Log linearizez NKM with heterogenous expectations

Stochastic simulations
1

Hello Everyone,

I am pretty new with Dynare and I wondered if someone could help me solve some questions about dynare.

I am trying to solve a Log linearised NKM model with heterogeneous expectations composed of a heuristic switching model. In this framework, we use a switching mechanism based on the forecasting error of each heuristic based on the forecasting accuracy of the previous period.

Can dynare compute this type of model? Or can it only solve rational expectations? When I try to compute the proportion of individuals that believe in a certain way I get wrong values (bigger than 1). Also important to mention that under RE, the mod file works perfectly.

For a methodological reference please see the first part of section 3.2 of Goy, G., Hommes, C., Mavromatis, K. (2020). Forward guidance and the role of central bank credibility under heterogeneous beliefs. Journal of Economic Behavior Organization.

Thank you!!

2

Dynare by default only does rational expectations. But my understanding is that it can also be used to solve these types of models, although it requires some tweaking. Essentially, you are explicitly specifying time t expectations instead of letting Dynare compute them based on RE.