we know that OLS estimator $\beta = (X'X)^{-1}X'y$
Under what conditions of the matrix X, the OLS estimator is well-defined?
Asked
Active
Viewed 15 times
1
-
Have you tried looking this up in a textbook? Dozens if not hundreds of textbooks contain the answer. – Richard Hardy Nov 29 '21 at 18:03
-
1the answer is that X is full-column rank matrix. Right? @RichardHardy – 1190 Nov 29 '21 at 18:06
-
I do not remember. I would look this up in a textbook. There is probably some condition on $\mathbb{E}(X'X)$, too, if we are talking about time series (you have used the tag, so I assume we are), but again I would need to check. – Richard Hardy Nov 29 '21 at 18:50