It is a consequence of our model of (flexible) exchange rate determination that, when capital markets are sufficiently integrated, interest rates across countries must be equated. Clearly, however, there are differences across countries. How can we reconcile this fact with our theory?