Investor Sentiment and Market Reaction to 8-K Filings

Yuexi Ding

College: College of Business and Public Management

Major: BS.ACCOUNTING

Faculty Research Mentor: Kakolyris, Andreas

Abstract:

This study examines whether social media attention amplifies short-term market reactions to earnings-related Form 8-K disclosures and whether the effect depends on how attention is measured. Focusing on a selected group of S&P 500 technology firms, we analyze Item 2.02 (“Results of Operations and Financial Condition”) filings over the sample period 2018–2023. Event dates are obtained from Bloomberg, and short-horizon market reactions are measured using cumulative abnormal returns around the disclosure window.Investor attention is proxied using StockTwits messages aggregated around each event. We distinguish between total attention (all messages) and human attention by filtering accounts with bot-like characteristics based on activity patterns and account features. This allows us to construct total message counts, human-only message counts, and bot ratios for each firm-event observation.We test whether higher pre-disclosure attention predicts stronger immediate price reactions and whether total attention produces larger estimated effects than human-only attention. By comparing regressions using alternative attention measures, the study evaluates whether automated or low-credibility participation inflates observed attention–reaction relationships. The results contribute to the investor attention literature by showing how measurement choices in social media data can affect inference about market efficiency and price discovery around corporate disclosures.