Computational Methods for Oblivious Equilibrium

Oblivious equilibrium is a new solution concept for approximating Markov perfect equilibrium in dynamic models of imperfect competition among heterogeneous firms. In this paper, we present algorithms for computing oblivious equilibrium and for bounding approximation error. We report results from computational case studies that serve to assess both efficiency of the algorithms and accuracy of oblivious equilibrium as an approximation to Markov perfect equilibrium. We also extend the definition of oblivious equilibrium, originally proposed for models with only firm-specific idiosyncratic random shocks, and our algorithms to accommodate models with industry-wide aggregate shocks. Our results suggest that, by using oblivious equilibrium to approximate Markov perfect equilibrium, it is possible to greatly increase the set of dynamic models of imperfect competition that can be analyzed computationally. ∗We have had very helpful conversations with Uli Doraszelski, Liran Einav, Feryal Erhun, Paul Glasserman, Hugo Hopenhayn, Ken Judd, Jon Levin, and Ariel Pakes, as well as seminar participants at Berkeley, Columbia, Duke, Iowa, INFORMS, Kellogg, Minnesota, NYU, SITE, Stanford, Rochester, UCLA, UIUC, Univ. of Chile, UT Austin, and Yale. Przemyslaw Jeziorski provided exemplary research assistance. This research was supported by the Federal Reserve Bank of San Francisco, General Motors, the Lillie Fund, the National Science Foundation, and the Office of Naval Research. Correspondence: [email protected]; [email protected]; [email protected].