Research
Working Papers
We explore multiple rational bubbles with varied bursting probabilities. Post-bursting, the sector contracts, while surviving bubbles gain value. Lower bursting odds lead to larger equilibrium bubbles, expanding the sector and deflating others. Fragile bubble assets, collectively, might rival a stable one due to diversification benefits. Additionally, our stationary model establishes the upper equilibrium limit, applicable to models handling price fluctuations, nonstationarity, and potential bubble formations in financial ecosystems. Our model has implications for crypto ETFs, CBDC issuance, and cryptocurrency regulations, illuminating diverse bubble dynamics.
I study market quality implications of the competition between traditional market making and high-frequency trading. A long-run traditional market maker responds to the competition from high-frequency traders by reducing both the spread and the amount of capital committed to market making. While a lower spread level is beneficial, less capital commitment deteriorates market quality. Specifically, the market's capacity to satisfy large demand is impaired. My model provides a more comprehensive illustration of high-frequency trading's implications on market quality by integrating both price and quantity effects. I further use this framework to analyze the implications of different high-frequency trading regulatory measures.
R&R at Review of Economic Studies
We study a dynamic contracting problem in which the principal can allocate his limited capacity between seeking evidence that confirms or that contradicts the agent's effort, as the basis for reward or punishment. Such flexibility calls for jointly designed monitoring and compensation schemes practically relevant but novel in the literature. When the agent's continuation value is low, the principal seeks only confirmatory evidence, but when the agent's continuation value exceeds a threshold, the principal switches to seeking mainly contradictory evidence. Moreover, the agent's effort can be perpetuated if and only if both synergy and flexibility in monitoring are sufficiently large.
We study trading the execution strategies of a large uninformed trader over the transition from human-based markets to electronic markets, by focusing on two elements of this transition: (1) the increase in trading frequency, and (2) the loss of commitment power. We develop a multi-period model with symmetric information where an uninformed trader must liquidate her position to several market makers. A trader without commitment power is unambiguously better off from an increase in trading frequency. A trader that exchanges commitment power for more frequent trading is worse off. Commitment power is especially valuable if it can be coupled with frequent trading.
Abstract: We study the interaction between the usual inside information (about asset values) and the information about its existence (i.e., about the existence of insiders) in an otherwise standard continuous-time Kyle-Back model. Interestingly, only the inside information conditional on its existence, rather than the information about its existence, is revealed in equilibrium and affects asset prices and market liquidity, and it is revealed asymptotically in the long run. Our model sheds light on the impact of stock market regulations on market liquidity and reconciles the relevant mixed empirical findings.
Abstract: This paper introduces a novel channel of belief updating with awareness. Awareness is represented by agents having state-dependent subjective state spaces. Moreover, an agent can form conjectures of his subjective state spaces under different state realizations. He updates his belief over the true state realization by excluding states in which his conjectured subjective state spaces are different from his current state space. We restrict agents' capacities of belief updating through awareness by imposing axioms on agents' conjectures. We further use this framework to analyze belief updating when an agent's state space expands. Importantly, when belief is updated with awareness, the belief process is not necessarily a martingale. We use persuasion games as examples to illustrate this point. Since the receiver's belief process induced by awareness signaling may not be a martingale, the sender might be better off sending an awareness signal than performing any feasible Bayesian experiment.