We ask whether risk or mispricing is the main driving force behind capital markets anomalies, which are empirical relations between average returns and firm characteristics not explained by standard asset pricing models. Over the past three decades, anomalies have become important in asset allocation, capital budgeting, security analysis, hedge fund strategies, and many other applications. Understanding their driving forces is one of the most important questions in capital markets research.
Two competing schools of thoughts have proposed an array of economic explanations for capital markets anomalies (see Appendix A for a brief review). Behavioral finance contends that investors make systematic mistakes in pricing assets, and that anomalies are driven by predictable pricing errors (mispricing) (e.g., Barberis, Shleifer, and Vishny (1998), Daniel, Hirshleifer, and Subrahmanyam (1998), and Hong and Stein (1999)). In contrast, retaining the assumption of rational expectations, new classical finance argues that risk and expected returns vary with firm characteristics in a systematic way, and that anomalies are driven by time-varying expected returns (risk) (e.g., Cochrane (1996), Berk, Green, and Naik (1999), Zhang (2005), and Liu, Whited, and Zhang (2009)).
We aim to further the risk versus mispricing debate by estimating accounting-based expected returns to anomalies-based trading strategies. The basic idea is simple. Average realized returns equal average expected returns plus average unexpected returns. If anomalies are driven by risk, average expected returns should account for the bulk of average realized returns. If anomalies are driven by mispricing, average unexpected returns should account for the bulk of average realized returns. Building on the latest accounting literature on expected return estimation, we use the residual income model to estimate expected returns for zero-cost trading strategies formed on a comprehensive list of anomaly variables. Under the model, the expected return can be calculated as the internal rate of return that equates the present value of expected future residual incomes to the current stock price (e.g., Gebhardt, Lee, and Swaminathan (2001)). Using accounting-based models has the added advantage that accounting identities always hold regardless of investor rationality or lack thereof.
Our key message is that expected return estimates differ drastically from average return estimates for most anomalies, suggesting that mispricing, not risk, is the main driving force behind anomalies. In particular, the expected return estimate of the value minus-growth quintile is 6.3% per annum, which is close to the average return estimate of 5.2% in terms of economic magnitude. The expected return estimate, which is about 12 standard errors from zero, is also more precise than the average return estimate, which is slightly more than two standard errors from zero. The expected return estimate of the small-minus-big quintile is 3.1%, which is close to the average return estimate of 3%. The expected return estimate is more than 7.5 standard errors from zero, whereas the average return estimate is within one standard error of zero.
Download
Does Risk Explain Anomalies? Evidence from Expected Return Estimates
