Short Constraints, Difference of Opinion and Stock Returns

Abstract

We present a new approach to analyzing the effect of short constraints on stock returns. Starting from the theory, we measure how short constrained a stock is by looking at month end outstanding short positions and other measures that proxy for heterogeneity in beliefs and ease of shorting a stock. Our proxies include analyst's forecast dispersion, institutional holdings, turnover (among others). Using monthly data for the period 1992 to 2000 on NASDAQ and NYSE stocks, we form portfolios using our measure of short constraints. We find that portfolio of stocks that are short constrained earn lower returns. We also observe that the most constrained portfolio earns a significant negative one month ahead abnormal return, after accounting for risk. One can interpret this evidence as being supportive of the common notion that constrained stocks might be priced higher today. We believe our work provides evidence that short constraints matter for asset returns. It also confirms the notion that arbitrage can sometimes be rendered ineffective in financial markets because of market frictions like short constraints which can lead to abnormal returns in stocks that are constrained. This is consistent with the idea of Limited Arbitrage of Shleifer and Vishny.