Short Sale Constraints and the Likelihood of Crashes and Bubbles

Abstract

In this paper, we investigate short sale constraints' impact on the incidence of extreme stock market movements. The latter can be used to proxy for the likelihood of tail events like crashes and bubbles in a market and, thus, is a crucial measure of stock market stability. Since crashes and bubbles are, almost by definition, unpredictable, we, unlike scarce prior research which relies on simple descriptive statistics, address only the component of the return which hits investors unexpectedly. To this end, we rely on long lasting short selling regimes in 3 Asian markets which, unlike the short-lived bans analyzed in existing studies, provide us with a setting to consistently estimate sophisticated time series models for the market return. Our evidence suggests that, during some market phases, short sale restrictions lead to an increased kurtosis of pricing errors which, in turn, indicates a higher probability for tail events.

Keywords: Short Selling Ban, Crashes, Bubbles, Kurtosis, Markov Switching GARCH

JEL Classification: G10, G12, G14, G15, G18