Negatively correlated shocks

[Mathilde]

Hello,

I am very new to Dynare, and I am currently trying to code a model in which I consider firms that have a stochastic access to the debt market. More precisely, I name theta the probability for a firm that was allowed to borrow on the financial market to remain so (1-theta is thus the probability that the firm cannot access the debt market the next period, and cannot borrow).
I computed my equilibrium manually so I have an analytical solution for my steady state.

I would like to obtain the IRF for a few variables considering two shocks: one shock on the interest rate r and one shock on theta (I'm in a stochastic framework). Basically, I would like these shocks to be negatively correlated, so that a positive increase of the interest rate occurs always with a decrease of theta.

I modeled my shocks the following way: knowing that the steady state values are theta = 0.95 and r = 0.01
I assume that tau is the cross persistence of the shocks.
theta = 0.95*exp(u);
u = 0.95*u(-1) + tau*v(-1) + eps;
r = 0.01*exp(v);
v = 0.95*v(-1) + tau*u(-1) + nu;

shocks;
var eps; stderr 0.01;
var nu; stderr 0.01;
var eps , nu = phi*0.01*0.01;

Basically, I don't know which value I should give to tau and phi, or which one should be negative so that they are negatively correlated. I'm not even sure that's how I should do it to be honest.

Thank you very much for any help.

Mathilde

[jpfeifer]

You are basically trying to specify a VAR process in your exogenous variables. See http://www.dynare.org/phpBB3/viewtopic.php?f=1&t=5415 for more on this. There is no general answer to your question. You have to decide how to specify this VAR-process.