Dynare forums

Hi, I understand calculating steady state requires art and luck. But is there anything I can do to improve the attached model to enhance my chance of reaching a steady state? Below seems the closest result I can get after numerous tries of initial values and parameters. Any advice would be much appreciated.

SOLVE: Iteration 11745
Spurious convergence.
0.0375
0.0274
1.0152
-0.0519
1.3113
0.6542
-0.5938
-0.6199
1.2337
1.4565
-1.2521
-0.0074
0.1131
0.5540
0.0087
-0.0734
-0.0000
0.0003
-0.0034
0.0308
0.0019
1.0429
1.1198
-3.2572
1.0000
0.1194
-0.0201
0
0
-0.5625

Residuals of the static equations:

Equation number 1 : -6.3667e-05
Equation number 2 : -0.038342
Equation number 3 : -0.023092
Equation number 4 : -0.00020226
Equation number 5 : -0.0093687
Equation number 6 : -0.0014208
Equation number 7 : -0.0082218
Equation number 8 : 0
Equation number 9 : 0
Equation number 10 : -0.00015816
Equation number 11 : -0.0088415
Equation number 12 : -0.0048723
Equation number 13 : -0.0029257
Equation number 14 : 0.093776
Equation number 15 : -0.014007
Equation number 16 : -0.0022533
Equation number 17 : 0.00091766
Equation number 18 : -0.10853
Equation number 19 : -0.002195
Equation number 20 : -0.0083212
Equation number 21 : 0.00043992
Equation number 22 : -0.073357
Equation number 23 : 0.0067228
Equation number 24 : 0.0012881
Equation number 25 : -0.00166
Equation number 26 : 0.020638
Equation number 27 : -0.0060324
Equation number 28 : 0
Equation number 29 : 0
Equation number 30 : 0


Error print_info (line 74)
Impossible to find the steady state. Either the model doesn't have a steady state, there are an infinity of steady states, or the guess
values are too far from the solution

Error steady (line 92)
print_info(info,options_.noprint, options_);

Error modelclean (line 395)
steady;

Error dynare (line 180)
evalin('base',fname) ;

Having Dynare compute the steady state of course only works if there is a unique steady state to be found. But your model almost surely does not possess a unique steady state. You have

Code: Select all

pi=p/p(-1);

which suggests a unit root in the price level and therefore a multiplicity of steady states.

Thanks Johannes. This is the definition equation for inflation rate. What would you suggest to resolve such unit root problem? Do I need to take transformation? Any suggestions would be much appreciated!

What I am saying is that if you have the price level in there, you cannot let Dynare endogenously compute the steady state. If you have a unit root, you need to provide an analytical steady state. Of course, you can drop the definition of the price level from your model.