I built a simple NK-DSGE model and have introduced mortgage constraints following the approach of Kiyotaki & Moore (1997). In my model, the loan for an entrepreneur in period t is restricted by the total assets held in period t+1, denoted as K. The specific model and corresponding steady-state solution can be found in the attachment.
However, the impulse response functions of my model under monetary policy shocks exhibit an odd feature that doesn’t appear in the same steps of articles studying the same content: There is a significant and abrupt change between the first and second periods in terms of aggregate output, household consumption, and labor supplied by households, as illustrated below.
As this feature looked so wrong, I had to check my operations and thinking. Given your experience, could you give me some suggestions on whether there might be issues with the structural setup of my model or if there are errors in my parameter selection? Thanks!
A11_get_Calculate_diff.m (172 Bytes)
A11_get_variables1.m (733 Bytes)
model_steadystate.m (1.0 KB)
untitled.fig (97.6 KB)
The main file did not correctly upload.
model.mod (2.6 KB)
I’m very sorry. model.mod couldn’t upload all the time. I changed the file suffix to .m and uploaded it successfully. You can change its suffix to .mod and then run it.
Dear Mr. Pfeifei, my mod file has been uploaded. I sincerely look forward to your reply. Thank you : )
This IRFs seem to follow the movement of the price of capital Q_t. It could be a timing error, but it could also be due to the movement of the real interest rate. Have you tried different Taylor rule parameterizations?
Thank you very much for your reply. I have tried to take various values of several parameters of the taylor rule in the model, but it will not change the simulation results.
Then it’s either a timing issue or a feature of the model.
Thank you very much for your reply. I’ve been trying to find timing issue and analyze the feature of the model these days.But I really don’t see what’s wrong. What is even more sad is that the model I built is basically the same as that described in an article. But the pulse image provided by the author did not encounter the problem I encountered.
Then you should try to ask the authors for their code.