Solving recent RBC models using linearization: further reserves

Abstract

Through a simple example, we show that the successive sophistications introduced in the early RBC models in order to improve their internal propagation mechanisms have actually increased their non-linearities, even locally. Accordingly, linearization-based resolution methods become much more disputableı than they were for early RBC models. Simple comparative studies ofı impulse-response functions are used to illustrate this point. We conclude by pointing at sorne alternative resolution techniques that allow the model builder to take non-linearities into account and/or to handle with the presence of large state spaces.