The Effect of Short Selling on Market Reactions to Earnings Announcements*

Abstract

This paper examines the effect of the inherent demand implied by short interest by studying how stock price reactions to earnings announcements depend on the level of short interest. We find that, for extreme good and bad news events, the inherent demand increases stock prices around the earnings announcement date, with the effect being stronger for good news relative to bad news. Specifically, the initial market reaction to an extreme positive earnings surprise is larger for firms with high levels of short interest. On the other hand, for an extreme negative earnings surprise event, the initial market reaction is less negative for heavily shorted firms. Furthermore, we find that the post-earnings-announcement drift is smaller (larger) in magnitude for extreme positive (negative) earnings surprises for the heavily shorted firms.

Unsophisticated Arbitrageurs and Market Efficiency: Overreacting to a History of Underreaction?

“…28 Although I use I/B/E/S earnings announcement dates, I also require the Compustat earnings announcement date to either be the same day or the day after, in order to eliminate potential errors in earnings announcement dates. 29 Several studies using either monthly data or limited sample periods find evidence of short-selling activity related to PEAD (e.g., Boehmer and Wu [2013], Berkman and McKenzie [2012], Lasser, Wang, and Zhang [2010], Christophe, Ferri, and Angel [2004]).…”

Short Interest and Lottery Stocks

“…Similar to Lasser et al. () and Prado et al. (), we focus on the top and bottom SUE quintiles for the positive and negative earnings surprises, respectively, to examine PEAD implications as stated in Hypotheses 3 and 4.…”

“…Furthermore, Hypotheses 1 and 2 have implications for post-earnings announcement drift (PEAD). We draw inspiration from Prado et al (2016), who find that higher institutional ownership concentration leads to greater PEAD due to lower speed of adjustment to news, and from Lasser et al (2010), who find that heavily shorted firms exhibit smaller PEAD for positive news and larger PEAD for negative news as compared with lightly shorted firms' PEAD. Nevertheless, our predictions are distinct from prior work.…”

“…(). Lasser, Wang, and Zhang () find that heavily shorted firms exhibit smaller PEAD for positive news and larger PEAD for negative news as compared with lightly shorted firms’ PEAD. We predict that this effect will be more pronounced as LOTT increases, because lottery stocks are associated with greater overvaluation in high RSI stocks and greater undervaluation in low RSI stocks.…”

Are Short Sellers Informed? Evidence from the 2007–2008 Subprime Mortgage Crisis

“…We thus include the dollar trading volume (TRADVOL) and the standard deviation of market model residual (VOLATILITY) as proxies for liquidity. We also include momentum (MOM) in our regression analysis, which is the 12‐month stock return from event month “−30” to “−19.” Furthermore, we control for dividend yield (DY) because short sellers are required to reimburse the security owner for any subsequent dividend declared over the period when the short position is open (Arnold, Butler, Crack and Zhang, 2005; Lasser, Wang and Zhang, 2010). Our empirical model is summarized in the following equation: …”