I am a PhD candidate in Economics at UNSW, Sydney. My research interests lie in Auction Theory, Mechanism Design, and Game Theory.
My supervisors are Carlos Pimienta and Juan Carlos Carbajal.
(With Alex Ellis) (Games and Economic Behavior) Link
We consider first-price auctions with independent and private valuations that have asymmetric valuation distributions and supports. We first show the existence of equilibrium in these auctions through a perturbation approach, thereby establishing that the limit of Bayesian Nash equilibria (BNE) of such perturbed auctions is indeed the Bayesian Nash equilibrium (BNE) of the limit auction with asymmetric supports. We then characterize this BNE and show that the ε-equilibrium (ε-BNE) of the auction with asymmetric supports is a BNE of “close” auctions with common supports. We then demonstrate some numerical examples.
(With Juan Carlos Carbajal) (Working Paper) Link
We investigate the equilibrium equivalence between pay-as-bid auctions and first-price auctions when bidders are budget-constrained. Such equilibrium equivalence hinges on a spread vs. concentrate trade-off. This trade-off can only exist in multi-unit auctions where bidders face hard budget constraints and has no analogue in single-unit auctions. We characterize the spread vs. concentrate trade-off through concavity of the bidder's expected payoff function. When the bidder's expected payoff function is concave, the 'spread' in the spread vs. concentrate dominates, and the symmetric equilibrium strategies in the pay-as-bid auction are a modification of the equilibrium strategies in Che and Gale (1998)'s first-price auction under budget constraints. In a sufficiently large market, the 'concentrate' in spread vs. concentrate dominates; the equilibrium equivalence fails in such markets. In at least one instance, the equilibrium equivalence fails (the 'concentrate' in spread vs. concentrate dominates) with three or more bidders. We also show that while reserve prices can restore equilibrium equivalence, the minimal reserve price required for this purpose approaches the maximal valuation of bidders as the market becomes large.
(In progress)
I study a seller's nonlinear pricing problem when buyers face consumption spillovers. The presence of consumption spillovers ensures that a buyer can receive a non-zero payoff even when the buyer does not participate in the seller's mechanism—the buyer's 'outside option' becomes endogenous. I compare the optimal mechanism—subject to full participation—in two cases: (i) when the seller has commitment power to issue the optimal non-participation threat, and (ii) when the seller lacks commitment power to issue a credible non-participation threat. When the seller lacks commitment power, the seller pays the buyers participation rents in order to induce a full participation equilibrium.