Spurious Convergence and errors

Dynare help (legacy posts)
1

Please help me…
I’m running a model with Dynare, but only getting this message

I think my initial value is wrong, but don’t know how to guess it correctly.
Could anyone take a look at my code and give me some advice, please?

BTW, I don’t know the message of ‘Spurious Convergence’. What does this mean?

Starting Dynare (version 4.1.0).
Starting preprocessing of the model file …
Substitution of exo leads: added 1 auxiliary variables and equations.
Found 13 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.

Residual for equation number 1 is equal to 0
Residual for equation number 2 is equal to 3.4213
Residual for equation number 3 is equal to -0.043418
Residual for equation number 4 is equal to -0.78368
Residual for equation number 5 is equal to 0.14247
Residual for equation number 6 is equal to -0.7355
Residual for equation number 7 is equal to 0
Residual for equation number 8 is equal to 0.17878
Residual for equation number 9 is equal to -0.0024868
Residual for equation number 10 is equal to 0
Residual for equation number 11 is equal to -0.2282
Residual for equation number 12 is equal to 0
Residual for equation number 13 is equal to 0

SOLVE: Iteration 20
Spurious convergence.
0.3826
3.9545
0.2019
0.2205
0.0139
-4.9425
0.1350
-0.0821
1.6269
0.0413
0.9908
0
0

??? Error using ==> steady_ at 132
STEADY: convergence problems

Error in ==> steady at 52
steady_;

Error in ==> frics at 190
resid(1);steady;

Error in ==> dynare at 132
evalin(‘base’,fname) ;

frics.mod (1.91 KB)

2

Either you compute the steady state by hand and enter the solution as initial values. Or you tweak the numerical values until it fits by using the residuals output as a guidance if you move closer. Or you try, if a different algorithm helps, i.e. use
steady(solve_algo=integer);
Possible values for integer are (copied from the manual page 24):
• 0: uses MATLAB® Optimization Toolbox FSOLVE
• 1: uses Dynare’s own nonlinear equation solver
• 2: splits the model into recursive blocks and solves each block in turn
• 3: Chris Sims’ solver
• 4: similar to value 2, except that it deals differently with nearly singular Jacobian
• 5: Newton algorithm with a sparse Gaussian elimination (SPE)
The default is 2.