Hi everyone
I have a problem to my model
I run Dynare then the likelihood increases overtime
I do not know what happens to this
can anybody detect this problem for me ?
My file as qaa.zip (16.7 KB)
What do you mean with
It is completely normal that the likelihood improves during mode-finding. What is not OK is that you are using a normal distribution for autocorrelation parameters that can only be between -1 and 1 (or 0 and 1)
Dear Prof. @jpfeifer
Sorry for confusing you
means that the initial likelihood is 429.304, then increases to over 1000. In Dynare, all of them are minus sign (-429.304 decreases to -1000)
I correct it by using bete distribution
But the likelihood still decreases (minus sign as Dynare reports)
I still don’t get your point. The likelihood/posterior kernel is supposed to increase during mode-finding. What Dynare reports is minus the likelihood/posterior kernel as a minimizer is used. Thus, the reported value is supposed to decrease.
Dear Prof. @jpfeifer
Sorry for confusing
Yes, I mean that my model is that the initial likelihood is -429 then decreases to -1000