Inflation Targeting In A Small Open Economy: The Colombian Case

This paper presents a dynamic stochastic general equilibrium model of inflation targeting in small open economy. We calibrate the model to the Colombian economy and present the response of some macroeconomic variables to different types of shocks that are relevant for emerging economies. We also analyze the sensitivity of those responses to some key parameters. Furthermore, using simulated data from the model we study the ability of the model to capture the spectra, the phase and the coherence of observed output and inflation. We follow a frequency domain comparison methodology proposed by Diebold, Ohanian and Berkowitz (1998,[19]). The Colombian data is characterized by: first, cyclical inflation and output gap (as measured by Hodrick – Prescott filter) are dominated by periodic movements between 2 and 25 quarters with a peak between 10 and 12 quarters. The cross spectrum and coherence show results in the the same direction. Second, the coherence does not show any significant dominance of frequencies for the cross movements but the correlation jumps to 0,6 for periodic movements around 5 quarters. These facts are compared to the data simulated from the model. We conclude that the simulated data spectra and cross spectra do not differ statistically from the respective population quantities for, at least, frequencies beyond 0,05 . Which correspond to periodic movements of up to at least 10 quarters. The model spectra presents more persistence than the observed data the population coherence is captured for most frequencies but the ones around the peack of the model's theoretical coherence and very long run periodic movements. Subsequent research will address these issues.