I do an event study to examine the effects of catastrophes on financial firm's stock prices. To test for the significance, I have two Z-Statistics: One as adjusted in Patell (1976) and one using the test statistic suggested by Boehmer, Musumeci, and Poulsen (BMP, 1991). They mostly differ from each other with respect to their significance. Do I interpret both test statistics? Some papers I read only interpret the BMP coefficients. Thanks in advance.
My suggestion is that since people use Event Study methodology to usually study equilibrium why don't you use some simple test like runs test with the excess deviations by assuming uniform distribution?
You really do not need to test for significance. Once you have estimated the Expected Returns, subtract them from the Actual Returns, and plot the residuals (usually in a cumulative manner). Good luck
Hi,
You may have a look at this paper:
A Note on Event Studies in Finance and Management Research.
By: de Jong, Abe; Naumovska, Ivana. Review of Finance. Jul2016, Vol. 20 Issue 4, p1659-1672. 14p. DOI: 10.1093/rof/rfv037.
Best regards
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics