PASA (Pensions Administration Standards Association) recently released guidance on data readiness for buy-ins, which was an excellent summary of the steps clients need to consider when planning the data elements of the endgame strategy. As my colleague Angela set out in the first in this blog series, we have pursued the improvement of pension scheme data with dogged obsession.
‘The PASA guidance is helpful and reinforces what we have been advocating, and implementing, for years.’ Angela Burns
It’s now time to roll up our sleeves and look at what actually takes place during a data cleansing exercise.
We usually approach these exercises in a structured three-step manner, focussing first on the data and then moving on to looking at the benefits themselves.
1. Data Audit
The first step is to get a full understanding of the scheme’s position through a data audit, bearing in mind what will need to be passed to an insurer. This is likely to require a more in-depth examination than your usual common and conditional data checks.
Important things to look out for here are the “corner cases” and making sure that things like pension sharing orders, transfers in, and any special benefits are recorded in a consistent and easily extractable manner.
This is also the opportunity to check the quality of “experience data” and making sure that good records exist for former members of the scheme. Any missing data items can be identified and sourced from archive records where available. We’ve built these checks into our pensions administration and actuarial system, meaning that our clients and us can see how cleansing is turning into results, in real time.
2. Member Self Data Check
For the second step, the results of the data audit will be improved by allowing members the opportunity to check the data against their own records. This also provides an important piece of risk mitigation against challenges from members once their benefits have been insured.
Traditionally this exercise is done through a paper mailing, but it is increasingly common to use online portals to allow members to view their own data. Of course it’s important to allow members who are not able or comfortable to use online routes the same opportunity to review and challenge their own data through the paper based route.
3. Benefit Audit
After those two steps have been completed, you can have a good level of confidence that the basic data held for members is as accurate as it can be. But what about the benefits themselves?
Defined Benefit pension schemes are complicated and even a simple retirement calculation can have dozens of data points. The most significant problems that can occur are systemic issues around how the rules of the scheme have been applied in practice, but given many legacy calculations have been done by hand, or on only partly automated systems, the potential for errors needs to be seriously considered.
A good way of ensuring any systemic issues are identified is a benefit audit that reviews a representative sample of calculations from first principles, and checks that all key items within the benefit specification have been captured in practice. It’s not only important to check different categories of members, but also a range of time periods – many pension schemes have been administered by many different parties over their lifetimes.
While this kind of calculation review can bring to light issues that require careful consideration and resolution, it’s better to be aware of these challenges at the earliest possible point in the process and to allow time for them to be managed carefully and sensitively with proper advice, than rushed towards the end of the buy-in process. Given the heavy-duty insurer warranties that trustees must sign as part of risk transfer exercise, we feel the value from such a review is considerable.
While any of the above exercises can be carried out in isolation, we find the most effective way to approach the data and benefit work required through a risk transfer transaction is to put in place a full end-to-end plan. This ensures that all parties involved are clear about what work will be carried out and when and who is responsible for the management of each element.
Look out for our next blog in the series that will talk about maintaining data quality over time, so it’s ready and waiting as and when you need it.