Dissertations, Theses, and Capstone Projects

Date of Degree


Document Type


Degree Name





Donal Byard

Subject Categories



Essay 1: Using data from StockTwits.com, the most popular investment-dedicated social media network in the US, I develop new and direct measures of investors' aggregate attention to the stock market, and to individual stocks. I then examine the behavior of investors' aggregate attention and firm-specific attention. First, I show that investors' aggregate attention to the stock market is higher during earnings season than non-earnings season, but is not lower on Fridays compared to other weekdays. Second, I show that investors' firm-specific attention is positively related to Earnings Response Coefficients (ERCs) and negatively related to Post Earnings Announcement Drift (PEAD), indicating that investor attention, as directly measured using social media data, enhances the processing of information and price discovery around earnings announcements. Finally, I find that when multiple firms announce their earnings on the same day, each earning announcement is not equally distracting: larger firms and firms with better information environment distract more attention from smaller firms.

Essay 2: Using data from StockTwits.com, the most popular social media network dedicated to the discussion of stock investment, I textually analyze the content of 11 million tweets written by both sophisticated and unsophisticated investors. From this analysis, I develop measures of the information content of sophisticated and unsophisticated investors' tweets and a daily measure of the degree of information asymmetry between these two classes of investors. Using sporadic management forecasts as a research setting, I examine the impact of public disclosures on the information asymmetry between sophisticated and unsophisticated investors. I find that: 1) information asymmetry between these two classes of investors increases in the short term (about a week) after public disclosures-specifically, sporadic management forecasts; 2) information asymmetry decreases in the longer term (roughly a week) after public disclosures; and 3) more precise public disclosures result in a smaller short-term increase (and a larger long-term decrease) in information asymmetry.

Included in

Accounting Commons