simulated moments for bkk example

[tonyxia324]

Hi all,

I'm running the bkk example and there's something that I find really weird.
The countries in the mod file seem to be symmetric. When I calculate the empirical moments using periods ranging from 10000-100000, the moments for the home variables and foreign variables are always a little different (usually the difference is around 10^-4). I had suspected that the correlation between the two shocks might be the cause, but making the correlation 0 still give different moments between the home and foreign variables. I wonder if this is just because the empirical moments are not precise enough due to limited number of periods in the simulation? I really appreciate your help!

[jpfeifer]

You can trace this out by setting the correlation to 0 and looking at theoretical moments (periods=0). It should be symmetric and therefore only due to random chatter. If not, please report back