Short Interest: Explanations and Tests

Abstract

Cross-sectional and time series tests are performed to explain levels and changes in short interest. Explanatory variables and tests are chosen based on tax, arbitrage, and speculative reasons for going short. Short interest is found to follow a seasonal pattern that is weakly consistent with tax-based trading. Stocks with high betas and the existence of convertible securities or options tend to have higher levels of short interest, which is con? sistent with arbitrage efforts. For firms with traded options, there is a positive association between the month-to-month changes in option open interest and short interest. Prior months' returns and changes in short interest are positively related, but there is no rela? tionship between changes in short interest and returns in the subsequent month.