winsor by none?
Posted: 06 April 2014 11:06 AM   [ Ignore ]
Newbie
Rank
Total Posts:  16
Joined  2014-04-04

Dear Joost,

I saw you have winsor by NONE in the example of restatement event study example. Will that be the same as not winsoring any variable? or winsoring all variable on the dataset??

a little bit lost here.

Thanks a lot,
Clark

Profile
 
 
Posted: 09 April 2014 10:30 AM   [ Ignore ]   [ # 1 ]
Administrator
Avatar
RankRankRankRank
Total Posts:  901
Joined  2011-09-19

hi Clark,

Winsorizing with ‘by=none’ treats the sample as one big pool of data.

The ‘by’ variable for winsorizing is helpful if the variables are potentially clustered on some dimension. Imagine for example that you want to winsorize stock return outliers, and the dataset covers both ‘bull’ and ‘bear’ periods. A ‘by=none’ would just slice off the 1% extremes, but it probably is better to repeat the winsorizing by year (‘by=year’).

best regards,

Joost

 Signature 

To reply/post new questions: Please use the group WRDS/SAS on Google Groups! http://groups.google.com/d/forum/wrdssas

Profile
 
 
Posted: 10 April 2014 11:38 AM   [ Ignore ]   [ # 2 ]
Newbie
Rank
Total Posts:  16
Joined  2014-04-04

Joost,

thanks a lot for clear up this concept!

Best Regards,
Clark

Profile