We tested the significance of abnormal returns surrounding the SEO issue opening during the window period is tested using an estimation period of 250 days: starting from day -31 up to day -280 from SEO issue opening date (0th day). We used -280 days to -31 days to estimate parameters of our expected return model.
Is it essential to use regressions with heteroscedasticity-adjusted t-stats to estimate parameters for expected return model?
It is recommend that the data analyst should: a) correct for heteroscedasticity using HCCM whenever there is reason to suspect heteroscedasticity; b) the decision to correct for heteroscedasticity should not be based on a screening test for heteroscedasticity; and c) if the sample is less than 250, a small sample version of the HCCM known as HC3 should be used.
Dear Ajit Kumar Sir.
Thank you for the answer. Can you please suggest me any good book or reference materials to learn this test.
Thanking you sir.
Yours
Iqbal
I agree with the answer given by Ajit kumar Roy .
Thanks to Ajit kumar Roy .
Best wishes to Iqbal Thonse Hawaldar .
Thanks to Iqbal Thonse Hawaldar for recommending the answer given by me.
Best wishes to him.
Thanks to Iqbal Thonse Hawaldar . I wish him success in life.
http://www.indiana.edu/~jslsoc/files_research/testing_tests/hccm/98TAS.pdf
With a title:
Correcting for Heteroscedasticity with
Heteroscedasticity Consistent Standard Errors in
the Linear Regression Model: Small Sample
Considerations.
Thanks to Minka Anastasova-Chopeva for supplying the link which will be helpful to Iqbal Thonse Hawaldar.
following GM assumptions .. one of the assumptions is that the error terms needs to be homoscedastic .. however with heteroscedastic behaviour in error term will violate BLUE estimates... so the coefficients will either be underestimated or overestimated because of SE calculation .. so use either Newey -West HAC corrected estimation for reporting results
Thank you Minka Anasrasova for the link. But unfortunately I could not open the link. Kindly send link to my email: [email protected]
Thanking you
Prof. iqbal
Thanks to Iqbal Thonse Hawaldar for recommending the answer given by me. Best wishes to him.
I wish him success in life.
I acknowledge the recommendation by Iqbal Thonse Hawaldar of my question.
Thanks to him.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics