Dependent variable is cumulative returns for a 40-trading-day window after an earnings announcement. Independent variable of interest is a variable related to investor attention around the earnings announcement.

When I risk-adjust the return (using market model or various factor models), the significance of the coefficient of interest is much stronger than when I just use a simple abnormal return (firm return minus market return) for the dependent variable. If anything, I might have expected the opposite.

Any general thoughts on why the risk-adjusted returns gives a stronger result?

  • $\begingroup$ Well you would hope that your predictor applies to the idiosyncratic risk of the stock in question so it seems rather food news that you see it is the case no ? $\endgroup$ – Ezy Dec 29 '18 at 5:58
  • $\begingroup$ Well, I thought there might be a systematic component to it. Even if there wasn't, I didn't expect the systematic component of the returns to seemingly create so much noise that it made such a difference in the results. I wondered if there was some fatal flaw with using risk-adjusted returns that I was overlooking. I'm open to additional critiques / ideas. Thanks $\endgroup$ – sjd_fin Dec 29 '18 at 22:16
  • $\begingroup$ Then regress against the systematic return and check directly what is the significance of the term for that return $\endgroup$ – Ezy Dec 30 '18 at 10:01

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Browse other questions tagged or ask your own question.