Working Paper: CEPR ID: DP18094
Authors: Frederico Belo; Yao Deng; Juliana Salomao
Abstract: Abstract Most investment-based asset pricing models predict a close link between a firm's stock return and firm-characteristics at any point in time. Yet, previous work typically examines the weaker prediction that this link should hold on average. We show how to incorporate the time-series predictions in the estimation and testing of investment-based models using the generalized method of moments. We find that standard specifications of the investment-based model with one physical capital input fail to match the time series properties of stock returns in the data, and discuss the implications of the findings for future research.
Keywords: Production-based asset pricing; Model testing; Estimation
JEL Codes: G11
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| firm characteristics (L20) | stock returns (G12) |
| investment-based asset pricing model (G12) | stock returns (G12) |
| increasing weight of time series moments (C22) | deterioration in fit of cross-sectional moments (C51) |
| aggregation bias and data misalignment (C43) | discrepancies in stock return behavior (G40) |