Error in output results - what does it mean?

[pycs]

I am trying to estimate a DSGE model with monopolistic competition and price rigidity according to Rotemberg (1983). The model, in levels, takes Peter Ireland's framework.

I have attached the mod.file

I'm getting the follow output "error". What does it mean? There is no impulse response functions neither.

STEADY-STATE RESULTS:

a 1
c 2.75635
d 1.99836
gam 0.362798
k 21.5948
m 4.07376
mc 0.833334
pi 1.00001
q 0.035101
r 1.01011
y 2.75635
z 1

MODEL SUMMARY

Number of variables: 12
Number of stochastic shocks: 3
Number of state variables: 3
Number of jumpers: 4
Number of static variables: 5


MATRIX OF COVARIANCE OF EXOGENOUS SHOCKS

Variables ea er ez
ea 0.000100 0.000000 0.000000
er 0.000000 0.000400 0.000000
ez 0.000000 0.000000 0.002500

POLICY AND TRANSITION FUNCTIONS
y pi r c r a
Constant 2.756355 0.986693 0.983210 2.756379 0.983210 1.000000
(correction) 0 -0.013315 -0.026899 0.000024 -0.026899 0
a (-1) 0 0.022547 0.045549 -0.000041 0.045549 0.950000
k (-1) 0.042121 -0.008219 -0.008886 0.042136 -0.008886 0
z (-1) 2.618537 -0.466495 -0.462613 2.619385 -0.462613 0
ea 0 0.023733 0.047946 -0.000043 0.047946 1.000000
er 0 -0.500099 -0.000192 0.000909 -0.000192 0
ez 2.756355 -0.491048 -0.486961 2.757247 -0.486961 0
a (-1),a (-1) 0 -0.057366 -0.115377 -0.055944 -0.115377 -0.023750
k (-1),a (-1) 0 0.005894 0.011881 0.040852 0.011881 0
k (-1),k (-1) -0.000653 0.000427 0.000654 -0.008102 0.000654 0
z (-1),a (-1) 0 0.477613 0.965261 2.318392 0.965261 0
z (-1),k (-1) 0.040015 0.012036 0.020639 -0.805452 0.020639 0
z (-1),z (-1) -0.065464 0.103753 -0.144180 -24.058127 -0.144180 0
ea,ea 0 -0.051072 -0.102606 -0.062011 -0.102606 0.500002
er,ea 0 -0.284911 -0.551609 2.617747 -0.551609 0
er,er 0 3.004225 5.816518 -27.579922 5.816518 0
ez,ea 0 0.529211 1.069542 2.568855 1.069542 0
ez,er 0 5.897886 11.418910 -54.160592 11.418910 0
ez,ez 1.378181 -0.143485 -0.416052 -25.206018 -0.416052 0
a (-1),ea 0 -0.097036 -0.194951 -0.117820 -0.194951 0.950003
a (-1),er 0 -0.270666 -0.524029 2.486860 -0.524029 0
a (-1),ez 0 0.502750 1.016065 2.440412 1.016065 0
k (-1),ea 0 0.006204 0.012506 0.043002 0.012506 0
k (-1),er 0 0.098724 0.191141 -0.906541 0.191141 0
k (-1),ez 0.042121 0.012669 0.021725 -0.847844 0.021725 0
z (-1),ea 0 0.502750 1.016065 2.440412 1.016065 0
z (-1),er 0 5.602991 10.847964 -51.452562 10.847964 0
z (-1),ez 2.618545 -0.272622 -0.790499 -47.891434 -0.790499 0


MOMENTS OF SIMULATED VARIABLES

VARIABLE MEAN STD. DEV. VARIANCE SKEWNESS KURTOSIS
y NaN NaN NaN NaN NaN
pi NaN NaN NaN NaN NaN
r NaN NaN NaN NaN NaN
c NaN NaN NaN NaN NaN
r NaN NaN NaN NaN NaN
a NaN NaN NaN NaN NaN


CORRELATION OF SIMULATED VARIABLES

VARIABLE y pi r c r a
y NaN NaN NaN NaN NaN NaN
pi NaN NaN NaN NaN NaN NaN
r NaN NaN NaN NaN NaN NaN
c NaN NaN NaN NaN NaN NaN
r NaN NaN NaN NaN NaN NaN
a NaN NaN NaN NaN NaN NaN


AUTOCORRELATION OF SIMULATED VARIABLES

VARIABLE 1 2 3 4 5
y NaN NaN NaN NaN NaN
pi NaN NaN NaN NaN NaN
r NaN NaN NaN NaN NaN
c NaN NaN NaN NaN NaN
r NaN NaN NaN NaN NaN
a NaN NaN NaN NaN NaN

Can someone have a look and tell me where the problem might be?

Merci/Thanks,
pycs

[MichelJuillard]

Currently, for second order approximation, Dynare runs th simulation by recursion with the 2nd degree polynomials decision/transition functions. This can lead to explosive trajectories and the empirical moments are reported as NaN (Not A Number). The alternative is to look at the 2nd order accurate moments reported when no simulation is done.
An alternative approach would be to run the simulation with the "pruning" algorithm described in Kim, Kim, Schaumberger and Sims, but it isn't implemented yet in Dynare. (The moments would be the same as the 2nd order accurate ones).

Kind regards

Michel