Question in Bayesian estimation


#1

Dear all,
I built a DSGE model. The estimates are as follows
‘kappa_r 0.500 0.4083 0.3267 0.5599 gamm 0.1000
kappa_y 0.250 0.7562 0.6826 0.8270 beta 0.1000
kappa_pi 1.500 1.2397 1.1150 1.3563 gamm 0.1000’
kappa_y of estimated value looks a little big. What can we do to make kappa_y 's estimate less ?
For example, make kappa_y close to 0.2


#2

That is impossible to tell without knowing a lot more details of the model. You need to understand economically why the data prefer such a large value. Here, we don’t even know what kappa_y represents. Mechanically, you could simply make the prior tighter, but that is not a good idea without knowing what the problem is.


#3

jpfeifer, I’m sorry I didn’t explain my problem. ‘kappa_r ’ stand for interest coefficient of Monetary Policy; ‘kappa_y’ stand for output gap coefficient of monetary policy, ‘kappa_pi’ stand for Inflation coefficient of Monetary Policy.
When I do Bayesian estimation,the other results seem to be correct, except for inflation. As picture followed.
I find when the value of kappa_y 's estimated is large, it’s going to have this problem. When the value of kappa_y 's estimated is small, there will be no such problem.
Why does inflation grow after monetary policy shocks? is it possible to tighten monetary policy to raise inflation?
eps_r
data2018.xls (31 KB)
price.mod (10.7 KB)


#4

jpfeifer, I follow US_CD08 in MMB completely. I really don’t understand why this problem arises. Is it possible that the model of this article (Christensen and Dib(2008) is problematic?


#5

Please search the forum. This model has been discussed over and over again and AFAIK nobody ever got exactly the same results as the paper.


#6

jpfeifer, thank you. I’m finally know it may be not my problem.