Laubach Williams (2003) State Space Model

[breakceck]

I want to implement the Laubach Williams (2003) model to EViews with some assumptions. The model is attached in a pdf.

My Code:

@ename e1
@ename e2
@ename e3

@evar var(e1) = exp(c(10))
@evar var(e2) = exp(c(11))
@evar var(e3) = exp(c(12))


@signal ygap=c(1)*ygap(-1) + c(2)*(r(-1)-rstar) + e1
@signal pi = c(3)*pi(-1) + c(4)*ygap(-1) + e2

@state rstar=c(5)*g + c(6)*z(-1) + e3
@state z=c(6)*z(-1) + e3



My problem: I get too big values for my rstar (natural rate of interest), because c(5) = -44'783. Can anyone help me or explain that to me?

Many thanks in advance!