Essays on and Beyond Market Microstructure

My dissertation consists of two parts. The first part examines standard microstructure topics from new perspectives. Chapter 2 investigates the market implications of high-frequency trading focusing on an ignored channel that existing market makers might reduce the capital exploited in market making facing the competition from high-frequency traders. I find that more high-frequency trading might not improve market quality exactly because of this channel. My model also generates several policy suggestions in improving the market quality with the presence of high-frequency trading. Chapter 3 studies the large traders’ execution problem to understand whether electronic trading brings execution advantage. I find that although an increase in trading frequency improves their executions, losing the ability to adopt reputation-based trading strategies due to the prevalence of anonymous trading in this electronic trading era can be so costly that large traders become worse-off in execution. Chapter 4 considers an insider trading problem with uncertainty over the insider’s existence. The competitive market making assumption is not consistent with this new element because the insider can achieve an unbounded payoff by taking advantage of the competition. Instead, I construct and analyze the equilibrium under the assumption of monopolistic market making. The model can explain the heterogeneous changes in liquidity of different stocks after a regulation.The second part of my dissertation study two elements outside the scope of current microstructure literature but (in my opinion) are important and should be incorporated to achieve a comprehensive understanding of financial markets. Chapter 5 studies an attention allocation problem under the context of contract and information design. The principal needs to jointly design compensation for the agent and his attention allocation in acquiring signals to motivate the agent to exert the unobservable effort. The flexibility to allocate attention can lead to perpetual cooperation between the principal and the agent. Chapter 6 studies belief updating when the state space contains elements that we are not aware of with an axiomatic approach. The agent can learn from what he is currently aware of and updates his belief when awareness is related to the state realization. In this case, besides leading to Bayesian updating, the arrival of a new signal can also expand the agent’s awareness and trigger further updating. Its implication is discussed under the context of persuasion.