Hi,
I am estimating a two-country monetary union model on 5 observables: GDP and inflation for both countries and the common interest rate. I have already estimated this model with Bayesian techniques and the absolute fit of the model (i.e. observables vs. 1-step ahead forecast) is acceptable. My problem is associated with the variance decomposition, since the decomposition of output (y) is not being explained by any of the shocks!! How could this be?
VARIANCE DECOMPOSITION (in percent)
e_a e_d e_pih e_ystar e_R e_pstar e_astar
pi 0.09 0.01 0.01 46.66 47.82 0.54 4.87
y 0.00 0.00 0.00 0.00 0.00 0.00 0.00
ystar 0.00 0.00 0.00 38.63 38.63 2.27 20.47
pstar 0.00 0.00 0.00 46.82 46.82 0.64 5.72
R 0.00 0.00 0.00 66.26 29.23 0.45 4.05
Any insights would be highly appreciated?
B.
could you please upload the mod file and the data? i have some experience with two country models.
Pls find attached both the .mod file and the data. The model is a standard small open economy model a la Gali/Monacelli augmented with habit formation and price indexation in a monetary union setting (but no fiscal policy).
The model solves ok but the theoretical moments for the main domestic variables are not being computed (output, consumption, terms of trade and RER/CPI differentials) and as a result, the variance decomposition for output is not computed as well.
I have been thinking hard on this problems. It seems somehow related to the absence of home bias in my model. Specifically, Jondue and Sahuc (2008) state that ‘when there is no home bias in preferences, the perfect risk sharing assumption does not allow us to determine the terms of trade anymore’. If this is indeed the case, then it explains why the above mentioned domestic variables are not being computed, as all these depend on TOT.
Your insight on this problems would be greatly appreciated. In particular, I would be grateful if you can show me how to incorporate home bias and/or share some two country/monetary union code.
Thanks.
data.m (2.87 KB)
soe_MU_index.mod (5.09 KB)