Datastream instructor training 20190724

Reviewing a variety of recent student enquiries

Last session’s challenge

Last session we looked at custom headers to combine time series and static request datatypes. I set a challenge to find which sector has the most employees in Colombia in 2017.

The learners used the following steps to answer the question:

  • Create a time series request.
  • Use the series ‘WSCOPECB for active companies in Colombia.
  • Use the datatype ‘WC0711 for number of employees.
  • Add a custom header with item ‘CLS or ‘INDM4 for industry code level 4.
  • Dates from 2017 to 2017, frequency ‘Yearly’.
  • Select to ‘Transpose’ the output, so the companies appear in a column not a row.
  • There are many results. Sort them and remove blanks in Excel.
  • Bonus points: use a pivot table to show sum of number of employees for each sector.

The answer was Banking.

Past enquiry: a data type is not available for all equity indices

We looked at a real, recent enquiry. I explained how I answered it.

We were asked why the volume (VO) datatype was returning an error when looking at two particular equity indices for Malaysia.

At first, I replied and asked for clarity of which dates were being requested. Often, an error shows when there is data but not for the date range requested.

We have to interpret the student’s enquiry. They might not type the name of the index exactly as it appears in Datastream (could be a typo or synonym).

After finding relevant indices, we can look at which datatypes are available, and from what date.

  • Click on the series name or blank space to show a preview of it at the bottom of the Navigator window.
  • Under the ‘Headline’ section, you see some of the datatypes available, such as ‘P’ or ‘RI’. Click on ‘More…’ to see a list.
  • The ‘Turnover by Volume’ code ‘VO’ datatype was not listed.
  • For some of the indices, a similar item ‘Turnover by Volume (Datastream Calculated)’ code ‘DSVO’ was listed. This might be good enough.

I compared it to a common, familiar equity index, the FTSE 100. That did indeed have the ‘Turnover by Volume’ code ‘VO’ datatype.

Less commonly used, more obscure series may not have the same range of datatypes available in Datastream. Sometimes they can be found in other sources.

Past enquiry: ASSET4 does not include data for Argentinian companies, why?

A student was using the ASSET4 company sample sheet. The workbook includes a data sheet with many companies covered, and which country they are registered in. There are companies from around the world, but none from Argentina. The student wanted to know why this is, are there data being withheld by the companies, in the collection process, or is something else going on?

I had to approach Refinitiv for an answer. They provided me with some documentation about the collection process, how they look for high quality data. There was nothing about Argentina in the response.

I asked the student if we should look further into this with Refinitiv. This enquiry is still ongoing.

Past enquiry: finding research and development impact for companies involved in cross-border deals

This is a more complex enquiry that involved using Thomson ONE to get a list of mergers and acquisitions where the target company was in another country to the acquiring company (cross-border).

We have to request the DS Code or Datastream Code for the target and acquiring company in Thomson ONE. This field can be used to cross-reference with Datatream.

Datastream has a datatype for Research & Development (R&D). The student explained that the R&D impact is calculated by dividing R&D by turnover or price, as a percentage. There is a field already in Datastream called Research & Development/Price. This is the correct field to use, the calculation is already done.

The process of mapping Datastream data to Thomson ONE data is one for the student to complete. They can use a VLOOKUP in Excel, but it is often better to use a more reproducible approach with software like Stata, SAS or Python.

It is possible to get the R&D impact value for a list of companies that you make from a Thomson ONE download:

  • Get the DS Code column.
  • Remove duplicates.
  • Use the ‘Create List from Range’ function in Datastream.
  • Perform a time series request on that list.

It is possible to get the R&D impact value for a list of all companies in a country, using the wscope codes we have looked at before. You can repeat this for multiple countries in a request table. Take care when copying the codes between rows, make sure the number in the datatype code does not increment! Completing this task is considered advanced and beyond the skills expected of assistants.

A look at Eikon

We used the rest of the time to look at our trial access of Eikon, how it compares to Thomson ONE.

No challenge this time

We will not have another Train the Trainer session for several weeks (after summer), so there is no challenge this time.




‘Train the trainer’ materials for the Library's specialist financial and business databases at The University of Manchester.

Recommended from Medium

A Conversation about Data Science

Evaluation Metrics in Machine Learning Models using Python

What learning unsupervised word embeddings across 51,000 journal articles can tell us about…

Toward a False xkcdization of Data

Mistakes I Made In My Machine Learning Career

How this scale-up keeps innovating with data-driven decisions

Joint Probability vs Conditional Probability

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Phil Reed

Phil Reed

Librarian Data Specialist, The University of Manchester. Supporting teaching, learning and research with financial databases, digital skills and scholarship.

More from Medium

5 Laws of Gold — The Richest Man in Babylon

Kindle Support Number Toll Free 1–815–940–5701

Charles Schultz Museum (Large Snoopys, Studio and Ice Rink!)

The Power of a Narrative — Youth Climate Action Solutions