Loading Events

Juuso Toikka (Massachusetts Institute of Technology)

21 March 2012

 

  • Past event

Details

Date:
21 March 2012
Event Category:

“Efficiency in Games with Markovian Private Information”

(joint with Juan Escobar)

abstract

We study repeated Bayesian n-player games in which the players’ privately known types evolve according to an irreducible Markov chain, type transitions are independent across players, and players have private values. Our main result shows that, with communication, any Pareto-efficient payoff vector above a minmax value can be approximated arbitrarily closely in a perfect Bayesian equilibrium as the discount factor goes to one. As an intermediate step we construct a dynamic mechanism (without transfers) that is approximately efficient for patient players given sufficiently long time horizon. The results apply to partnership games, models of relational contracting, collusion among asymmetrically informed firms, and insurance with private income shocks.