Optimal stopping in a model of speculative attacks

When faced with a speculative attack, banks and governments often hesitate, attempting to withstand the attack but giving up after some time, suggesting they have some ex-ante uncertainty about the attack they will face. I model that uncertainty as arising from incomplete information about speculators’ payoffs and find conditions such that unsuccessful partial defences are possible equilibrium outcomes. There exist priors over the distribution of speculators’ payoffs that can justify any possible partial defence strategy. With Normal uncertainty, partial resistance is more likely when there is more aggregate uncertainty regarding agents’ payoffs and less heterogeneity among them. ∗I am deeply indebted to George-Marios Angeletos, Ricardo Caballero, Bengt Holmström and especially Iván Werning for their invaluable guidance. I also thank Daron Acemoglu, Sergi Basco, Emre Ozdenoren and seminar participants at the MIT Macro Lunch for useful suggestions and comments. All remaining errors are my own. Correspondence: Department of Economics, Stanford University, Stanford, CA 94305. Email: [email protected]