Bayesian Estimation of a Simple Macroeconomic Model for a Small Open and Partially Dollarized Economy

I describe a simple new-keynesian macroeconomic model for a small open and partially dollarized economy, which closely resembles the Quarterly Projection Model (QPM) developed at the Central Bank of Peru (Vega et al. (2009)). Then I use Bayesian techniques and quarterly data from Peru to estimate a large group of parameters. The empirical findings provide support for some of the parameters values imposed in the original QPM. In contrast, I find that another group of coefficients – e.g., the weights on the forward-looking components in the aggregate demand and the Phillips curve equations, among several others – should be modified to be more consistent with the data. Furthermore, the results validate the operation of different channels of monetary policy transmission, such as the traditional interest rate channel and the exchange rate channel. I also find evidence that in the most recent part of the sample (2004 onwards), the expectations channel has become more prominent, as implied by the estimated values of the forward-looking parameters in the aggregate demand and the Phillips curve equations.