New Keynesian Models and the Kydland and Prescott Test

This paper evaluates New Keynesian models using RBC methods for a number of key macroeconomic variables. Its main findings are that the NK model provides a good description of the behaviour of real variables but performs very poorly when nominal variables are considered. The latter result is puzzling, given the success of NK models in replicating impulse response function and presents a challenge for current models.