Forum Navigation: Select WHO'S ON: 02:34 AM General Forum Technical Forum Economics Forum Numerical Methods Forum Trading Forum The Quantitative Finance FAQs Proje... Student Forum Book And Research Paper Forum Programming and Software Forum The Quantitative Finance Code Libra... Careers Forum Events Board Brainteaser Forum Off Topic Forum and Website Bugs and Suggesti... Wilmott / BookMap Competition

 new topic
 search
 jobs board
 magazine
 news
 help
 file share
 audio visual
 articles
 blogs
 wiki
 forums
 home

 FORUMS > Student Forum < refresh >
 Topic Title: cumulative abnormal returns in event study Created On Mon Mar 01, 10 06:18 AM Topic View: Branch View Threaded (All Messages) Threaded (Single Messages) Linear

yipyy
Junior Member

Posts: 6
Joined: Feb 2010

Mon Mar 01, 10 06:18 AM

Is it a must to calculate test statistics for cumulative abnormal returns in event study? If yes, how to do it by using Eviews?

csa
Senior Member

Posts: 296
Joined: Feb 2003

Tue Mar 02, 10 05:05 AM

Yes, because you don't know whether the cumulative abnormal return is significantly different from zero or not. For example, if you are calculating a two-day window for a certain event and one day was positive and the other day was negative, then the resulting CAR could be statistically insignificant, statistically significant and positive, or statistically significant and negative. Another example would be if you are looking at a two-day event window and the days are individually positive (negative) but not significant but the CAR could be positive (negative) and significant.

I don't know if there is a function in E-views to do this. But, you can read the reply I posted under your thread titled "event study" for calculating t-statistics. Note that there is no single way to calculate t-statistics for event studies, so you may find different methods from different sources. In practice, the resulting t-statistics aren't too different from one another.

yipyy
Junior Member

Posts: 6
Joined: Feb 2010

Tue Mar 02, 10 05:26 AM

Thanks CSA.

Another question.

For the CAARs. If I use 31 days event window (day -15 to day +15, where day 0 is the event date), the way that I get the average CAAR for day -5 to day +5 , for example , should be divided by 11 ( day -5 to day +5)or 31( total days in the event window)?

Edited: Tue Mar 02, 10 at 05:33 AM by yipyy

csa
Senior Member

Posts: 296
Joined: Feb 2003

Wed Mar 03, 10 04:09 AM

Why would you want to get the average CAR over the 11 day window? Shouldn't you be cumulating the CARs over the event window?

Typically, you would do the following. For each company, you would cumulate the abnormal returns during the event window (whether it is 31 days or 11 days). Then, the cumulative average abnormal return is averaging the event window CARs across the firms in your sample.