Problem with the BGG model

[qazi.haque]

Hi,

I have been looking at the DYNARE code for replicating the BGG (1999) paper provided by Ambrogio Cesa-Bianchi on his webpage (please find the mod file attached).

However, the problem I am encountering is that when I am trying to modify eq 4.19 as q(+1) = psi*(i(+1) - k) to allow for investment delays as in section 4.2.1.1 in the paper, the model turns out to be indeterminate (as reported by DYNARE). But BGG seem to simulate the model under this alternative formulation and show impulse responses in their paper which now got me confused.

Any suggestions?

Cheers,
Qazi

[jpfeifer]

My guess is that 4.23 the law of motion for capital also gets a different timing. Because if you leave the timing for investment, investment will go live immediately instead of with a delay.

[carondon]

I had the same issue....the following works

//6. Eq. 4.19
q=psi*(i_del-k(-1));
% 6.1 W/o investment delays
%i_del=i;
% 6.2 With investment delay
i_del(-1)=i;

[Ramental]

I had the same issue....the following works

//6. Eq. 4.19
q=psi*(i_del-k(-1));
% 6.1 W/o investment delays
%i_del=i;
% 6.2 With investment delay
i_del(-1)=i;

Works perfectly. Thanks.
I've used this model as basis: http://www.nviegi.net/research/bgg.txt