Dissecting Arbitrage Costs

Abstract

We dissect the relative importance of nine commonly used measures of arbitrage costs on the extent of mispricing in the cross-section of stock returns. Using the relative valuation index developed in Stambaugh, Yu, and Yuan (2012) to measure mispricing, we find that arbitrage risk, measured using idiosyncratic risk, is the predominant limiting arbitrage cost. Information uncertainty and transactions costs are also important limiting factors to arbitrage, but only among small stocks. A principal components analysis shows that arbitrage costs can be reduced into two significant dimensions. Overall the evidence suggests that arbitrage costs explain much of the average returns associated with stock anomalies.