EViews.com

Dear Eviews-Forum-Users,

I am completely new to this program but it got referred to me by a friend to solve problems like mine. My task is to look at abnormal returns around the unlock date of lock-up agreements after an IPO. The dataset I created consists of an Dax index (first column) and than 10 columns of company stock prices. Each company has an observation range of -50 days before the unlock date, day 0 the unlock date and 50 days after the unlock date. The stock prices are placed next to the Dax for every day. (File is attached). I would like to know which commands i need to type into the "white box". Attached is also a file on what the output is kind of supposed to look like...

My task is now to run the following regression:

Ri,t = β1 - Rm,t + γ1(X i,u)

Cummulative Abnormal Return: CAR = u(γ1)

where
Ri,t = Return of firm i at time t
Rm,t = Dax30’s return at time t
β1 = Dax30 Return Coefficient
γ1 = Dummy Variable Coefficient
u = CAR testing period (measured in days = 5)
X i = Firm i dummy for u days surrounding lockup
CAR = Cumulative Abnormal for 5 days surrounding lockup

Step 2 using a z-statistic to test for whether the CAR is significantly different from zero. The z-statistic is equal to the sum of the t-statistics for the dummy variable coefficient for the five days surrounding the lockup date divided by the square root of the total number of regressions.

Okay, i started my eviews learning process a little bit and maybe my so far findings get this "discussion" started a little.

First i added a new dummy row in my excel file. I attached the new excel file, so you can see what i mean.

Now in a first step i can get the index and single dax_x series raw data to returns with:

genr returnindex = dlog(index)
genr returnddax_1 =dlog(dax_1)

now i could continue to to it all in single steps, but i guess there is an option to do it more easy, however if i try it, i only get an error message:

for !i = 1 to 27
series retrunddax_!i = dlog(dax_!i)
next

i also tried it with

for !i = 1 to 27
genr retrunddax_!i = dlog(dax_!i)
next

not working either.

Anyhow, for a single regression now i did:

equation reg_1.ls returndax_1 c -returnindex +dummy

Would you guys say, that this would sufficiently portray:

Ri,t = β1 - Rm,t + γ1(X i,u)

Best