Error in mode finding

Simply because the news shocks provide agents with time to adjust their behavior while the actual shock has not realized. See my last post at [Deterministic Shock in A Stochastic Model- IRF)