Page 1 of 1

can't find steady state

PostPosted: Tue May 22, 2007 1:33 pm
by student6
Could you pleae help me to undertand what's wrong with mu model ( file attached). i'am trying to make a tochatic small open economy model to detect shockcs, impulse response, iRF.

when i try to run code in dynare, i have this mesage:
dynare smallopen1.mod;

rhoz =

0


SOLVE: Iteration 291
Spurious convergence.
0.0093
0.0939
0.0000
-0.0909
0.4614
0.8764
-3.8089
0.0000
0.0122
0.4168
0
-0.0015
0.0000
0.0000
0.0000

??? Error using ==> steady_ at 27
STEADY: convergence problems

Error in ==> steady at 7
steady_;

Error in ==> smallopen1 at 125
steady(0);

Error in ==> dynare at 26
evalin('base',fname) ;

>>
thanks for help

PostPosted: Thu May 24, 2007 9:01 pm
by student6
please someone can help me

PostPosted: Thu May 24, 2007 9:19 pm
by MichelJuillard
sg and strf are declared as both variables and parameters!

rhoz =0,95 comma instead of dot

in initval:
lamda = mu

but mu hasn't a value

why to choose so small guess value for the steady state?

Best

Michel

PostPosted: Thu May 24, 2007 10:38 pm
by student6
thank you for quick replay

- about steady state values:
in my papers i have found the steady state definition of every variable so to find these steady state values i have just remplace parametres by their own values.
- after making modification, i find this result:
SOLVE: Iteration 29
Spurious convergence.
0.0044
0.0677
-0.0042
-0.0567
0.2549
0.8630
-3.2522
-0.0000
0.0108
-0.0047
0.0026
-0.0000
0.4315


STEADY-STATE RESULTS:

b 0.00442133
c 0.0677117
g -0.00424411
i -0.0567016
k 0.254877
lamda 0.862957
mu -3.25221
n -1.85183e-008
r 0.0108048
tb -0.00469968
trf 0.00257212
y -7.05946e-006
z 0.431491
>>
but because my aim is to check chocks, impulse response... i have to made stochastique simulation. so when i add stoch_simul; dynare give me this resultat:
OLVE: Iteration 29
Spurious convergence.
0.0044
0.0677
-0.0042
-0.0567
0.2549
0.8630
-3.2522
-0.0000
0.0108
-0.0047
0.0026
-0.0000
0.4315


STEADY-STATE RESULTS:

b 0.00442133
c 0.0677117
g -0.00424411
i -0.0567016
k 0.254877
lamda 0.862957
mu -3.25221
n -1.85183e-008
r 0.0108048
tb -0.00469968
trf 0.00257212
y -7.05946e-006
z 0.431491

SOLVE: maxit has been reached
??? Error using ==> print_info at 30
Impossible to find the steady state. Either the model doesn't have a unique steady state of the guess values are too far from the solution

Error in ==> stoch_simul at 49
print_info(info);

Error in ==> smallopen1 at 127
info=stoch_simul(var_list_);

Error in ==> dynare at 26
evalin('base',fname) ;

>>
the file attached is dynare code after modification

thanks a lot

PostPosted: Fri May 25, 2007 7:45 am
by MichelJuillard
If I understand correctly, you have solved analyticaly for the steady state of the model. Unfortunately, the nubers that you enter in INITVAL aren't a good solution.
1) with your values c-psi*n^nu is negative and Matlab gives a complex value for the fractional power of a negative number
2) trf should be equal to zero to satisfy the last equation

Best

Michel

PostPosted: Fri May 25, 2007 11:18 am
by student6
Thanks again for you help.
i have a last question: how can i do to get a correct value for initval to find the steady state.

Best

PostPosted: Fri May 25, 2007 11:47 am
by MichelJuillard
I'm surprised by the question because I was under the impression that you could solve for the steady state analytically.

There is no general recepee for good starting value, but it is better to take them in the order of magnitude of 1 unless you have information to the contrary

In your case, I believe that there is a mistake in the model and that the set of equations that you have don't have a solution.

Watch out also for the timing of your stock variables k and b. In Dynare they must be stock at the end of the period and lagged by one in comparison with what you have currently in the file

Best

Michel

PostPosted: Sat May 26, 2007 6:30 pm
by student6
sorry for my stupid question, but really i have find analiticaly the steady state.

I don't know what's wrong with my model, may be there is no steady state solution ,but i will do my best to undertand.

I would like to thanks you.

thanks a lot