Recursive Thick Modeling and the Choice of Monetary Policy in Mexico

By following the spirit in Favero and Milani (2005), we use recursive thick modeling to take into account model uncertainty for the choice of optimal monetary policy. We consider an open economy model and generate multiple models for only the aggregate demand and aggregate supply. Models are constructed by matching the rankings of aggregate demand and aggregate supply and adding other specifications for the rest of the variables. The main results show that recursive thick modeling with equal and different weights approximates the recent historical behavior of nominal interest rates in Mexico better than recursive thin modeling