrained winners”—will have gotten too far ahead of themselves and therefore underperform the market going forward.
The professors compiled a list of stocks that satisfied their criteria of being a constrained winner at any point between 1980 and mid-2020. They found that, on average, these stocks lagged the market for five straight years after making it on to the list, lagging the market over those five years by 60%. I think we can all agree that a loss that big satisfies the definition of a bubble bursting.
Today’s Bubble Stocks To determine which of today’s stocks satisfy the professors’ “constrained winner” criteria, I replicated, as closely as possible, their methodology with all exchange-listed U.S. stocks (courtesy of FactSet data). A total of 131 stocks survived, a high number—nearly three times the average since 1980 of 49. The only other month of the past four decades when the number was any higher was at the top of the dot-com bubble in early 2000, when 140 stocks satisfied the criteria. Since there are many fewer exchange-listed stocks today than in 2000, the 131 constrained winners today likely represent a record as a percentage of the total market. In an interview with Barron’s, Daniel said that he and his co-authors didn’t specifically study whether a larger constrained-winner group indicated that the overall market was more vulnerable to a decline. But he said it is certainly plausible that the large current number of constrained-winner stocks is a warning sign for the market as a whole. Prior Bubble Research The approach that the professors devised differs crucially from prior research into predicting bubble stocks. Many of those earlier efforts focused on performance alone, on the theory that “the higher they go, the further they fall.” And while that’s definitely a first step, this new research goes a step further to differentiate between winning stocks that are easily shorted and those that aren’t. It’s the latter that are particularly vulnerable, according to the professors. To appreciate the increased predictive power of focusing on stocks that are less easily shorted, consider the performance of stocks that are otherwise similar to those in the professors’ constrained-winners portfolio—but which are not difficult to short. On average, these “unconstrained” winners didn’t lag the market over the subsequent five years—in contrast to the constrained winners. The full list of 131 stocks is beyond the scope of this column. The table below lists the 25 with the largest market capitalizations. Bubble Candidates These 25 market-beating stocks are relatively difficult to sell short, making them impervious to the market's checks and balances.