Short Sellers as Informed Traders

Abstract

The purpose of this thesis was to examine whether short sellers are informed traders. To measure this, a three-fold approach including several different methods was applied. Firstly, the relation between short interest and stock returns is examined to see whether short sellers predict future abnormal returns. Secondly, the short interest is inspected by adding firm-specific controls to the model. Thirdly, the relation between shorting and trading volume is examined to provide additional information of short sellers’ actions.

The study sample consisted of return and short interest data from 85 New York Stock Exchange listed S&P100 stocks between 15 January 2010 and 30 September 2011. Daily returns were used to form bimonthly return figures to match the short interest announcing frequency, which is bimonthly. Returns were examined by measuring raw returns. Additionally market adjusted returns were used as a robustness check. The relation between short selling and returns was examined by using both linear and cross-sectional regression models.

The main findings of the study were consistent with previous literature in that short sellers seem to be informed in their trading activities. An inverse relationship between increased short selling and future returns was found suggesting that an increase in shorting predicts negative abnormal returns. In addition, after controlling for market, size and book-to-market factors, short selling was found to be a significant return determinant. The negative relation found in the cross-sectional regression model reinforces the findings of the linear model. The empirical evidence supported the information hypothesis, which implies that short sellers are informed traders.