Author: Anqi Liu
Advisor: Dr. Steve Yang
Date: April 26, 2017
Department: Financial Engineering
Degree: Doctor of Philosophy
Advisory Committee:
Dr. Steve Yang, Chairman
Dr. Zhengyu Cui,
Dr. Babak Heydar,
Dr. Khaldoun Khashanah,
Dr. Mark E. Paddrik
Abstract: Information digitization and advanced data mining techniques have increased popularity of using business news, social media messages or other sources of information in trading and investment decisions. Investigating investors’ responses to both internal and external information is a mainstream of research in behavioral finance per which sentiment analytics is the key to quantify investor perspectives and to evaluate influence of trading activities. Moreover, a series of studies ar gued that investor sentiment can become systematic and hence has broad impact to financial systems. This dissertation explores information propagation in financial markets based on sentiment transition in investor community, market price/investor sentiment shocks transmission, and sentiment-drive/price-driven information flow in equity markets. First, by applying agent-based framework to model social traits of sentiment transition, we conclude that information propagation across investor community facilitates formation of homogeneous social beliefs about financial markets. Second, we identify shocks in market price and business news and model arrival of these events through Multivariate Hawkes point processes which repre sent interactive responses in financial markets. We document that there are cross excitation effects between market returns and investor sentiment. Generally, price jumps affect social beliefs about potential financial markets performance, and in return, trading decisions based on investor prospects further influence price movements. Third, we propose a novel approach to identify irrational trading behavior using information entropy. We introduce corresponding news-driven and price driven trading activities in the irrationality identification experiments.
Overall, this dissertation introduces a behavioral approach to model financial markets dynamics under different rational vs. irrational regimes. It contributes to field of behavioral finance at both micro and macro level. By interpreting market responses to investors’ behaviors, we conclude that financial markets perform as an ecosystem which evolve with investors’ adaptive trading behaviors.
For full Dissertation, click here.