Working Paper: NBER ID: w12360
Authors: Jonathan Lewellen; Stefan Nagel; Jay Shanken
Abstract: It has become standard practice in the cross-sectional asset-pricing literature to evaluate models based on how well they explain average returns on size- and B/M-sorted portfolios, something many models seem to do remarkably well. In this paper, we review and critique the empirical methods used in the literature. We argue that asset-pricing tests are often highly misleading, in the sense that apparently strong explanatory power (high cross-sectional R2s and small pricing errors) in fact provides quite weak support for a model. We offer a number of suggestions for improving empirical tests and evidence that several proposed models don't work as well as originally advertised.
Keywords: Asset Pricing; Empirical Tests; Cross-Sectional Analysis
JEL Codes: G12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| high cross-sectional R-squared values (C21) | misleading conclusions about the effectiveness of asset-pricing models (G19) |
| strong factor structure of size and book-to-market portfolios (G32) | high cross-sectional R-squared values (C21) |
| high cross-sectional R-squared values (C21) | weak support for models (C52) |
| theoretical restrictions on slopes ignored (C20) | diminishes economic significance of findings (F62) |
| using GLS instead of OLS (C20) | more accurate assessment of model performance (C52) |
| reporting confidence intervals for R-squared values and pricing errors (C59) | better convey uncertainty and sampling variability (C46) |
| including broader set of test assets (G19) | more reliable results (C90) |