Question: The model is Yi X0 iei E[ei j Xi ] 0 E e2 i j Xi 2i diag{21
The model is Yi Æ X0 i¯Åei E[ei j Xi ] Æ 0 E
£
e2 i
j Xi
¤
Æ ¾2i
§ Æ diag{¾21
, ...,¾2 n}.
The parameter ¯ is estimated by OLS b¯ Æ
¡
X 0X
¢¡1 X 0Y and GLS e¯ Æ
¡
X 0§¡1X
¢¡1 X 0§¡1Y . Let be Æ Y ¡X b¯
and e Æ Y ¡ X e¯ denote the residuals. Let b R2 Æ 1¡be0be/(Y ¤0Y ¤) and e R2 Æ 1¡ee0ee/(Y ¤0Y ¤) denote the equation R2 where Y ¤ Æ Y ¡Y . If the error ei is truly heteroskedastic will b R2 or e R2 be smaller?
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
