While it’s best practice for the incumbent administration team to run regular checks on data and benefits for accuracy and completeness, BAU activities can often get in the way of this, meaning that checks are watered down or not done at all. If issues are not identified and resolved quickly, they may become the root cause of a wider systemic issue.
A regular independent and in-depth check can mitigate the risks of issues not being identified or even sidestepped. Therefore, it’s good practice that checks are completed at least annually.
Independent checks should be completed by data specialists and pension calculation experts. Third party input can validate information being provided to members and Trustees and also provide an additional perspective, for example ensuring the correct interpretation of Trust Deed and Rules has been applied.
Of course in a perfect world, processes, systems and procedures should have sufficient checks built in to avoid issues occurring or mistakes happening. In the real world, when working on a case by case basis, things do and will go wrong.
Running checks independently, in bulk and at a member level, can be extremely beneficial to keep on top of your dataset; the checks should be repeatable and cost effective. Often a two-pronged attack is best, one to review base data and the other to review the pension scheme benefit calculations and the application of Trust Deed and Rules:
1. Data Assessment
TPR has set out guidance on common, conditional and numerical data checks but do they go far enough?
Looking at your data for missing entries, obvious errors, illogical date sequences or implausible results is one step that can be done quickly and will provide a degree of comfort. Underlying each pension scheme administration platform will be a database; the data can be extracted and once some configuration is completed, a set of results can be produced.
DataSure, Equiniti’s data analysis and migration software, has a bank of over 600 tests and checks that can be applied to any given dataset. Results are presented in a way that indicates whether an issue is systematic or one-off, and the likely impact of the data errors.
The data assessment step is an important one during the assessment journey although it shouldn’t be a one-off exercise completed in isolation. Once you have established a baseline, checks should be repeated frequently (e.g. annually) to ensure continuous improvement and that no new issues are introduced into the dataset.
2. Benefit Audit
Checks on scheme data can highlight issues but, when run in parallel, a benefit audit can be especially beneficial.
Reviewing calculations at a member level allows the auditor to get underneath the system data itself by reviewing case paperwork and the calculations that have been performed. Cross-checks against the Trust Deed and Rules that governed the scheme should be completed, for example a return to first principles.
Checks should include areas such as pensionable service, final pensionable earnings and pensionable service splits. Not all details may be contained in the Trust Deed and Rules, e.g. pension commutation factors, early retirement and late retirement factors, so cross checks against other scheme documentation will often be required.
Selecting a sample of members to ensure the audit has sufficient coverage is important and should be done with the input of scheme stakeholders. Equiniti has a well-tested methodology to assist in selecting audit samples.
Having an independent review of your data and pension scheme benefits is good governance: at best it will provide a clean bill of health for your data and benefits, whilst at worst it will help identify issues that need to be resolved. Resulting in cost savings in the long run and providing Trustees with the correct information for reporting purposes and the information to make key improvements.