Dissertations, Theses, and Capstone Projects
Date of Degree
9-2025
Document Type
Doctoral Dissertation
Degree Name
Doctor of Philosophy
Program
Business
Advisor
Sebastiano Manzan
Committee Members
Xi Dong
Lin Peng
Subject Categories
Finance | Finance and Financial Management
Keywords
Textual Analysis, News, Sentiment, Corporate Social Responsibility, Economic Policy Uncertainty, Behavioral Finance
Abstract
This dissertation consists of three chapters that encompass textual analysis applications in finance and economics, with a focus on firm-level news sentiment, corporate social responsibility disclosure under policy uncertainty, and market reactions to business event news.
Chapter 1: This chapter presents an introduction to the dissertation.
Chapter 2: We conduct a textual analysis of firm news, focusing on three key aspects: news sentiment, its dispersion across articles, and news coverage. We find that high sentiment stimulates trading activity, though its impact is nonlinear, with neutral sentiment having a stronger effect than either positive or negative extremes. Our evidence suggests that neutral news (lacking a clear directional signal) carries a higher proportion of fundamental information, while tail news (with a strong positive or negative tone) contains more emotional content. We also find that news dispersion significantly influences market behavior by increasing both turnover and volatility. While broader news coverage boosts trading activity, it is associated with lower stock returns. Furthermore, the effects of sentiment, dispersion, and coverage are more pronounced for small firms. These findings underscore the importance of accounting for text structure, firm size, and neutral news when evaluating the market impact of firm-level news.
Chapter 3: We examine whether firms use corporate social responsibility (CSR) disclosure as part of their human capital and corporate governance strategies in response to economic policy uncertainty (EPU). Using a large panel of U.S. firms from 2005 to 2023, we find that firms increase CSR transparency and the use of EPU-related keywords during periods of heightened macroeconomy uncertainty. In turn, greater CSR disclosure is associated with reduced firm idiosyncratic risk in these periods. We document a shift in disclosure emphasis on human capital and corporate governance when economic uncertainty rises. Furthermore, we find that firms with stronger corporate culture rely less on CSR disclosure to deal with uncertainty, suggesting a substitution effect between embedded cultural norms and external CSR disclosure. The results are robust to alternative measures of CSR disclosure and policy uncertainty.
Chapter 4: We investigate how business event news affects stock market activity using Capital IQ Key Developments data. We find that news with unclear sentiment leads to higher trading volume and volatility, while extreme positive or negative news has a smaller effect. Stock returns increase linearly with the strength of the sentiment. Under high economic policy uncertainty (EPU), the return impact of sentiment becomes stronger, and investors react more to positive news. When individual sentiment differs from overall market sentiment, trading and volatility decrease, suggesting more cautious behavior. However, returns continue to follow individual sentiment regardless of market sentiment.
Chapter 5: This chapter concludes the dissertation.
Recommended Citation
Kong, Mingyuan, "Essays on Textual Information and Financial Markets: Evidence from News Sentiment, Corporate Disclosure, and Market Dynamics" (2025). CUNY Academic Works.
https://academicworks.cuny.edu/gc_etds/6364
