Dynare forums

I attempt to use the Guerrieri-Iacoviello Occbin toolkit. When I run my codes, the following errors appear:

There are 9 eigenvalue(s) larger than 1 in modulus
for 9 forward-looking variable(s)

The rank conditions ISN'T verified!

??? Error using ==> print_info at 46
Blanchard Kahn conditions are not satisfied: indeterminacy due to rank failure

Error in ==> stoch_simul at 81
print_info(info, options_.noprint);

Error in ==> Basetaxram_alleqelse at 195
info = stoch_simul(var_list_);

Error in ==> dynare at 120
evalin('base',fname) ;

Error in ==> solve_one_constraint at 54
eval(['dynare ',modnamstar_,' noclearall nolog '])

Error in ==> runBasetaxram_alleq at 44
[zdatalinear zdatapiecewise zdatass oobase_ Mbase_ ] = ...

Personally I don't understand how it could be if I have the exact number of eigenvalues greater than 1 equal to forward-looking variables but still the rank condition fails.
Please help me check the model. Thanks a lot!

Because we are talking about the rank condition and not the Blanchard-Kahn eigenvalue conditions. They are two different things. Typically, linearized DSGE models can be written as:

Ay_t=BE_t(y_t+1)

If a A does not have full rank, you cannot invert it to bring it to the other side. This is your rank failure.

Checking your model is impossible, because you did not tell us what file to run. Running runBasetaxram_alleq.m results in a problem with finding the steady state. In short: I cannot replicate your error message due to other problems with your file.

Dear all,

I got the same with a deterministic simulation:

There are 2 eigenvalue(s) larger than 1 in modulus
for 2 forward-looking variable(s)

The rank conditions ISN'T verified!

However,
MODEL SIMULATION :
Total time of simulation :0.64
Convergency obtained.

I am a bit puzzled what I have to whatch out for when using deterministic simulation. Do I need to check the Blanchard Kahn conditions? How do I know that the (numerical) solution will be unique? Is it problematic that the rank condition is not verified? My guess is, that the solution won't be unique then, correct?

Best,
Peter

Edit: If someone knows were I can read up those things please let me know. I searched for it in the internet and the User Guide but could not find anything useful..