Hello,
I am running a Fama-MacBeth (1973) regression and trying to show the change in the resulting coefficients over an event period. For clarity, I am trying to see if the coefficients on particular factors increased or decreased after a certain date. I have run two separate models (before and after) and can see the results using this method.
But, I am also trying to run the model with interaction variables (time period dummy * Factor) to show the impact. I am running into a multicollinearity issue since the regressions are run cross-sectionally one period at a time. I have yet to see a study show results with interaction variables in this manner - wondering what I am missing.
Has anyone had success running Fama-MacBeth with time period interaction variables as opposed to running two separate models (i.e. before and after models)? Thoughts are much appreciated.
Thank you,
I am running a Fama-MacBeth (1973) regression and trying to show the change in the resulting coefficients over an event period. For clarity, I am trying to see if the coefficients on particular factors increased or decreased after a certain date. I have run two separate models (before and after) and can see the results using this method.
But, I am also trying to run the model with interaction variables (time period dummy * Factor) to show the impact. I am running into a multicollinearity issue since the regressions are run cross-sectionally one period at a time. I have yet to see a study show results with interaction variables in this manner - wondering what I am missing.
Has anyone had success running Fama-MacBeth with time period interaction variables as opposed to running two separate models (i.e. before and after models)? Thoughts are much appreciated.
Thank you,
Comment