Hello everyone,
I am trying to estimate the following model with SolveDSGE:
states:
a, k
end
jumps:
c, muc, n
end
shocks:
ε
end
parameters:
α = 0.4
β = 0.99
δ = 0.1
γ = 0.1
ρ = 0.9
θ = 5.0
sd = 0.01
end
equations:
c^(-1) = β * muc(+1) * (α * a(+1) * k(+1)^(α - 1.0) * n(+1)^(1.0 - α) + 1.0 - δ)
n = (θ^(-1) * c^(-1) * (1.0 - α) * a * k^α)^(1.0 / (γ + α))
k(+1) = a * k^α * n^(1.0 - α) - c + (1.0 - δ) * k
log(a(+1)) = ρ * log(a) + ε * sd
muc = c^(-1)
end
I wish to use a projection method with Chebyshev polynomials. But unfortunately, I am not succeeding.
The problems start already in computing the steady-state.
Do you have any suggestions about the vector of initial conditions?
I just tried the [0, 0, 0, 0, 0] but it does not work.
By trial and error, I discovered that [0.5, 1.5, 0.5, 1.5, 0.5] works, but this is not a very clever approach…
Anyway, I computed the grid with:
P = ChebyshevSchemeStoch(ss,chebyshev_nodes,[21,21],9,4,[26.0 26.0; 0.1 0.1],tol,tol,maxiters)
But the model is not solvable, giving a DomainError with -0.7941962588004425.
Does anyone know what am I doing wrong?
Thank you!