Datastream & Turnover by volume

A recent post by Judith Gulpers made me aware of the fact that Datastream handles some data types rather differently than many people expect. When you do a regular search for items like Prices, Datastream returns the price on a specific day (depending on the start date). Turnover by volume is different, however. The data type works similar to other data types when you download data with the frequency Daily. When you download VO with the frequency Monthly or Weekly, however, the result is a cumulative number for the previous month or week. The start date for a download in this case does matter a lot!

Below you see an example for a download using the Google share for Months:

  • Search 1 is a regular download with frequency Daily
  • Search 2A is a download with the same starting date as search 1: 1 January 2013. The result here is incorrect as Datastream adds up the result of last Months’ trading days. The number 41708.3 for 01/01/2013 is the sum for December 2012.
  • Search 2B is a download with the starting date 31 January 2013. The result correctly starts with a cumulative number for the first Month of the year. As a comparison I manually summed the days of the months (column G) using search 1.

Below here you see an example for the Google share for Weeks:

  • Search 1 is a regular download with frequency Daily
  • Search 3A is a download with the same starting date as search 1: 1 January 2013. The result here is incorrect as Datastream adds up the result of last Weeks’ trading days. The number 6231.6 for 01/01/2013 is the sum for the last week of December 2012.
  • Search 3B is a download with the starting date 8 January 2013. The result correctly starts with a cumulative number for the first week of the year. As a comparison I manually summed the days of the weeks (column L) using search 1.

Email

%d bloggers like this: