Alright thanks, I'll try that ... View more

I have attached a text file with one company only. Each week is assigned a number and I just need the weekly returns for each one of them. I have other companies too with returns in the same weeks as the one attached so it needs by week and by permno and by year as well cause you'll find the same number assigned to a  week every year. Does this help?     ... View more

Maybe I can share my data set to show you what I mean. Can I share my dataset or maybe screenshot would do? ... View more

Sorry, that works for my example but I just realised it won't work (I think) for what I actually have. Like this: PERMNO DATE RET Weekly Ret 10158 28/04/2014 -0.01 10158 29/04/2014 -0.006 10158 30/04/2014 0.006 10158 01/05/2014 -0.009 10158 02/05/2014 0.002 -0.16983 10158 05/05/2014 -0.04 10158 06/05/2014 -0.006 10158 07/05/2014 0.002 10158 08/05/2014 -0.003 10158 09/05/2014 0.005 -0.04195 10234 28/04/2014 -0.02 10234 29/04/2014 -0.005 10234 30/04/2014 0.001 10234 01/05/2014 -0.007 10234 02/05/2014 0.003 -0.027850 So I've added an extra week for 10158 cause for each company I've got multiple consecutive weeks. Hope that makes sense? ... View more

Just weekly 🙂 So for 10158 I should get -0.01698. That's (-0.01+1)*(-0.006+1)*(0.006+1)*(-0.009+1)*(0.002+1) - 1 For 10234 I should get -0.02785. That's (-0.02+)1*(-0.005+1)*(0.001+1)*(-0.007+1)*(0.003+1) - 1 ... View more

yep my data is sorted so that works too 🙂   Thanks! ... View more

Excellent that works 🙂    Thanks so much for you help! ... View more

Yes that's right 🙂 ... View more

Alright so I want to keep the ID if it has at least 120 observation for a given event date.    So if one ID has 1 obs for one date but 121 obs for another date I only want to delete it where it has 1 obs but keep it where it has 121 (or more).   Is it clearer?     ... View more

Yep that works. Thanks a lot, I really appreciate it!   What do you think I should do if I want to get rid of the ID that have less than 120 for a given event date?  ... View more

yep that works.   Thanks so much! ... View more

Thanks for that, that helps. So the code I got previously does work, however ti actually doesn't do precisely what I was after.   for example, I get this:   n       total_buyvolume      avg   1        38224.373                 . 2        50590.5011               . . . . 252    292598.3558             . 253    550276.994          3078902.8083   the problem here is that the avg figure on observation 253 is from observation 2 to observation 253. However, I'd like the avg figure on observation 253 it to be from observation 1 to observation 252. Meaning that I'd like to have the previous 252 days on a given day without including the figure on that particular day. Does it make sense? How do I do that?  ... View more

I think I can answer no to all your questions. ... View more