Non-Existent Beta Matrix

#1
Help please:

Consider only a simple linear regression model for this please.

so, Y(vector) = X matrix * Beta matrix + Error(vector)

under what conditions is the estimate of Beta nonexistent?

thanks,
 

vinux

Dark Knight
#2
The model is
\( y = X\beta + \varepsilon, \)
OLS estimate of beta is
\( \hat\beta = (X'X)^{-1} X'y \)
If X'X is singular matrix then beta estimates will not be generated using the above equation.
This may happen when you have less observations (or obs < parameters.)