Haoyuan Zeng 曾皓原

This paper studies sellers' and buyers' incentives to engage in shill bidding in single-object auctions. An auction is seller identity-compatible if the seller cannot profit by posing as one or more bidders through fictitious identities, and buyer identity-compatible if no buyer can profit by posing as multiple bidders. We distinguish lit auctions, which reveal the number of bidders, from dark auctions, which conceal it. Our results highlight the importance of concealing the number of bidders in deterring shill bidding: no optimal lit auction can achieve our design goal, whereas the dark first-price auction uniquely does.

This paper introduces a novel criterion, persuasiveness, to select equilibria in signaling games. In response to the Stiglitz critique, persuasiveness focuses on the comparison across equilibria. An equilibrium is more persuasive than an alternative if the set of types of the sender who prefer the alternative would sequentially deviate to the former once other types have done so—that is, if an unraveling occurs. Persuasiveness has strong selective power: it uniquely selects an equilibrium outcome in signaling games à la Spence (1973). More generally, persuasiveness refines predictions beyond existing selection criteria. Notably, it can also select equilibria in cheap-talk games, where standard equilibrium refinements for signaling games have no selective power.

We study the design of mechanisms—e.g., auctions—when the designer does not control information flows between mechanism participants. A mechanism equilibrium is leakage-proof if no player conditions their actions on leaked information; a property distinct from ex-post incentive compatibility. Only leakage-proof mechanisms can implement social choice functions in environments with leakage. Efficient auctions need to be leakage-proof, while revenue-maximizing ones not necessarily so. Second-price and ascending auctions are leakage-proof; first-price auctions are not; while whether descending auctions are leakage-proof depends on tie-breaking.

In the theory of multi-stage games, correlated action profiles can be implemented using an autonomous device that sends private messages to the players at the beginning of each stage. In this paper, we assume that such messages, while initially confidential, are publicly observable at the end of the stage. We first show that the revelation principle holds for ex-post observable mediated play. Then, we derive necessary and sufficient conditions for the implementability of a given payoff profile. These conditions coincide with the standard conditions for two-player games and for games that satisfy the NEU condition, but are more flexible otherwise.