Lewellen, Nagel, and Shaken give us a Skeptical Appraisal of Asset-Pricing Tests
Short version: we really do not have much of an idea as to what asset pricing model is correct and to make matters worse existing tests may not be showing what we think they do.
A few quick views in the authors' words:
"The finance literature has offered an explosion of new asset-pricing models in recent years, motivated by evidence that small, high-B/M stocks have positive CAPM-adjusted returns. The new models – formal equilibrium theories as well as simple econometric models – propose a variety of risk factors beyond the market return or aggregate consumption"Maybe the key paragraph in entire paper:
"Empirically, many of the proposed models seem to do a good job explaining the size and B/M effects, an observation at once comforting and disconcerting: comforting because it suggests that rational explanations for the anomalies are readily available, disconcerting because it provides an embarrassment of riches. Reviewing the literature, one gets the uneasy feeling that it seems a bit too easy to explain the size and B/M effects. This is especially true given the great variety of factor models that seem to work, many of which have very little in common with each other."In the authors' own words:
"Our paper is motivated by that suspicion. In particular, we explain why, despite the seemingly strong evidence that many proposed models can explain the size and B/M effects, we remain unconvinced by the evidence."And then:
"The heart of our critique is that the literature has often given itself an extremely low hurdle to meet in claiming success: high cross-sectional R2s (or low pricing errors) when average returns on the Fama- French 25 size-B/M portfolios are regressed on their factor loadings. This hurdle is low because size and B/M portfolios are well-known to have a strong factor structure, i.e., the Fama and French’s (1993) three factors explain more than 90% of the time-series variation in portfolios’ returns and more than 75% of the cross-sectional variation in their average returns. Given those features, obtaining high cross-sectional R2 is very easy because almost any proposed factor is likely to produce betas that line up with expected returns...."Then after giving suggested solutions (sorting on other factors and emphasizing "theoretical restictions", and reporting confidendence intervals) to these problems, the authors test popular asset-pricing models.
Unfortunately they find little support for any of the asset pricing models:
"We apply these prescriptions to a handful of proposed models from the recent literature. The results are disappointing. None of the five models that we consider performs well in our tests, despite the fact that all seemed quite promising in the original studies."Uuch. Well, I guess it's back to the drawing board.
Lewellen, Jonathan W., Nagel, Stefan and Shanken, Jay A., "A Skeptical Appraisal of Asset Pricing Tests" (January 2006). AFA 2007 Chicago Meetings Paper Available at SSRN: http://ssrn.com/abstract=891434