Short interest on NASDAQ: Good news, bad news, or no news

I study the information content in short interest data using NASDAQ-listed stocks from 1988 to 2004. I show that stocks with relatively high short interest subsequently experience negative abnormal returns. However, this effect is generally limited to larger stocks and is of debatable economic significance. In contrast, I find that intensively traded stocks with low short interest experience both statistically and economically significant positive abnormal returns. These positive returns are much larger (in absolute value) and more persistent than the negative returns observed for highly-shorted stocks. I also empirically test Diamond and Verrecchia's (1987) theory on the information content in unexpected short interest. I find that large increases (decreases) in short interest relative to turnover implies smaller (greater) expected stock returns for the subsequent month. Because I show that stocks with greater short interest are less mispriced, I provide support for the hypothesis that short selling promotes market efficiency. However, I also show that both positive and negative information in publicly-available short interest data is only slowly incorporated into prices, thereby raising a significant market efficiency issue. My results also cast serious doubt on existing theories of the impact of short-sale constraints and the information asymmetry argument.; Keywords. Market efficiency, price correction, short interest, short-sale constraints, liquidity, information asymmetry.