RBC news shock puzzles

Stochastic simulations
1

Dear dynare team, I read example about RBC news shock. Now I have some problems:

  1. Here you list two TFP shocks, i.e., eps_z_news and eps_z_surprise. And I think the shock eps_z_news would anticipate by agents (such as households) when it happened, but the shock eps_z_surprise cannot anticipate by agents when it happened. Is right?
  2. In the model block, you entered the eps_z_news(-8), what means? Could you give me more details about this ?
  3. Why you set two shocks, as you said ‘having two exactly offsetting shocks from continuous distributions to keep the exogenous variable constant is a 0 probability event for the agents of the mode’. I cannot understand very well about this, is there any more introduction?’
  4. Finally, I want to know if I only set shock eps_z_news, is OK? And at steady state, we use the simult_ to simulate the process after given shock, this doing means the shock can anticipate by agents, what is different from the stoch_simul ?
    Please give me some help, thanks a lot.

My best,
ZL

2
  1. Yes, that is the idea.
  2. It means that the shock eps_news enters the information set at time t (as all shocks do by convention), but it only affects TFP 8 periods later, i.e. 8 periods after it has become known that this shock will happen.
  3. Any event from a continuous distribution has likelihood 0. It’s the density times 0. Thus, having one shock that exactly offsets the other one is extremely unlikely (loosely speaking).
  4. The part with simult_ and the pure news shock is for didactical purposes. It is along the lines of the Beaudry/Portier (2004)-paper. But nothing prevents you from having only news shocks or doing different experiments.
1 Like