DSGE Model with Structure Variance
In this paper we develop full Bayesian inference for a rich class of DSGE models (Dynamics Stochastic General Equilibrium). It is well known that any DSGE model after being log-linearized could be written as a state space model with Gaussian shocks in the state equation, in this context we propose to explore a DSGE with structure variance on the shocks, building a normal model with stochastic volatility and a Student-t model with stochastic volatility. The methodologies are applied to real data and is compared with the traditional approach.