I have a data set that follows a specific portfolio monthly from 05-2004 to 05-2008. What I am trying to do is track which stocks remain in the portfolio, which ones are removed, and which ones replace the ones that were removed. Basically, trying to count the number of changes to see if the financial crises approached (2008) if the portfolio manager started making more changes to the portfolio. I'm not exactly sure how to tackle this problem. Frequency crossed my mind and maybe separating my data in to monthly data sets and then trying to merge them based on stock name. Stock names are a character value in the $14 format, and the date is in YYMMDD10. format. Any advice or thoughts would be greatly appreciated. Thanks! ... View more