Find a interesting feature of function simult_ (or bug?)

Stochastic simulations
1

Dear professors,

I have found a new feature (or bug?) of the simlut_ function. When setting the stateshock_vec as indicated by the method in RBC_state_dependent_GIRF.mod, the shock can be correctly expressed as a percentage of its previous value. However, when setting the value of stateshock_vec starting from period 1, the shock is expressed as an absolute value.

stateshock_vec=zeros(irf_periods+burn_periods,M_.exo_nbr); %initialize impulse vector 
    stateshock_vec(:,1) = stateshock_vec(:,1) + 0.005;

    starting_point=oo_.dr.ys; %define starting point of simulations
    shocks_baseline = stateshock_vec;
    y_baseline = simult_(M_,options_,starting_point,oo_.dr,shocks_baseline,options_.order);
    adatass(:,i) = y_baseline(:,end);
2

Sorry, but I don’t understand what the problem is you are experiencing. Can you explain in more detail? For example, what do you mean with

? You always specify the absolute shock size. This absolute shock size can only be interpreted as a percentage if the variable you are adding the shock to is in percentages, e.g. being logged.

3

Sorry, Professor jpfeifer. I forgot the fact that I set the variable in log form. :rofl: