I am a new Dynare user and have the following problem:
I am using an estimated DSGE model with financial frictions to run some counterfactual scenarios. In practice I use the series of structural shocks and use the 2007-2009 period shocks to forecast model variables today, in order to re-play the crisis.
My problem is that the matlab code which re-calls the saved parameters from the mod file, works well with Dynare 4.4.3 and earlier versions, BUT gives erratic forecasts with Dynare versions later than 4.4.3. Were there some changes that can affect this? Maybe in the storing matrices M_ and oo_dr?
Thanks a lot and looking forward for a reply!