
42 Current Trends in Bayesian Methodology with Applications
period and p (= 8)
4
being the order of the VAR. MSE
0
i
is the MSE of variable
i with the prior restriction imposed exactly (w = 0), while, the baseline Fit is
defined as the average relative MSE from an OLS-estimated VAR containing
the four variables, i.e.
F it =
1
4
4
X
i=1
MSE
∞
i
MSE
0
i
(2.5)
In addition, we also look at the forecast performances of the large-scale BVAR
model for a F it = 0.25(w = 0.14), 0.50(w = 0.26) and 0.75(w = 0.58). Note
that the classical VAR attains a highest possible fit of 0.65 for w = ∞, where
a fit of 0.75 is unattainable. However, the fits of 0.25 and 0.50 are obtained
under