deterministic simulation

[rockyyao]

Dear all,

I am simulating a deterministic version of NK model with taylor rules. I want to simulate a preannounced contractionary policy by restricting nominal interest rate (i) from changing in the first 4 periods. Because "i" in my model is an endogenous variable, I can not use the commend "schocks" to specify the path. "histval" also doesn't work in my case. Can anyone help me to figure out how to do it?

Thanks

Rocky

[jpfeifer]

From what I understand, restricting i might be wrong. If the shock is preannounced, people will already react today, leading to endogenous changes in inflation and output and thus in the interest rate via the Taylor rule. What you could do is specify the monetary policy shock as a news shock, i.e.

Code: Select all

i=i_ss+rho*(pi-pi_ss) + eps_interest(-4);

where eps_interest is an exogenous shock. In this case, the interest rate at time t is shocked by the value of the shock announced 4 periods ago. If you now use stoch_simul to generate impulse responses, the interest rate will be exogenously shocked after 4 periods, but will also react endogenously starting at time t.

[rockyyao]

Thank you, jpfeifer. It is a good advise. I agree that fixing i is not a good idea. What you proposed is nice in the stochastic simulation, but its IRF only gives me the dynamics around one steady state. What I really want to do in this deterministic simulation is similar to what Mankiw Reis did in their 2002 paper. To simulate the dynamics between two steady states, one is high inflation and the other is low inflation. I want to see the transition of a NK model with a simple taylor rule from high inflation to low inflation (attaching my mod file here). Can you give some further ideas to do that?

Thanks

Rocky