Date of Award

Spring 5-15-2021

Author's School

Graduate School of Arts and Sciences

Author's Department

Business Administration

Degree Name

Doctor of Philosophy (PhD)

Degree Type



The 2008 financial crisis has highlighted the challenges faced by financial systems in aggregating information, ensuring coordination among various market participants and providing adequate liquidity. In this backdrop, the three chapters of the dissertation explore (i) the role of disagreement in enabling communication and trade among strategic investors (ii) how uncertainty about what others know could be exploited in preventing coordination failures (ii) pricing liquid and illiquid assets in response to unforeseen liquidity demand.

When do competing traders, endowed with different pieces of information pertaining to a security payoff, exchange information before trading? The first essay shows that competing traders share information when they disagree enough. Traders can lose competitive rents by sharing private information, but with sufficient disagreement, they can engage in profitable belief arbitrage by trading against each other's signal. Traders, however, would gain by over-reporting their signals so that competitors make large opposing trades. When information is verifiable, truthful disclosure emerges due to an “unraveling” argument. Mediators (say, sell-side analysts or brokers) could facilitate partial information sharing by aggregating and distributing information in an incentive compatible manner. Disagreement makes the market more liquid, but information sharing undermines the liquidity benefits.

IPOs, currency attacks, bank runs and stag hunt games all admit a sub-optimal equilibrium in addition to a superior equilibrium where agents co-ordinate on a Pareto optimal, but risky action. The second essay studies situations where players are uncertain about what others know, and analyzes how this second order uncertainty can be useful for achieving coordination among agents. Holding first order beliefs fixed and modifying second order uncertainty, the policymaker designs simple information structures with contagion to induce his preferred action as often as possible. The policymaker chooses information structures where states in his favor are allowed to infect other states, while states that are not in his favor are quarantined. In a stylized bank run model, a bank manager allows perfect information when bank fundamentals are weak, but allows agents to be uncertain about the other agent's knowledge in all other scenarios. This provides a rationale for accounting conservatism in banking: all probable loan losses are publicly recorded when they are discovered, while knowledge of investment gains are privately dispersed among agents.

The third essay studies a model of cash management using a portfolio of assets with varying degrees of liquidity. The agent faces a probabilistic interim liquidity shock and chooses which assets to liquidate. Assets with high liquidation value are valuable in states when the liquidity shock is high or there is a scarcity of other liquid assets in the economy. The downsides of such assets are protected by their expected terminal value, which delivers an option-like convex pricing rule: the price of a liquid asset is an American put option with strike price process given by the liquidation value in each period. The model in this essay explains the empirically observed convex liquidity premium even with a single representative agent.


English (en)

Chair and Committee

Philip H. Dybvig Jonathan Weinstein

Committee Members

Brett Green, Armando Gomes, Mina Lee,