Dynamic Mechanism Design

Paper Session

Friday, Jan. 6, 2017 7:30 PM – 9:30 PM

Hyatt Regency Chicago, Dusable
Hosted By: Econometric Society
  • Chair: Tibor Alejandro Heumann, Yale University

Non-Linear Pricing and Optimal Shipping Policies

Aron Tobias
,
Yale University

Abstract

A monopolist produces a good for sale to a buyer with uncertain valuation. The seller seeks to implement a profit-maximizing non-linear pricing scheme, which includes the time at which the good is shipped to the consumer. If the buyer discounts future payoffs and the seller does not, then delayed shipments act as an almost-perfect substitute for complete information and the monopolist can extract almost all of the surplus from trade. Shipping policies thus serve as a powerful tool of enhancing price discrimination. However, if the seller is as patient as the buyer, then she cannot benefit from delaying allocations.

Dynamic Mechanism Design With Unknown Arrival

Dirk Bergemann
,
Yale University
Philipp Strack
,
Berkeley University

Abstract

We analyze a dynamic revenue-maximizing problem when the arrival time of the agent is uncertain and unobservable to the seller. We consider a large class of dynamic allocation problems in continuous time, including time-separable allocation problems and stopping problems. The valuation of the agent is private information and changes over time.<br />
We derive the optimal dynamic mechanism, characterize its qualitative structure and derive a closed form solution in special cases. As the arrival time of the agent is private information, the optimal dynamic mechanism has to be stationary to guarantee truthtelling. The truthtelling constraint regarding the arrival time can be represented as an optimal stopping problem. The resulting value function of the agent has be convex and continuously differentiable everywhere.

Ascending Auctions in Informationally Rich Environments

Tibor Alejandro Heumann
,
Yale University

Abstract

We study ascending auctions when agents observe multidimensional signals. We provide a novel methodology to solve for equilibria when agents observe multidimensional Gaussian information structures. We provide novel predictions of ascending auctions with multidimensional signals. We show that an ascending auction can have an equilibrium arbitrarily close to the collusive outcome (that is, price is arbitrarily close to 0 in distribution), even when the fundamentals are arbitrarily close to common knowledge. This equilibria is in undominated strategies and does not rely on off-path beliefs. Second, we show that an ascending auction can have multiple symmetric equilibria. Each equilibria induces a different allocative efficiency and different profits for the seller. The multiplicity of equilibria arises from a complementarity in how agents use their signals to determine their drop out price. The methodology developed in the paper extends to a variety of different trading mechanisms (e.g. supply function equilibria, double auctions and generalized VCG mechanisms).
JEL Classifications
  • D0 - General