Cross sectional analysis of dividend payments.

[rayshonali]

Okay! So I've the data of different firms with the dates on which they made a dividend change. Other data that I have is of the historical stock prices of the firms and the fama french factors. I need to perform regressional analysis so that I can find the coefficients. My problem is that my estimation window should be -310 to -10 days from the event but the event dates are different for each firm. How can I regress this considering -310 to -10 days for each of the firms without doing the obvious way of performing it one at a time? Any help in this regard would be much appreciated. Thanks! :)