OLG model: solving for steady state

Hi,

Its my first time using Dynare, so I would appreciate any advice on the following.

I have built an OLG model with private pensions savings with three types of agents (young, old and a capitalist). I have convergence issues for my variables c3, k and f: consumption of the capitalist, capital, and the fund value of the assets held in a pension fund.

But from the equation of motion for capital the steady state of k can be solved by ((alpha * beta)/((1 + n) * (1 + g) - (1 - delta) * beta))^(1/(1 - alpha)). So I’m confused by the residual.

Furthermore if I amend kss from 1.0 to ((alpha * beta)/((1 + n) * (1 + g) - (1 - delta) * beta))^(1/(1 - alpha)) which I thought might solve the problem, I get the error message:

Error using print_info (line 83)
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

Thanks.
db.mod (3.63 KB)

Check your equations for mistakes. In Dynare 4.5, I get

[quote]Residuals of the static equations:

Equation number 1 : 0
Equation number 2 : 0
Equation number 3 : 0
Equation number 4 : 0
Equation number 5 : 0
Equation number 6 : 0
Equation number 7 : 0
Equation number 8 : 0.85067
Equation number 9 : 0.026258
Equation number 10 : -0.4598
Equation number 11 : 0

y 0.372439
r 1.49372
w 0.207655
c1 0.165775
c2 0.155741
c3 0.0748923
k 0.050138
f 0.0170803
tau 0.403358
mu 0.05
junk 0[/quote]

and the IRFs are oscillating, suggesting a timing problem

Thank you Johannes.

I’ve corrected the timing which has improved my output but still leads to some residuals and oscillations (though dampened vs before). Would this still suggest to you that there is some issue with timing? I’m still confused as to why it struggles to calculate the steady state for k.

On another note, I built this code off the OLG example from davidrpugh where T was set to 30. I wanted to change it to 1 to reflect that the exogenous variable mu (more accurately the shock e) - the returns generated by the pension fund - was a 1 period return. This led to the following error:

There are 3 eigenvalue(s) larger than 1 in modulus
for 1 forward-looking variable(s)

The rank condition ISN’T verified!

Would you be able to give any further insights into why this arises?

Thanks.
predetermined.mod (4.27 KB)

Hi Johannes,

I’ve amended my code again. But I still have an issue with the residuals. Any advice would be much appreciated.

Thanks
productivity2.mod (3.59 KB)

What do you mean with

`issue with the residuals`?
If you analytically computed the steady state, there are only two possibilities explaining the inconsistency of these computations with your model: either the steady state computation is wrong or the model equations are wrong (or both).