EVENT STUDY

[barts2]

dear all,
I am doing an event study comprising of 30 firms over a 50 day event window. I have used eviews to calculate the standard errors and thus allow for covariance between individual abnormal returns not to be strictly zero. I have done this by letting the regression be something like
LS FIRM C MARKETRETURN D1 D2 D3........D30 where D1 etc is a dummy of 1 for event window day one and 0 otherwise. I am also able to get the cumulative abnormal return for each firm by using the regression LS FIRM C MARKETRETURN DX2 DX3 DX4 ...........DX30 where DX2 is D2-D1 and DX3 is D3-D1. the problem I am having is calculating a test statistic using eviews for the cumulative average returns over all firms. I have no idea how to get a standard error for my value and was wondering is there any way of doing so on eviews.
regrads,
niall