Publication Type
PhD Dissertation
Version
publishedVersion
Publication Date
4-2025
Abstract
My dissertation consists of three studies on how investor behavior affects the outcomes in financial markets. The first study focuses on the investors’ limited capacity on processing the complicated information transmitted through interconnected firms. The other two studies explore the investor heterogeneity on preferred time of trading, due to their personal traits or irrational judgements on the market conditions.
In Chapter 2, I focus on the return predictability across economically linked firms, due to investor limited attention. I document significant outperformance of stocks whose industry peers currently experience salient (attention-grabbing) gains, especially for those with strong social ties. The long-short strategy based on this effect can earn a risk-adjusted return of 12.5% (t = 5.58) per year. The pricing results differ from well-known crossstock momentum and can’t be attributed to risk-based explanations. The effect is more pronounced when focal firms are more centered on social networks, receive less attention, or are faced with greater limits of arbitrage. I show that these findings align with categorical thinking, where the salience of peer firm performance distracts investor attention from the focal firm, leading to investor underreaction. My study offers an approach to establishing a microfoundation of investor behavior in explaining the return co-movement among economically linked firms.
The following two chapters examine the heterogeneity of investor behavior by digging into the puzzling “overnight-intraday return gap”. Chapter 3 shows that retail trading proportion (RTP) is a major explanatory variable for this puzzle by using data from Korea where accurate and exhaustive retail trade flows are available. To establish a causal relationship between the return gap and RTP, the study uses an instrumental variable approach that exploits retail investors’ tendency to more actively traded stocks with low per-share prices. I attribute this relationship to the recurrent retail ebb and flow stemming from the retail investors’ demand for daytime stock market exposure.
Chapter 4 digests the "night-day" puzzle by revealing the role of trading volume. The study documents a novel high-volume overnight premium (intraday discount). Specifically, stocks with high trading volume exhibit remarkable future outperformance (underperformance) during the overnight (intraday) period. This phenomenon is also prevalent in global equity markets, robust to alternative measures of trading volume, and persistent for at least three years after portfolio formation. In a tractable equilibrium model, it is shown that such a heterogeneous night/day effect is the result of the cycling pattern of belief dispersion being high/low at market open/close, due to the overconfidence of late-informed investors. To justify this idea, this study uses high frequency returns and documents the decreasing cross-sectional dispersion of market betas over the intraday period is especially strong for high-volume stocks. Several testable model predictions are proposed and empirically shown that our main results are more pronounced when uninformed investors are more optimistic, or more dominant in overall trading activity.
Degree Awarded
PhD in Business (Finance)
Discipline
Finance | Finance and Financial Management
Supervisor(s)
HUANG, Dashan
First Page
1
Last Page
208
Publisher
Singapore Management University
City or Country
Singapore
Citation
FAN, Qi.
Essays on empirical asset pricing. (2025). 1-208.
Available at: https://ink.library.smu.edu.sg/etd_coll/686
Copyright Owner and License
Author
Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.