Cassese, Gianluca (contributor); … - 2005
model for time t (de�ned by day/hour of the day) as
y
it
= β
0
+x
0
it
β
1
+�
it
i =1,...,20 or
y
t
= β
0
ι
20
+X
t
β
1 … vector x
0
it
contains the regressors characterizing models 1 - 9. Let�s now stack the T observations
(for instance 3,434 for … GLS estimator
�
β
GLS
1
=(X
0
ΩX)
−1
X
0
ΩY
is consistent and efficient, and also yields consistent estimates of the …