Working Paper: NBER ID: w22894
Authors: Juhani T. Linnainmaa; Michael R. Roberts
Abstract: Using data spanning the 20th century, we show that most accounting-based return anomalies are spurious. When examined out-of-sample by moving either backward or forward in time, anomalies' average returns decrease, and volatilities and correlations with other anomalies increase. The data-snooping problem is so severe that even the true asset pricing model is expected to be rejected when tested using in-sample data. Our results suggest that asset pricing models should be tested using out-of-sample data or, when not feasible, by whether a model is able to explain half of the in-sample alpha.
Keywords: stock returns; accounting-based anomalies; data snooping
JEL Codes: G11; G12; G14
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| accounting-based return anomalies (G14) | spurious (Y90) |
| sample period (C41) | average returns (G12) |
| sample period (C41) | volatilities (C58) |
| sample period (C41) | correlations with other anomalies (C10) |
| unmodeled risk or mispricing (G19) | anomalies (Z13) |
| data snooping (C52) | accounting-based return anomalies (G14) |
| insample period (G14) | presample and postsample periods (C83) |