2381/39922 Xin Li Xin Li Essays in Behavioral and Computational Finance University of Leicester 2017 IR content 2017-06-20 09:02:15 Thesis https://figshare.le.ac.uk/articles/thesis/Essays_in_Behavioral_and_Computational_Finance/10202375 This thesis consists of two essays on behavioral finance and financial market microstructure with computational approaches. Chapter 2 investigates the effects of steroid hormones and trader composition on financial markets in a mathematical model. We focus on the composition of traders in financial markets, namely, female traders and male traders, as risk preferences change in different ways with the mediation of steroid hormones. Firstly, we examine the effects of testosterone on financial risk preferences and market stability in the model. The results from simulation show that the effects of a more balanced gender composition are more nuanced. An increase in the proportion of female traders may actually increase the volatility of returns; however, the chances of extreme events are reduced. Secondly, we analyze the effects of cortisol on traders' risk preference and market behavior in our model with traders' risk preferences influenced by market uncertainty via the mediation of cortisol. Results from our model show that concerns about heightened market uncertainty mitigate traders' excessive risk-taking behaviors and performance of traders is largely affected by market sentiment. In the third part of Chapter 2, we examine the overall effect of testosterone and cortisol on market behavior with traders having heterogeneous behavioral and physiological responses to trading outcomes and market uncertainty. Results from simulation show that male-dominated market is less volatile as the effect of concerns about market uncertainty outweighs the effect of trading outcomes on traders' behavior. Chapter 3 examines the impact of two different types of information on high frequency market microstructure. We present a dynamic trading game in the limit order market with computerized traders and human traders trading in one risky asset, where traders might have lags in observing the contemporaneous fundamental value and the order book status. Optimal strategies and market characteristics are determined through a unique numerical technique. Our results show that these two types of information have different values for traders with information on contemporaneous fundamental value being more valuable than the information on contemporaneous limit order book status.