Expectations and optimal monetary policy: A stability problem revisited