Reducing overreaction to central banks' disclosures : theory and experiment

Financial markets are known for overreacting to public information. Central banks can reduce this overreaction either by disclosing information to a fraction of market participants only (partial publicity) or by disclosing information to all participants but with ambiguity (partial transparency). We show that, in theory, both communication strategies are strictly equivalent in the sense that overreaction can be indi-erently mitigated by reducing the degree of publicity or by reducing the degree of transparency. We run a laboratory experiment to test whether theoretical predictions hold in a game played by human beings. In line with theory, the experiment does not allow the formulation of a clear preference in favor of either communication strategy. This paper, however, makes a case for partial transparency rather than partial publicity because the latter seems increasingly diffcult to implement in the present information age and is associated with discrimination as well as fairness issues.