I have faced a problem. I took a mod from this forum, more specifically Matrix close to singular or badly scaled, Boja_SOENK.mod which is quite workable. Then I checked the mod against the original paper and found two small mistakes in the mod. What I found were following.
In the original Boja mod(line 104), mc_h = (sig/(1-h))(c-hc(-1))+vphil-z_h**+*gam2(1-gam1)v+gam1mu;
I changed it to mc_h = (sig/(1-h))(c-hc(-1))+vphil-z_h**-*gam2(1-gam1)v+gam1mu;, (only sign before gam2) following appendix 5, equation 5.8, original paper.
In the original Boja mod(line 142), q = psi**+(1-gam2*(1-gam1))v-gam1mu;
I changed it to q = psi-**(1-gam2*(1-gam1))v-gam1mu; (only sign after psi), following appendix 5, equation 5.19, original paper.
After these correction, Dynare gives an error, saying that
There are 4 eigenvalue(s) larger than 1 in modulus
for 5 forward-looking variable(s)
The rank conditions ISN’T verified!
??? Error using ==> print_info at 42
Blanchard Kahn conditions are not satisfied: indeterminacy
Error in ==> stoch_simul at 71
Error in ==> Boja_SOENK at 346
info = stoch_simul(var_list_);
Error in ==> dynare at 120
It seems that the first correction doesn’t have any effect. But the second one changes everything when I make the corrections one by one.
Help help me out here. Thank you in advice.