The starting point of this article is the observation that many of the Granger-causality tests for policy effectiveness that have been performed are likely to be severely biased because of omitted variables problems. An alternative procedure that does not suffer from this handicap is proposed to test for the stabilizing effect of economic policy interventions. The procedure is applied to test whether the Swiss National Bank's monetary interventions have managed to stabilize the Fr/$ and the Fr/DM exchange rates. The answer is yes.