Buying the Good Stocks Can Be Bad

A well-known stylized fact in investing is that, once academics discover and publish an anomaly—some pattern in stock prices that can be used to get predictable superior risk-adjusted returns—the anomaly tends to go away. The intuition is that anomalies represent stocks that are consistently over- or undervalued by the market, and publishing them corrects that. Once everyone knows that, say, companies run by chief executive officers who are scratch golfers underperform companies run by non-golfers, hedge funds will buy non-golfer companies and short scratch-golfer companies until their prices are correct and there are no more gains to be had by following the strategy. This is not completely true—some apparent anomalies might represent real risk factors that don’t go away; others might represent deep psychological factors that can’t easily be arbitraged; etc.—but it is a useful stylized fact, and makes a certain amount of intuitive sense.1

Here is a wild paper from Alex Horenstein of the University of Miami that is sort of a pure generalization of that intuition. The title is “The Unintended Impact of Academic Research on Asset Returns: The CAPM Alpha,” and the basic argument is that, once the capital asset pricing model was developed and published beginning in the 1960s, investment managers began to consciously seek out “alpha,” which is the now-standard term for the portion of a portfolio’s return that is not explained by the overall market return. Achieving positive alpha (rather than just high returns from levering up a market portfolio) became a main goal for investment managers. In practice this tended to mean that they bought stocks with high “realized alphas,” that is, stocks whose historic returns suggested that they added alpha. And so stocks with high realized alphas became more valuable and stocks with low or negative realized alphas became less valuable. And so the strategy of buying stocks with high alphas became bad: