When doing the cross sectional approach, i.e. first performing the time series to obtain the betas and then regressing against the estimated betas, does the dependent variable HAV to be an excess return? can I still interpret my lambda coefficients as factor risk premia if I model regular instead of excess returns?
@3.04
The GLS estimator of the risk premium should end with the expected returns, not with the estimated beta from TS.
When doing the cross sectional approach, i.e. first performing the time series to obtain the betas and then regressing against the estimated betas, does the dependent variable HAV to be an excess return? can I still interpret my lambda coefficients as factor risk premia if I model regular instead of excess returns?
how do you derive many betas?
Should it be fitting first before estimating !
2:03
not clearly explained
Which time series approach is usually used? VAR?
The voice quality is poor