eventstudy2 command: Estimation Window spans both before and after the event window

I am using the eventstudy2 command. In my research design, the estimation window spans before and after the event window.

The event window is time t₀, while the estimation window is composed of values t_-100 to t_-1 and values t₁ to t₁₀₀. The eventstudy2 command is as follows:

Code:

eventstudy2 firm_id date using security_file, returns(daily_ret) mod(COMEAN) evwlb(0) evwub(0) car1LB(0) car1UB(0) eswlb(-100) eswub(100) replace

The variable of interest is daily_ret; firm_id is the firm identifier; and the Constant Mean Model is being used. The upper and lower boundaries of the event window as well as the window for calculation of Cumulative Abnormal Returns is only the event date itself, i.e. (0,0).

I have specified the upper and lower boundaries of the estimation window as eswub(100) and eswlb(-100) respectively.

My question is whether this is this the right way to do it, keeping in view my research design as stated above?
Is Stata including the value for t₀ as well, while it calculates the mean for the estimation window? For the event study, the values within the event window should not be part of the estimation window.

Share on FacebookPlus on Google+