about the effectiveness of sterilised intervention and the channels through which current account imbalances influence exchange rates. Evidence of a substantial risk premium, however, is not required to verify that current account imbalances can have a substantial influence on exchange rates, since wealth effects are not the only channel through which current account imbalances can exert their influence A portfolio balance model of Germany is set out and estimated by Obstfeld (1983) and the effects of sterilised and non-sterilised foreign reserve sales are studied through dynamic perfect-foresight simulations of the empirical model. A sample of monthly data from January 1975 to October 1981 is used estimate the model The study makes inquiry into whether imperfect substitutability between foreign