Commission-free trading apps like Robinhood and coordinated action by retail investors on Twitter, WallStreetBets, and other social media have created an unprecedented force on Wall Street that specifically targets short sellers. One result has been a massive run-up in the stock prices of GameStop, AMC, and other “meme” stocks in 2021, presenting the SEC with the challenge of promoting the efficiency of capital markets – even as the prices of these meme stocks substantially deviate from firm fundamental value – while simultaneously protecting retail investors. AMC specifically warned its investors in a June 3, 2021, filing that, “We believe that the recent volatility and our current market prices reflect market and trading dynamics unrelated to our underlying business, or macro or industry fundamentals, and we do not know how long these dynamics will last.”
The trading frenzy has also left short sellers with substantial losses and fueled a debate about whether the SEC should require short sellers to disclose their short positions daily. On one hand, regulators argue for more disclosure and better transparency, and some retail investors want to know who is shorting the stocks. On the other hand, short sellers are increasingly concerned about revealing their short positions and becoming the targets of coordinated action aiming for a short squeeze.
The current disclosure requirement for short positions in the U.S. is that the aggregate short interest for each firm must be disclosed. In response to recent congressional concerns, the SEC is reviewing its role in mitigating potential detrimental effects of short sales. One option is for it to mandate that each investor disclose its individual short positions above a certain threshold, similar to requirements in the UK and other European countries. In our new paper, we examine the effect of mandatory short position disclosure on investor behavior by utilizing the UK setting, where short sellers have since November 1, 2012, been required to disclose their short position when it exceeds 0.5 percent of outstanding shares of a stock.
We employ two approaches and find evidence mainly consistent with the idea that mandatory short selling disclosure leads to investor herding behavior. First, we partition each fiscal quarter into three sub-periods: pre-announcement (30 days prior to earnings announcement), post-announcement (30 days after earnings announcement), and a no-information period. Our premise is that information-based trading is more likely to occur in either pre- or post-earnings announcements, especially for bad news. That is, if short sale disclosure clustering were driven by new fundamental information, we would expect to see these clusters only in high information windows, especially when that information was negative. Instead, we find that short sale disclosure occurs with similar frequency across the pre-earnings announcement, post-earnings announcement, and no-information windows. More importantly, the clustering of short sale disclosure does not vary significantly between good and bad earnings news, suggesting that information-based trading is not a major factor of short sale disclosure clustering. Second, we construct a matched sample of firms with similar short interest but no short selling disclosure and then examine subsequent changes in short interest between disclosure and matched non-disclosure stocks. By matching based on prevailing short interest, we isolate common information that may have affected information-based short positions of these investors. If disclosure itself induces investor herding behavior, then firm-level short interest is less likely to reverse (i.e., closing of short positions) for disclosure stocks than for matched non-disclosure stocks. Consistent with the herding hypothesis, we find that changes in short interest are persistently higher in the next 30, 60, and 90 days for disclosure stocks than for non-disclosure stocks. Indeed, short interest shows a strong reversal for non-disclosure stocks as high short interest tends to reverse, but barely exhibits a reversal for disclosure stocks.
Overall, we document an undesirable consequence of mandatory short selling disclosure, which informs regulators when they are considering related disclosure policies. The UK has institutional features very similar to those in the U.S., as both countries follow common law and have advanced financial markets with active short sellers. What happened in the UK following this short selling disclosure regulation has direct implications for the U.S., should the SEC adopt a similar regulation. Short sellers can always voluntarily disclose their intentions and positions, and mandatory disclosure is likely to increase the risk and costs of shorting selling. Short sellers’ herding behavior is not based on fundamental firm information and therefore could potentially drive stock prices away from firm fundamentals, making capital markets less efficient.
This post comes to us from professors John C. Heater at Duke University’s Fuqua School of Business, Ye Liu at Fudan University, Qin Tan at City University of Hong Kong, and Frank Zhang at Yale School of Management. It is based on their recent paper, “Does mandatory short selling disclosure lead to investor herding behavior?” available here.