Dynamic policy games feature a wide range of equilibria. This paper provides a methodology for obtaining robust predictions. We begin by focusing on a model of sovereign debt although our methodology applies to other settings, such as models of monetary policy or capital taxation. The main result of the paper is a characterization of distributions over outcomes that are consistent with a subgame perfect equilibrium conditional on the observed history. We illustrate our main result by computing, conditional on an observed history, bounds on the maximum probability of a crisis, and bounds on means and variances over debt prices. In addition, we propose a general dynamic policy game and show how our main result can be extended to this general environment.