Working Papers
“Dynamic Survival Bias in Learning from Doubly Censored Data” (Job market paper)
This paper studies a survival bias that emerges when an analyst seeks to learn about a random payoff from observing a binary sequence of doubly censored signals. There's firstly a joint right censoring such that a signal can only be observed if its precedents jointly passes a threshold. The second censoring is a truncation in time such that the analyst only sees more recent signals but not the historical ones. The bias is dynamic in the sense that the inference about the un-observed signals evolves as the observation continues. I characterize the analyst’s posterior belief about the random payoff adjusted for the bias by considering all the possible historical signals consistent with the censoring process.
“Contract Design in Self Investment Products” (with Can Sun)
This paper studies a new business model in the self-investment goods market, in which the firm proposes a two-part tariff composed of an up-front fee and a later rebate. This contract exploits consumers’ time-inconsistent preference by over-charging the up-front fee and makes profit by its crowdfunding nature.
Work in Progress
“Population Game with Strategic Substitution”
This paper studies a less-understood class of games --- population games with heterogeneous agents and a payoff structure that exhibits strategic substitution. It establishes the special equilibrium existence and uniqueness conditions and develops the comparative statics results accordingly. The results generalize the classes of models in the literature.
“Statistical Evidences Supporting Subsidy Policies for Low-income Students”
This project studies the survival bias in the National Higher Education Entrance Exam (Gaokao) score: Students from low-income families are more prone to early dropout when they perceive poor schooling performance. Low-income students who survive this self-selection mechanism are more likely to have higher overall performance compared with normal students with the same Gaokao score. Starting with observing this phenomenon, this study measures the survival bias according to a drop-out model, then proposes an equilibrium optimal policy that correct this bias by analyzing the game between students and policy makers.
“Dynamic Survival Bias in Learning from Doubly Censored Data” (Job market paper)
This paper studies a survival bias that emerges when an analyst seeks to learn about a random payoff from observing a binary sequence of doubly censored signals. There's firstly a joint right censoring such that a signal can only be observed if its precedents jointly passes a threshold. The second censoring is a truncation in time such that the analyst only sees more recent signals but not the historical ones. The bias is dynamic in the sense that the inference about the un-observed signals evolves as the observation continues. I characterize the analyst’s posterior belief about the random payoff adjusted for the bias by considering all the possible historical signals consistent with the censoring process.
“Contract Design in Self Investment Products” (with Can Sun)
This paper studies a new business model in the self-investment goods market, in which the firm proposes a two-part tariff composed of an up-front fee and a later rebate. This contract exploits consumers’ time-inconsistent preference by over-charging the up-front fee and makes profit by its crowdfunding nature.
Work in Progress
“Population Game with Strategic Substitution”
This paper studies a less-understood class of games --- population games with heterogeneous agents and a payoff structure that exhibits strategic substitution. It establishes the special equilibrium existence and uniqueness conditions and develops the comparative statics results accordingly. The results generalize the classes of models in the literature.
“Statistical Evidences Supporting Subsidy Policies for Low-income Students”
This project studies the survival bias in the National Higher Education Entrance Exam (Gaokao) score: Students from low-income families are more prone to early dropout when they perceive poor schooling performance. Low-income students who survive this self-selection mechanism are more likely to have higher overall performance compared with normal students with the same Gaokao score. Starting with observing this phenomenon, this study measures the survival bias according to a drop-out model, then proposes an equilibrium optimal policy that correct this bias by analyzing the game between students and policy makers.
