« Back to Results

The Cross-Section of Stock Returns

Paper Session

Saturday, Jan. 8, 2022 10:00 AM - 12:00 PM (EST)

Hosted By: American Finance Association
  • Chair: Richard Sias, University of Arizona

Open Source Cross-Sectional Asset Pricing

Andrew Chen
,
Federal Reserve Board
Tom Zimmermann
,
University of Cologne

Abstract

We provide data and code that successfully reproduces nearly all cross-sectional stock return predictors. We collect 319 characteristics from previous meta-studies, and compare reproduced t-stats to the original papers' results. For the 161 characteristics that were clearly significant in the original papers, 98% of our long-short portfolios find t-stats above 1.96. For the 44 characteristics that had mixed evidence, our reproductions find t-stats of 2 on average. A regression of reproduced t-stats on original long-short t-stats finds a slope of 0.90 and an R^2 of 83%. Mean returns are monotonic in predictive signals at the characteristic level. The remaining 114 characteristics were insignificant in the original papers or are modifications of the originals created by Hou, Xue, and Zhang (2020). These remaining characteristics are almost always significant if the original characteristic was also significant.

Ambiguity, Investor Disagreement, and Expected Stock Returns

Lawrence Hsiao
,
Northwestern University

Abstract

I set up a disagreement model in which traders not only have different interpretations of a public signal that conveys information of a stock, but are also uncertain about the information quality of others' interpretations. The model along with traders being ambiguity-averse predicts a positive relation between investor disagreement (ID) and expected stock return. Consistent with the model's prediction, I find that stocks in the highest ID decile outperform stocks in the lowest ID decile by 9.24 percent annually, adjusted for exposures to the market return as well as size, value, momentum, and liquidity factors. In addition, I find that stocks with higher ID prior to earnings announcements earn significantly higher earnings announcement returns.

Fundamental Extrapolation and Stock Returns

Dashan Huang
,
Singapore Management University
Huacheng Zhang
,
Southwestern University
Guofu Zhou
,
Washington University in St. Louis
Yingzi Zhu
,
Tsinghua University

Abstract

We study both a naive extrapolation from individual fundamentals and pooling extrapolations from all fundamentals simultaneously. We find that, while naive extrapolation generates no alphas except one case, pooling extrapolations generate economically and statistically significant alphas larger than those extrapolated from analysts' forecasts. Since our extrapolations have positive slopes at the common monthly frequency unexplained by existing theories, we provide a model to show that the fundamental extrapolation has dual effects on stock price: a cash flow effect and a discount rate effect. The former pushes stock price up relative to its fundamental value, whereas the latter increases the representative investor's expected volatility and depresses today's stock price. Our empirical results suggest that the discount rate effect dominates the cash flow effect.

Sectoral Labor Reallocation and Return Predictability

Esther Eiling
,
University of Amsterdam
Raymond Kan
,
University of Toronto
Ali Sharifkhani
,
Northeastern University

Abstract

Sectoral labor reallocation shocks change the optimal allocation of workers across industries. We show that the cross-sectional dispersion in industry-specific stock returns (CSV) serves as a good proxy for sectoral reallocation shocks. This measure significantly predicts aggregate unemployment growth, as well as an ex-post measure of the mismatch between job seekers and vacancies across sectors.

We find that CSV is a remarkably strong and robust predictor of future stock market returns. In predictive regressions, the one-year out-of-sample R-squared is as high as 12.56%, outperforming a long list of alternative variables and leading to large economic gains.

We propose a production-based asset pricing model in which sectoral labor reallocation shocks generate stock return predictability through time-varying exposure to aggregate productivity risk. When the need for labor reallocation across industries arises, industries are more likely to hire workers from other industries. This involves more search and training costs. Hence, industries face higher labor adjustment costs, which impedes them from fully responding to aggregate economic fluctuations. In turn, their aggregate risk exposure decreases leading to lower expected returns.

The model provides new testable implications which are supported by the data. For example, we show that the predictive power of CSV is concentrated among industries with the strongest dependence on high-skill labor, which is costlier to replace. We also rule out a number of alternative explanations, including capital adjustment costs, investor over- or underreaction and slow information diffusion.

Discussant(s)
Jeffrey Pontiff
,
Boston College
Jon Garfinkel
,
University of Iowa
Huseyin Gulen
,
Purdue University
Harry Turtle
,
Colorado State University
JEL Classifications
  • G3 - Corporate Finance and Governance