HI,

Please find enclosed a mod file containing a very simple model with human capital. The model looks perfect when solved by hand. However, dynare is not able to find steady state. Can anyone help.

Thankssfr_2017_final.mod (1.6 KB)

HI,

Please find enclosed a mod file containing a very simple model with human capital. The model looks perfect when solved by hand. However, dynare is not able to find steady state. Can anyone help.

Thankssfr_2017_final.mod (1.6 KB)

Increasing `maxit=1000`

I get

```
STEADY-STATE RESULTS:
c -6.91256e-11
y -1.6123e-09
u 0.319332
e -0.172096
epsl 101
h -3.56907e-09
k -1.56213e-08
w 1.44308
r_k 0.035101
i_k -9.72279e-10
z 1
m 1
```

This indicates there is still something wrong with your model equations.

Thanks I corrected. But I am confused with how dynare is working.

When I block equation 11 (law of motion of human capital) and allow the hhd budget constraint instead, the system solves. I get the impulse response.

When I introduce equation 11 but block budget constraint, I get the following error.

““The Jacobian matrix evaluated at the steady state contains elements that are not real or are infinite””

When I block equation 10 and introduce equation 11, I get the following error

““Blanchard Kahn conditions are not satisfied: no stable equilibrium””

HOw is it possible for dynare to produce different errors by tinkering with some equations when the system solves manually so well. What is dynare doing? Attached is the corrected mod file.sfr_2017_final.mod (1.8 KB)

Do not use

`predetermined_variables k ;`

if you already use the stock at the end of period timing.

In that case the model does not solve

Error using print_info (line 51)

The Jacobian matrix evaluated at the steady state contains elements that are not real or are infinite

I get as the steady state in this case

```
STEADY-STATE RESULTS:
c 0.821942
y 1
u 0.333333
e -54.6774
epsl 50.5
h 1.55925
k 7.1223
w 0.481002
r_k 0.035101
i_k 0.178058
z 1
m 1
```

I am using Dynare 4.5.4.

sfr_2017_final.mod (1.8 KB)

The impulse responses are not converging to zero. Does that mean the system is exploding? I want to get the peak multiplier of a shock to human capital actually which seems difficult here.

What happens if you set

`irf=1000`

?

It converges. But how do we interpret that?

Interpret what?

If the impulse responses dampens after 1000 quarters, this seems like a permanent shock. Is that true?