Deterministic Linear Model

[dbunger]

Hello,

I just learned using Dynare a few weeks ago and I have this very simple question for which I haven't found an answer.

I have a stochastic linear model with a normal stochastic shock and need to change it to a deterministic model, where the economy is shocked for the first 4 periods and then an endogenous variable is held fixed for the following 8 periods.
The model starts from the steady state, which is 0 for all variables since it’s a linear model.

So, am I right that there are no values to put in the initval-block?

Until now I have:

var …;
varexo e_B;
parameters …..;

model(linear);
…..
end;

steady;

shocks;
var e_B;
periods 1:4;
values 1;
end;

simul(periods=500);

How should I advance from here?
Thanks very much!

[jpfeifer]

Yes, that looks correct. When not providing initval and endval explicitly, Dynare takes them as 0. For your model, this is correct. The anticipated shock is also correct. However, fixing an endogenous variable exogenously to some value is generally not possible without introducing an exogenous shocks that causes the constancy. What did you have in mind?

[dbunger]

Thanks a lot for your reply!

I’m trying to replicate the model developed by Chen et al. (2012). The large-scale asset purchase shock on long-term bonds lasts for 4 periods and then they hold the long term bonds balance sheet constant for the following 8 periods. Short term nominal interest rates remain at the zero lower bound for the first 4 periods.

I first created a stochastic version with a normal shock and it works perfectly. However, in the deterministic version there seems no response to the shock and everything remains at steady state. Is there somehow a mistake?

I figured out that the only way to hold long term bonds constant for these 8 periods is with some smaller shocks. What do you think?