# Share repurchase

I am doing a project for share repurchase in the US. I am trying to estimate the operating performance before and after the repurchase announcement. And I am trying to adjust the performance to the industry level. The paper saying the results was estimated by the two tailed Wilcoxon test. Regarding this test, could I just test the unadjusted performance as a whole and then the adjusted performance as a whole? or do I have to put them as two different samples to compare? And how to use Stata to obtain such results? Do I have to find a code for that?

• It might help if you explain your data a bit. Do you have performance of firms before and after repurchase announcements? What does "adjust" mean? – Dimitriy V. Masterov Oct 15 '13 at 1:43
• You need to change the title and edit the question. You seem to be interested in application of stat tests, but your title suggest a larger out-of-scope topic. You should focus the title and the text on the statistics test of interest. Another reason to be focused is that share repurchase study must take into account certain issues that are well known in corporate finance field. For instance, operating performance clearly impacts share repurchase decisions, so you got endogeneity problem to deal with. – Aksakal Feb 16 '16 at 15:18

In Stata you can use ranksum for that. As to the test, you can compare the performance before and after, which will give you two samples. You can report the results for both industry adjusted performance and unadjusted performance. But, adjusted performance makes more sense, since you are measuring the performance against your competitors.
• It sounds like the OP has matched data, so it may be better to use the Wilcoxon matched-pairs signrank test, rather than a test that assumes independence over time. – Dimitriy V. Masterov Oct 15 '13 at 1:37