Dynare forums

Dear all,
I wrote a dynare code for a DSGE model including Real Estate sector.
But the processing result is that

Configuring Dynare ...
[mex] Generalized QZ.
[mex] Sylvester equation solution.
[mex] Kronecker products.
[mex] Sparse kronecker products.
[mex] Local state space iteration (second order).
[mex] Bytecode evaluation.
[mex] k-order perturbation solver.
[mex] k-order solution simulation.
[mex] Quasi Monte-Carlo sequence (Sobol).
[mex] Markov Switching SBVAR.

Starting Dynare (version 4.4.2).
Starting preprocessing of the model file ...
Found 47 equation(s).
Evaluating expressions...done
Computing static model derivatives:
- order 1
Computing dynamic model derivatives:
- order 1
- order 2
Processing outputs ...done
Preprocessing completed.
Starting MATLAB/Octave computing.


STEADY-STATE RESULTS:

ih 0
ihR 0
ihF 0
h 0
hF 0
hR 0
hRO 0
hRR 0
n 0
nO 0
nR 0
miu_O 0
miu_F 0
miu_H 0
miu_R 0
c 0
cO 0
cR 0
y 0
kF 0
iF 0
kH 0
iH 0
b 0
s 0
rFK 0
qL 0
rH 0
rHFK 0
rRR 0
qH 0
w 0
R 0
ggamma 0
Xi 0
v 0
fiH 0
lambdaZ 0
lambdaA 0
epsilonS 0
lambdaI 0
DLogY 0.0281454
DLogIHR 0.0281454
DLogIHF 0.0281454
DLogIF 0.0281454
DLogQH -0.0174283
DLogS 0.0281454
??? Error using ==> print_info at 51
The Jacobian matrix evaluated at the steady state contains elements that are not real or are
infinite

Error in ==> check at 76
print_info(info, 0, options);

Error in ==> para_est at 412
oo_.dr.eigval = check(M_,options_,oo_);

Error in ==> dynare at 180
evalin('base',fname) ;

I tried to adjust parameter value but it got the same result.
Can you fix it?

Your model local variables nO_bar and nR_bar are complex numbers.

Thanks Pro. Jpfeifer.
Afterwards, I tried to estimate the value of some parameters, but the result is this:

Warning: GAMINV did not converge for a = 0.0016, b = 1.25, p = 0.001.
> In gaminv at 98
In draw_prior_density at 53
In plot_priors at 55
In dynare_estimation_init at 253
In dynare_estimation_1 at 81
In dynare_estimation at 89
In para_est at 490
In dynare at 180
Loading 69 observations from CH_data_1997Q2014Q2.m

Initial value of the log posterior (or likelihood): -77439200145974846000000000000000000
Warning: Matrix is singular, close to singular or badly scaled.
Results may be inaccurate. RCOND = NaN.
> In mr_hessian at 212
In newrat at 67
In dynare_estimation_1 at 337
In dynare_estimation at 89
In para_est at 490
In dynare at 180
Gradient norm Inf
Minimum Hessian eigenvalue 1
Maximum Hessian eigenvalue 1
...
...
and finally

STEADY: numerical initial values or parameters incompatible with the following equations
Columns 1 through 16

1 3 4 5 8 9 10 14 16 18 19 20 21 22 23 25

Columns 17 through 32

28 29 30 31 32 33 34 35 36 37 38 39 41 42 43 44

Columns 33 through 35

45 46 47

Check whether your model in truly linear
lambda = -6.2723e-007; f = 77439200145974846000000000000000000.0000000
Norm of dx NaN
Try gradient direction
Predicted improvement: NaN
lambda = 1; f = Inf
lambda = 0.33333; f = Inf
lambda = 0.11111; f = Inf
lambda = 0.037037; f = Inf
lambda = 0.012346; f = Inf
lambda = 0.0041152; f = Inf
lambda = 0.0013717; f = Inf
lambda = 0.00045725; f = Inf
lambda = 0.00015242; f = Inf
lambda = 5.0805e-005; f = Inf
lambda = 1.6935e-005; f = Inf
lambda = 5.645e-006; f = Inf
lambda = 1.8817e-006; f = Inf
lambda = 6.2723e-007; f = Inf
lambda = 2.0908e-007; f = Inf
lambda = 6.9692e-008; f = Inf
lambda = 2.3231e-008; f = Inf
lambda = 7.7435e-009; f = Inf
lambda = 2.5812e-009; f = Inf

lambda =

-6.2723e-007

STEADY: numerical initial values or parameters incompatible with the following equations
Columns 1 through 16

1 3 4 5 8 9 10 14 16 18 19 20 21 22 23 25

Columns 17 through 32

28 29 30 31 32 33 34 35 36 37 38 39 41 42 43 44

Columns 33 through 35

45 46 47

Check whether your model in truly linear
lambda = -6.2723e-007; f = 77439200145974846000000000000000000.0000000
Norm of dx NaN
No further improvement is possible!

I checked my model and think it is linear.
Why GAMINV did not converge? And why isn't my model linear?
Thanks

Fully calibrate your model and make sure it runs with stoch_simul. When that works, use the use_calibration option of estimated_params_init to initialize estimation.
Currently, there are still issues with your model.