Hello everyone.
I’d like to ask you some clarifications about a replication that I try to set up
The model is a New Keynesian type of Galì Valles and Lopez-Salido, with rule-of-thumb consumers. I trying to use a non-linear model with the NKPC, monetary policy rule and fiscal policy rule expressed in linear terms. Firstly, is this a problem?
Secondly, the Blanchard and Kahn conditions are not satisfied (probably for the reasons mentioned), with an additional unit root detection which is not matched by enough forward looking variable. Using that linear form, is it possible that the inflation is not detected as forward looking? What can be other problems?
Mod and steady state file are attached.
tesigali2.mod (2.2 KB)
tesipar.m (1.0 KB)
- The unit root is expected as you have the price level in the model.
- Indeterminacy in that type of model may be caused by the transfer rule, which is not explicitly specified in the original paper. From a student paper
Thanks for the answer. I solved the indeterminacy problem by deleting the government budget constraint, which Dynare considers as redundant equation. However, the results are far from an acceptable interpretation. From the image you kindly share with me I understand that the solution is undertaken through the reduced form system, which I won’t use since the non-linear model is easier to extend with a further analysis of the model. How can I solve collinearity in this specification? Using two different fiscal rule for the two types of consumers seems to not improve the results
I sent you the replication files for the nonlinear version of the Gali et al. (2007) model.