Question on model with specified trend

Stochastic simulations
1

Dear Professor Pfeifer,

In a model with specified trend, e.g. technology shock A_t is composed of a permanent component A_{t}^{p} and a transitory component \varepsilon_t, and A_{t}^{p} has A_{t}^{p}=A_{t-1}^{p}\varphi _{t}. What’s the economic difference between \varepsilon_t and \varphi _{t}? And could I just neglect the transitory component \varepsilon_t and only consider about the permanent component \varphi _{t}?

Thank you for time on this post.

2

Permanent and transitory shocks have different effects. For the transitory one, agents know that it will ultimately vanish, allowing them to smooth the time path. That is not possible for permanent shocks.

3

Dear Professor Pfeifer,

If I do not have enough observed variables, could I neglect the transitory component \varepsilon_t and only consider about the permanent component \varphi_t?

4

What do you mean with

Also, what you consider in your model depends on the purpose of your model.

5

Dear Professor Pfeifer,

I mean that I do not have enough data series to run the estimation, so I want to reduce the number of shocks.

And also, what does this transitory technology shock refer to in reality?

6
  1. You can have more shocks than observables.
  2. Transitory shocks are mean-reverting. They are the standard shocks considered e.g. in the RBC model.