Guaranteed Minimum Pension (GMP) equalisation has been on every pension scheme trustee's radar since the 2018 High Court judgement that schemes are required to undertake it, though it's been the elephant in the room for many years.
Schemes needed to equalise Normal Retirement Ages nearly three decades ago, following the Barber ruling in May 1990. GMP equalisation is different, however. As the calculation of GMPs is set out in legislation, it's not possible to alter the GMP benefit itself. Instead, trustees have to mitigate its unequal effects by amending the non-GMP pension.
At Capita we've completed more than 100 GMP equalisation projects, mainly for schemes that were winding up or buying out benefits with an insurer. We also led an exercise to equalise GMPs for the Pension Protection Fund's backbook, covering about 220,000 members in 500+ schemes.
We would like to share with you what our experience has taught us about two areas that are key to successful equalisation: data and calculations.
Data, data, data
In our experience, many pension scheme trustees don't prepare their data in advance of starting their GMP equalisation projects. But it's much more efficient if you ready your data for GMP equalisation well in advance of starting the calculations. You should also consider cleansing your scheme data more thoroughly in a single exercise to help you with your journey plans and running the scheme in the future.
The Pensions Administration Standards Association (PASA) also sees early preparation of data as critical to a GMP equalisation project's success. In July 2019 it published a Call to Action guide that includes a prominent section on data; it echoes our experience that GMP equalisation may require more data than administrators normally hold for the day-to-day running of their scheme. In respect of pensioners and dependents, for instance, administrators only need to have enough information to be able to increase pensions and pay spouses' pensions. They often don't have data such as pensions valuations at date of leaving and retirement for current pensioners or information relating to deceased members for current dependents.
When we carry out a GMP equalisation project, we need key member data. But our pragmatic approach means that we need less data for our calculations than most other organisations would.
Reducing reliance on data
Our pragmatic approach to calculations for pensioners and dependents reduces our reliance on data that may not be available. We're able to use key data that most schemes hold electronically on their administration systems. Some schemes may need to obtain some additional data from backing files; however, we've carefully designed our approach to minimise this requirement.
Starting with the current pension, this tried-and-tested method ensures that rolling back and forward is done in a consistent way, minimising distortions and the effect of any assumptions. This means that the revised pension can be calculated without extensive, costly and time-consuming data mining.
For current pensioners, we start with the pension in payment and roll this back to retirement by removing pension increases. We then remove the effect of retirement options like cash commutation or early retirement reductions and, for those who retired from deferred status, roll this unreduced pension amount back to the date of leaving pensionable service.
If the scheme's pensions have not been rectified following the GMP reconciliation exercise, at this point we apply any changes required from this exercise to the pension.
We then roll the pension forward by reapplying the retirement options and the pension increases to both the same-sex and opposite-sex benefits.
Alternative calculation approaches
A full reconstruction, or other calculation that starts from the member's date of leaving, is only possible when a significant amount of extra data is available, typically obtained through a large data mining / extraction project.
This type of approach is likely to be very expensive and time-consuming because you need to have full and accurate data and to make assumptions which could materially affect the rectified pension.
