Approximating Multisector New Keynesian Models

Carvalho, Carlos; Nechio, Fernanda

doi:10.24148/wp2017-12

Working Paper

Approximating Multisector New Keynesian Models

Abstract: We show that a calibrated three-sector model with a suitably chosen distribution of price stickiness can closely approximate the dynamic properties of New Keynesian models with a much larger number of sectors. The parameters of the approximate three-sector distribution are such that both the approximate and the original distributions share the same (i) average frequency of price changes, (ii) cross-sectional average of durations of price spells, (iii) cross-sectional standard deviation of durations of price spells, (iv) the cross-sectional skewness of durations of price spells, and (v) cross-sectional kurtosis of durations of price spells. These results provide the tools for a growing literature that tries to estimate empirically-relevant multisector models with much reduced computational costs.

JEL Classification: E12; E22; J60;

https://doi.org/10.24148/wp2017-12