I analyze a model with moral hazard and limited enforcement in a small open economy. I find that when state contingent contracting is allowed adding the moral hazard friction improves the model's predictions along several dimensions. First, it justifies why non-contingent debt is an optimal way...