Jump variable and predetermined variable

Stochastic simulations
1

i want to reproduce the result of below article (cox and harvie, 2010):
(Resource price turbulence and macroeconomic adjustment for a resource exporter:A conceptual framework for policy analysis)

i write the final equation (20 equation) in dynare and i want to do simulation analysis, but harvie said in his article that:

"In this case, there are eight differential variables in the model: kp, kg, m, b, w, f, q and e; twelve algebraic variables: r, Nod, Nos, T, wp, y, yp, R, one, c, l and B; and ten exogenous parameters that are used to derive a solution for the long run steady state: mbar, rstar, kgstar, oa, pres, op, Nosp, c̄g, pstar and ystar

he said: **Of the eight differential variables, the first six are predetermined non-jump variables that adjust only gradually. The last two differential variables, q and e, are assumed to be non-predetermined or jump variables. **
For dynamic stability, the system must generate six negative and two positive eigenvalues.

i guess it means that six variables(kp, kg, m, b, w, f) is predetermined an others(q, e) is jump variable.
my question is how can i tell Dynare which of my variables are predetermined (backward looking) and non predetermined (forward looking)?

i write predetermined variable in this format, for example:
kp(+1)-kp = eta*q;

an i write jump variable in this format, for example:
q-q(-1) = (1/delta3)*q +(delta2/delta3)*r -(delta1/delta3)R -(delta2/delta3)(m(+1)-m);

as you see in second equation we have one jump variable(q) and one predetermined variable(m)
my dynare is correct or not?

I’m looking for your kindly response.

2

I am not sure that you simply use Dynare for the equations in that paper. Dynare solves difference equations, while the paper is concerned with differential equations. That is why you need to discretize. Unfortunately, I don’t have experience with this.

However,

kp(+1)-kp = eta*q;

seems wrong to me. Predetermined variables have a timing kp(-1). But you could get around this by using the predetermined_variables command to specify the predetermined variables. That would probably be the easiest way.

3

Dear jpfeifer,
can I say “Predetermined” is just a thing of “timing” in dynare?
That is all variables look like x(-1) are predetermined variables. If I have last period consumption c(-1) in some equations, can I say “c(-1)” is a predetermined variable?

4

Predeterminedness is a general concept referring to whether a variable can be changed at time t. c(-1) as decided yesterday and of course is predetermined. That is the reason why habits in consumption introduce a new state variable in models.

1 Like
5

Thank you!

6

Hi jpfeifer,

I’m new to Dynare, and I have a few questions in a similar vein.

  1. I’ve been following Eric Sims’ guide to using Dynare (https://www3.nd.edu/~esims1/using_dynare_sp15.pdf)[Using Dynare for DSGE models], and he mentions that predetermined variables (like capital stock) have to be re-written with a lag: k_t has to be replaced by k_t-1, k_t+1 by k_t and so on. Is this always the case?
  2. Does the order for declaring variables matter in Dynare?

Regards,
Saunok

7
  1. No, there are two types of timing you can use. But you need to be consistent.
    image
    image1602Ă—1166 125 KB
  2. No, the order does not matter.
8

Thank you for clearing that up!