DISSECTING THE SHORT SQUEEZE

Abstract

In this thesis, I conduct an empirical study of the phenomenon called the “short squeeze”. I argue that a short squeeze is more likely to happen when three conditions are met: 1) the stock must be under relatively high level of short interest; 2) there must be a binding capital constraint on short sellers, potentially as a result of margin calls; 3) the stock must have binding short-sale constraint so that short sellers cannot cheaply re-establish their short positions. I generate the exact historical short interest level report dates based on NASDAQ documentation rather than assuming the 15th of each month as the report dates for this thesis to avoid sample error, and I find heavy short squeeze activities happen before the official dates where exchanges make the data available for the public. I propose a model that estimates the likelihood of a short squeeze ex ante with 79.6% accuracy. A long-short strategy that purchases (shorts) stocks with a high (low) probability of a short squeeze generates a return of 44.5% per year.