Study the Transition Behavior from Arbitrary Point

[lm280299]

I want to see the transition behavior from, say, 1/3 of the steady state in the stochastic environment. I just want to simulate the economy in a short period, like 100. Can I still specify the initial condition and then use the command stoch_simul (periods=100, drop=0)? I'm not sure since according to the reference manual, if I specify the simulation periods, the simulation will start from the steady state which I don't want.

[jpfeifer]

What exactly are you tying to achieve? If you want to study deterministic transitions from an initial point to the steady state, there is no point using a stochastic setup. Note also that using a stochastic setup only makes sense if you go at least to second order. At first order, you will have certainty equivalence.

[lm280299]

I want to study the transition behavior from an initial point to the steady state with different shock paths and then take the average. Is that the same as the deterministic transition you mentioned?

[jpfeifer]

This is not equivalent, unless your model is linear. If your model is linear, the linearity will imply that the different shock paths average out. That is why this approach with shock paths only makes sense at higher order. With determinstic simulations, you can capture the full nonlinearity, but not different shock paths.

[lm280299]

OK, I understand your explanation. But then can I achieve my objective with Dynare?

Thank you!

[jpfeifer]

You can do that up to third order. See the manual on histval:

In the context of stochastic simulations, histval allows setting the starting point of those simulations
in the state space (it does not affect the starting point for impulse response functions).
As for the case of perfect foresight simulations, all not explicitly specified variables are set to
0. Moreover, as only states enter the recursive policy functions, all values specified for control
variables will be ignored