The effect of short-sale constraints on volatility, skewness, and kurtosis of intraday return

Abstract

This study examines the effect of short-sale constraints on volatility, skewness, and kurtosis (collectively, the distributional characteristics) of intraday return distribution. The purpose is to provide marginal evidence to address whether short-sales should be allowed to practice in financial markets. The study conducts tests during SET50 index addition and removal events. Since only SET50 index members were allowed to be sold short during 2002-2010, this constitutes a direct proxy for short-sale constraints. Furthermore, this study examines the effect of short-sale constraints on the distributional characteristics before and after earning announcement. The results show that when unconditional on earning announcement, short-sale constraints do not systematically associated with volatility, skewness, or kurtosis of intraday return. However, when the tests are conditional on earning announcement, short-sale constraints actually increase volatility, skewness, and kurtosis of intraday return. The results are similar in both before and after earning announcement. As skewness decreases along with kurtosis when short-sales are allowed, we can reconcile the results and conclude that reduction in skewness is due to fewer occurrence of extreme positive return, rather than higher frequency of extreme negative return. This is in line with the view that allowing short-sales increases arbitrageurs’ ability to correct mispricing. The implication is that regulators can prohibit short-sales during an information event if they see the need to distort intraday return distribution to be more right (positive) skewed at a price of higher volatility and kurtosis of intraday return.