Short Sale Constraints: Effects on Crashes, Price Discovery, and Market Volatility

Abstract: The recent SEC ban on short selling has presented an unrivaled opportunity to explore the effects of short selling constraints on crashes, market efficiency, and volatility. In this paper I carry out two groups of empirical tests on the individual banned stocks and a series of portfolios created from them: the first tests the hypothesis that short sale constraints increase the frequency and magnitude of crashes, by testing Hong & Stein’s (2003) model of market crashes. The second tests the hypothesis that short sale constraints reduce market efficiency, by testing Miller’s (1977) model in which stocks that are hard (or impossible) to short tend to exhibit overpricing. In regards to the first group of tests, the results are ambiguous: the frequency and magnitude of crashes increased during the ban period, while the skewness of the returns distribution of the portfolios became more negative, as expected, but these changes hold for the market as a whole, as well. On the other hand, the skewness of the returns distribution of the individual banned stocks became more positive. The second group of tests provides ample support for Miller’s model, as the results coincide with the models predictions: banning short selling leads to positive abnormal returns (overpricing) in the affected stocks. The ban is also related with a decrease in volatility relative to the market, an important result from a policy perspective.