Baxter & King as an endogenous variable

It seems that if I increase the higher frequency (I set p_h=2), I get a smoother responses:

do you have an intuition for this?