On the Uniqueness of Fully Informative Rational Expectations Equilibria