I need to estimate 47 unknown parameters with 14 observarble variables
Before estimating, to make sure that I do not have identification problem then I conduct the test for identification problem
In Dynare, I use the following command
identification(ar=10);
The result shows that
All parameters are identified in the model and the figure as the attached fileIdentification.pdf (11.9 KB)
So my question is that the test I conducted by Dynare is enough to conclude there is no identification problem in my model
Your attached document shows that there is no local strict identification issue. That is already a very good and important first step. Of course, you may still have issues with weak identification. You can see this clearly in your graph. Some parameters are identified in the data, but just weakly. As a result, your posterior bands may be wide are be very similar to your prior. But there is not fundamental problem evident here that you need to worry about.