Simulating AR(1) with trend


I tried to simulte an AR(1) equation in dynare with a growth factor. I plotted several stochastic simulations, but I don’t see the growth although it should clearly increase as time elapses. Can you interpret the periods in the stochastic simulations as time in dynare?

See the model:
Simplified2.mod (4.9 KB)

Your AR-process does not have a trend. It is still stationary. However, if it really did have a trend, no steady state would exist and you could not use stoch_simul. In such cases, you must work with a detrended, stationary model and then add back the trend.

Thank you. Do you maybe have an example? does something like that.

Thank you. My objective is to find the influence on investment, consumption etc. of increasing life expectancy. Therefore I think that an extended path might be a good way to provide these insights, instead of ivfs.

In the following mod file I added aging (so increasing life expectancy). Is there a way that I can include the fact that agents see the trend in aging, but not the shocks? I have now inserted that agents include the current mortality in their decision to spend etc. (in the euler equations).

Moreover, I have a question about extended_path. Although I loop over set_dynare_seed(i), I do not get a different random shocks path in each iteration.

Simplified2.mod (5.3 KB)