Calculating AAR and CAARS and then testing for statistical significance

[Felix Ziegler]

Hello everyone,
i've calculated abnormal returns for my event window and now i want to calculate average abnormal returns (AAR) for every day t in my event window and also my cumulative average abnormal return (CAAR).
Therefore i want to use the following formula:

AARt=1N∑i=1NARi,t

CAARi=∑t=T1+1T2AAR,ti,t

with N: Number of observations = 151 Companies
ARi,t: abnormal Return for security i on day t
CAAR: cumulative abnormal return for the event_window

here is my file with abnormal return data (sorry, i don't have stata at home. so i've uploaded it to dropbox. hope someone will give it a look anyway )

https://www.dropbox.com/s/4xxisqckqt...turns.dta?dl=0

---

Follwing commands i used to calculate AARs and CAAR:

sort dif
by dif: egen AAR = mean(abnormal_return)
egen CAAR = sum(AAR)

Now i want to test for statistical significance. Therefore i want to use the Cross-section test provided here (see photo below).
Here are my commands i've used:

sort dif
by dif: egen AAR_sd = sd(abnormal_return)
gen test =(sqrt(151)) * ( AAR /AAR_sd) //Formel: Wurzel(N)* (AAR/AAR_SD)

so here is my question. my results for the t-value seem to be wrong. have i done something wrong? i searched for this problem here but couldn't find any answer.

thanks for your help !

Felix

[Felix Ziegler]

Sorry, here is the link for calculating the AARs and CAAR. http://www.eventstudytools.com/significance-tests#Csect

[Felix Ziegler]

Ok, problem solved....
Now i want to create my variable CAAR.
My data for my AARS looks like this.
I want to create one variable "CAAR" which is the sum of the current AAR for Date t with the AARS of the dates before. For example the value of the CAAR for the 24oct2012 is: AAR253+AAR254+AAR255
how can i do that?
thanks in advance
felix

       +-----------------------+
       |      Date         AAR |
       |-----------------------|
  253. | 22oct2012    .0016806 |
  254. | 23oct2012    .0033284 |
  255. | 24oct2012   -.0026343 |
  256. | 25oct2012    .0013081 |
  257. | 26oct2012   -.0070921 |
       |-----------------------|
  258. | 31oct2012   -.0009182 |
  259. | 01nov2012   -.0100833 |
  260. | 02nov2012    .0010243 |
  261. | 05nov2012   -.0027614 |
  262. | 06nov2012    .0042589 |
       |-----------------------|
  263. | 07nov2012   -.0007319 |
  264. | 08nov2012    .0078595 |
  265. | 09nov2012   -.0025169 |
  266. | 12nov2012   -.0049136 |
  267. | 13nov2012   -.0040264 |
       |-----------------------|
  268. | 14nov2012    .0032638 |
  269. | 15nov2012   -.0016879 |
  270. | 16nov2012    .0022905 |
  271. | 19nov2012   -.0070145 |
  272. | 20nov2012    .0043224 |
       |-----------------------|
  273. | 21nov2012    .0030437 |
  274. | 23nov2012     .004581 |
  275. | 26nov2012    .0012823 |
  276. | 27nov2012    .0046554 |
  277. | 28nov2012   -.0160497 |
       |-----------------------|
  278. | 29nov2012     .013489 |
  279. | 30nov2012   -.0006666 |
  280. | 03dec2012   -.0008592 |
  281. | 04dec2012    .0113461 |
  282. | 05dec2012    .0030912 |
       |-----------------------|
  283. | 06dec2012    -.003946 |
  284. | 07dec2012    .0003843 |
  285. | 10dec2012   -.0095248 |
  286. | 11dec2012   -.0024391 |
  287. | 12dec2012    .0019716 |
       |-----------------------|
  288. | 13dec2012    .0041912 |
       +-----------------------+

.

[Felix Ziegler]

No one can help me with this?

[Lina Schuhmacher]

Hey Felix! I am facing the very same problem now... Did you eventually manage to figure it out? Would highly appreciate if you could share your insights!