Back to Portfolio for the Future™

Formula: Information + Liquidity + Short Time Horizon = Short Sales Profits

February 14, 2017

Jaewon Choi and two other scholars recently presented a paper that offers “direct evidence about the profitability of hedge fund short trades in equities.”  What is intriguing about this paper is not only what it finds about the profitability of short trades, but its method, why its authors consider their evidence more “direct” than discussions already in the literature. These scholars call their work “a first glimpse” at hedge fund shorting. This might bring some readers up short, so to speak. Hasn’t the field been much debated? Haven’t there been lots of “looks”? Well … yes. But these authors consider their “look” the first “direct” examination of the matter because they identify the opening and covering of equity short sales with the help of a novel algorithm that combines data from different data sets and gets to the level of individual management companies. Who are they? Choi is an assistant professor of Finance at the University of Illinois at Urbana-Champaign. Co-author Neil Pearson is a professor at the same institution, and Shastri Sandy, as associate at the Brattle Group, has a Masters of Engineering from Massachusetts Institute of Technology. What are Their Findings? Choi et al make the following findings:

  • Short sales concluded by hedge funds and covered within five trading days are highly profitable;
  • If a short position remains uncovered for longer than that, it will likely prove unprofitable;
  • Institutional (non-hedge-fund) investors tend to lose on their short positions, regardless of the time horizon;
  • Some of the profitability of those quickly covered positions comes from an information edge, while another part comes from the provision of liquidity;

Those points are closely related. For example, the reason institutional investors don’t make profits on their short positions is that they don’t play for liquidity providing function in the markets that hedge funds do, and so can’t (in effect) charge for doing so. A Natural Experiment How did these scholars determine that their algorithm is reliable? Ah, that is an intriguing subject for the methodological nerds who will surely read this paper.  They know because of a natural experiment performed by the Securities and Exchange Commission during the global financial crisis. On September 19, 2008, the SEC banned all short sales of certain listed stocks, a list designed to protect the financial services industry from predation. On October 9, the SEC allowed the resumption of short sales.  Choi et al presumed that, if their algorithm worked properly, it should identify the quantity of short sales in the financial services industry’s issued stock as close to zero during that period. Indeed, when the algorithm is run it shows that the average dollar value of shares sold short plummeted close to zero exactly when one expects that it should. Thus, the algorithm was vindicated. The algorithm and the data it produces tell them that hedge funds earn 14 bps per day on their short positions on the sort of short-horizon trade described in the first bullet point above. Returning to the fourth bullet point above, one might well ask: how do these authors determine that the hedge funds are getting some of their profit from an informational advantage?  They find that hedge funds’ short sales predict future negative earnings surprises. That points to one, but presumably not their only, informational advantage. Hedge funds “earn returns of 13.5 bps per day during non-earning announcement periods,” after all. From the Horse’s Mouth Separately, 50 South Capital Partners, a Chicago-based investment firm that makes both direct hedge fund and funds of hedge funds investments, has issued its own document, a review of Hedge Fund Strategies for 2017 that indicates some points of agreement with Choi et al. This review says that for several years, the “backdrop” for active management in the equity markets has been negative, but that the firm now has high expectations for dispersion, and low expectations for correlations, which means a favorable environment for a long/short equity strategy. It plans to focus on “fundamentally biased managers that can demonstrate alpha generation on both sides of their portfolios” employing “unique short selling expertise.” Or, in other words, 50 South will be using the informational edge discussed in the Choi article. Meanwhile, 50 South Capital also has high expectations for an event-driven equity strategy, given the new administration in the U.S. It expects “a more accommodating regulatory review environment [which should be] a tailwind for merger arbitrage opportunities.”