It’s time to talk data

Topic:

Hot topic

Date published:

Monday, 27 November 2017

Author:

Gillian Graham

The Pensions Regulator announced in November 2016 that they would be adding some questions to the Scheme Return about pension scheme data. They recently provided further clarification and issued some practical guides.

Pension trustees will need to state when they last reviewed their common and scheme specific data (previously known as conditional data), and how much of this data is present and accurate.

If trustees take their responsibilities seriously and have good administrators, I don’t think they need worry, as the Regulator seems to understand the issues. The Regulator’s ‘quick guide to measuring your data’ helpfully states “rather than measuring all this data, you should focus on data which is key to running your scheme”.

Now’s the time for pension trustees to re-focus

When the Regulator issued its original guidance on recordkeeping back in 2010 it was very specific about how the 11 basic data items (common data) should be measured, and pension trustees should be receiving regular updates from their administrators. As these items are straightforward, it is quite easy to provide a ‘data score’, although maintaining this score is a challenge. Indeed, it boils down to a constant need to keep track of member addresses - and I think there should be more emphasis on member responsibility to keep pension trustees informed of address changes.

In reality the scheme specific data is just as important and more likely to be in a mess but, as it is scheme specific, the Regulator did not spell out how to measure it. Consequently it fell into the ‘too difficult’ pile for some scheme administrators and pension trustees.

The Regulator has recognised this and the quick guides they have issued are helpful, but I suspect there will be a few people hitting the panic button when they realise they now need to report on this. Moreover, pension schemes now have to report on accuracy as well as presence of data. I welcome this approach, and the Regulator provides clear examples of the controls (including consistency and validation checks) they expect to be applied to establish accuracy.

Eating an elephant one bite at a time

In the past I have reviewed standard conditional data reports that contained the results of around 150 tests. This made it incredibly difficult to focus on what was important and meant peripheral items tended to distort the overall score. I think the Regulator’s new pragmatic approach means there is significant latitude available in how you measure pensions data, but you need to go through a rational process (that you can demonstrate) and also put an improvement plan in place.

In September, the Regulator published a ‘quick guide to improving your data’. This guide mentions that your improvement plan does not need to be complicated and suggests pension trustees and scheme managers should review their data at least once a year. As this will need to be reported on annually in future, there seems some sense in taking the ‘eating an elephant one bite at a time’ approach.

So, pension trustees, now is the time to engage with your administrator. Take time to talk to them and agree which scheme specific data items you should focus on to measure and assess for accuracy. If you haven’t previously measured any conditional data you should agree what you want to achieve and prioritise accordingly. You may, for example, decide to just focus on:

• date of leaving
• pensionable pay history
• lifetime allowance crystallisation event details
• total original deferred benefit
• tranches of original deferred benefit

You can then draw up your improvement plan to define what work is to be done and when it will be delivered. As most schemes are in the midst of GMP reconciliation, one would hope to see an improvement in the measurement from 2018 to 2019, which I believe is the Regulator’s desired outcome.

More regular contact with your pension administrator should help to build a better understanding of the state of your data, which will be invaluable for ensuring you are well placed for the future. Whilst it generally isn’t essential to go to an independent firm and commission an expensive conditional data report with its 150+ tests, it would be advisable to have an independent review from time to time. In some circumstances, this review process could shine an unwelcome light on the state of scheme data and result in a reduction in confidence in the administrator. The trustees may then conclude that a comprehensive independent review is required.

Back to opinions