Document Type

Dissertation

Date of Degree

Spring 2011

Degree Name

PhD (Doctor of Philosophy)

Degree In

Business Administration

First Advisor

Ashish Tiwari

Abstract

In this dissertation, I consider a range of topics in cross-sectional asset pricing. The primary research focus is twofold. First, I provide new insights on analyzing and testing capital market "anomalies" or patterns in equity returns that are not well explained by the traditional models used in the finance literature. Second, I propose and examine a methodology for pooling asset-pricing models to better characterize the cross section of stock returns.

The first chapter offers an explanation for the financial distress anomaly, i.e., the previously documented poor stock-price performance for financially distressed firms. I first show that market betas for distressed firms are highly volatile and tend to be low during bad economic times. After properly controlling for exposure to market risk, the low historical returns on these stocks are consistent with the conditional Capital Asset Pricing Model (CAPM). I then explain these findings through a theoretical model in which a levered firm's equity beta is negatively related to uncertainty about the unobserved value of its underlying assets. Empirical tests support the main predictions of the theory.

The second chapter proposes a hierarchical Bayes approach for evaluating and testing asset-pricing anomalies using individual firms as test assets. The empirical results indicate that much of the anomaly-based evidence against the CAPM is overstated. Anomalies are primarily confined to small stocks, few characteristics are robustly associated with CAPM alphas out of sample, and most firm characteristics do not contain unique information about abnormal returns.

Lastly, the third chapter proposes a new econometric methodology to combine predictive densities from a set of competing asset-pricing models to better characterize the cross section of stock returns. Using a variety of test portfolios, the optimal pool of models consistently outperforms the best individual model on both statistical and economic grounds.

Pages

viii, 159 pages

Bibliography

Includes bibliographical references (pages 151-159).

Copyright

Copyright 2011 Michael Shane O'Doherty

Share

COinS