Sensivity analysis

Hi everyone,

I am facing an issue when performing a sensitivity analysis in a Multivariate Filter (MVF). In particular, I want to analyze how potential output reacts to changes in the steady-state trend growth parameter of non-mining GDP, g_ss​.

While running the same exercise for the steady-state unemployment rate (u_ss​) lleads to small but noticeable changes in the estimated potential output, modifying g_ss appears to have no effect on the results. I tested this in two ways: by changing g_ss​ directly in the calibration block and by modifying it within the .mod file (i also tested using with set_param_value) , but in both cases the estimated potential output remained unchanged.

As pointed out by my advisor, the parameter does in fact affect potential output (he show me his results).

I would appreciate any guidance on how to correctly implement this sensitivity analysis or on potential reasons why changes in g_ss​ may not be reflected in the estimated potential output.

Ejercicio_2.mod (3.6 KB)

Here is the data file

FMV.xlsx (13.6 KB)

g_ss only enters in

g_ytend = theta*g_ss + (1-theta)*g_ytend(-1) + w_gytend;                    

while

g_ytend 

does not affect any other variable.