Working Paper: CEPR ID: DP3058
Authors: Qing Li; Maria Vassalou; Yuhang Xing
Abstract: In this paper we present a simple model where asset returns are functions of multipleinvestment growth rates. The model is tested for its ability to price the 25 Fama-Frenchportfolios using the Generalized Methods of Moments (GMM) methodology, as well asFama-MacBeth cross-sectional regressions. Comparisons on the basis of several metricswith other models, such as the CAPM, the Fama-French (1993) model and Cochrane's(1996) model, reveal that it consistently outperforms the CAPM and Cochrane's model. Italso outperforms the Fama-French model in several tests. Our model can explain asignificantly larger proportion of the cross-sectional variation in the 25 Fama-Frenchportfolios than the Fama-French model does. Specification tests in the context of GMMand the Fama-MacBeth regressions show that in the presence of the investment growthfactors included in our model, the size and book-to-market characteristics lose theirability to explain asset returns. Our model is successful in pricing size- and book-to-market-sorted portfolios, although it includes exclusively macroeconomic variables asfactors.
Keywords: GMM
JEL Codes: G12
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Investment growth rates (E20) | Asset returns (G19) |
| Investment growth rates (E20) | Future GDP growth (O49) |
| Investment growth rates (E20) | Cross-sectional variation in asset returns (G19) |
| Investment growth factors absorb priced information in Fama-French model (G31) | Loss of explanatory power in Fama-French model (C29) |