Dissertations, Theses, and Capstone Projects

Date of Degree


Document Type


Degree Name





Lin Peng

Committee Members

Zhi Da

Jian Hua

Dexin Zhou

Subject Categories

Finance and Financial Management


Social Connectedness Index, Household Finance, Investor Attention, Market Return Predictability


This dissertation consists of three chapters that span behavioral and household finance. Chapter 1 is the introduction. Chapter 2 to 4 are the main content. Lastly, chapter 5 concludes.

Chapter 2: I show that daily aggregate retail attention to firms (ARA) strongly and negatively predicts the one-week ahead market returns, whereas aggregate institutional attention (AIA) positively predicts market returns around scheduled major news announcements. Results are robust in out-of-sample tests and the effect of ARA is causal. The patterns are consistent with attention-induced retail buying which generates a transitory price pressure that quickly reverts and an increased institutional attention preceding the accrual of risk premium. The attention patterns also shed light onto the puzzling pre-announcement premium observed on days of clustered after-hours earnings announcements. Additional cross-sectional tests confirm that illiquidity and market risk exposure contribute to the predictive power of ARA and AIA, respectively.

Chapter 3: Using IRS tax filing data, I show that social network and word-of-mouth communications play an important role in stock market participation decisions. Using a novel dataset from Facebook, I construct a measure of social network friends' participation for US counties and find that a one-percentage point increase in friends' participation increases the focal county participation by 14 to 25 basis points in the following year. For identification strategy, I employ the revelation of financial misconducts as an exogenous negative shock to local participation rate and show that the instrumented change in friend participation significantly and positively predicts the change in focal county participation rates. The increase in participation rates among the low-income households induced by friends' participation decreases the Gini coefficients in metropolitan counties in the following two years. The evidence suggests that social influences and peer effects contribute to the cross-sectional differences in the stock market participation rates across US counties and may lead to lower income inequality.

Chapter 4: This chapter researches the peer effects on stock market participation decisions and portfolio characteristics using Taiwanese compulsory military service as random peer groups. I first identify compulsory military servants and military units in Taiwan using the anonymous tax return and wealth registration data in Taiwan. I examine and show the randomness in social status of the military servants and their military peers. Exploiting the random peer group design, I document peer effects on stock market participation among military peers. This effect is persistent and causal. Lastly, I find heterogeneity in peer effects on portfolio returns and level of diversification across demographics such as age, fields of study, and prior portfolio characteristics.