Transition to the new pension system: Embrace the importance of data quality

The data quality policy provides an assessment of risk attitude, permitted qualitative deviations and the design of management processes to ensure that the fund and PUO actually apply this policy. In addition, the fund must demonstrate by means of data analyses and partial observations that the data quality is actually in order. These steps have been described by the Pension Federation in the Data Quality Framework.

But what's the best way to go about it? What are the most important steps in this process towards a solid data quality policy? This was recently described by seven Deloitte pension and data specialists in an extensive article in the Journal for Pension Issues. From different angles and perspectives. The common thread of their argument in five key messages: embrace data quality, pay attention to it in the administrative dialogue, listen to your conscience and seek cooperation in the pension chain; Start on time because the outcome is unpredictable. We explain them briefly.

1. Embrace data quality and the guidelines of the Pension Federation
In the transition to personal pension assets, a solid data quality policy is more than ever in the interest of participants. In the current pension system, the retirement date is often the moment of truth: that is when the benefit/entitlement is granted. In the transition to the new pension system, this moment will be brought forward for all participants; During the entry into force, the collective pension assets are allocated to the personal pension assets of the participants on one day. In order to be able to determine the correct amount, all data must be correct and checked. After all, changing and correcting errors afterwards is time-consuming and expensive; And last but not least, it can cause considerable damage to the company's image.

Until now, most funds did not see their data quality as a major problem. This is not always justified, incidents and errors occur regularly. But those errors can be manually corrected when participants retire. With the transition to the Wtp, this situation is no longer tenable, the entry operation is too extensive for that. Pension administrators really need to embrace the topic of data quality.

The Data Quality Framework of the Pension Federation offers pension administrators a good foundation to structurally analyse their data quality and, where necessary, to correct or improve errors in the fund administration.

2. Give data quality the administrative attention it deserves
In the analysis of data quality, it is essential to pay full attention to the possible risks. Pension administrators sometimes shy away from this; Wrongly, because such an analysis and risk inventory really does not have to be as time-consuming, expensive and complex as people often think. Together with the PUO, the entire history of the fund is traced, but can focus on historical events (such as a regulation or system change), incidents and complaints and the profile of the participants. The information provided by this exercise is very valuable: it provides direction for further data analysis and the formulation of additional control measures and possibly even a renewed data quality policy.

3. Fund and PUO Board: Listen to Your Conscience
The Data Quality Framework of the Pension Federation identifies possible process steps but does not provide any normative frameworks. So a lot is left to the fund managers (key function holders) and their PUOs. That means they will have to make conscientious choices to give the participants what they are entitled to. What plays a role here: finding the right balance between accuracy and efficiency; between looking for possible errors and rising costs. Deviations must be corrected in any case, but high research costs are ultimately at the expense of the balance of all participants. An important tool in this assessment is the Maximum Permissible Deviation (MTA). This indicates when you, as a fund, will look for errors or which (minor) errors you will allow. To determine this MTA, you need to consider a number of elements. Think of the qualitative standard of the fund as laid down in the data quality policy and the possibility of correcting errors after entry can also be considered.

4. Start on time; Any data recovery is unruly
In order to have sufficient time for the possible recovery phase, pension administrators must start on time. This way, the process of entry is not jeopardised. This seems obvious. Unfortunately, some pension administrators underestimate the time required to formulate a balanced quality policy, followed by the steps of data analysis, risk assessment and eventual recovery. For example, if you want to enter in 2026, you will have to submit a plan approved by an accountant to DNB by mid-2025. Risk identification, remediation and the preparation of documentation can easily take a year, so the data quality policy must already be formulated and approved by mid-2024. So postponing is not an option. Another reason to start on time is that the outcome of the whole process is unpredictable. In principle, the range of possible errors and complex files is very wide. How much time does it take to assess those mistakes, fix them where necessary and compensate participants or not? And how do you communicate a negative correction to a participant? In any case, a good and open collaboration between the fund and PUO is crucial in this process: it will lead to a single, integrated view of the risks that the fund and its participants actually run and the considerations that the fund makes.

5. Seek cooperation in the chain at an early stage
We consider an open and integrated implementation (i.e. with a combined fund-PUO working group) of the data quality programme to be a best practice for the sector. Open collaboration and information sharing between the fund and the PUO should lead to a single, integrated view of the risks that the fund and its participants actually run and the considerations that the fund makes.

Funds must also be aware of the enormous workload that the execution of the data analysis and partial observations for entry entails for the implementer. Multi-client PUOs, especially those with a sisable client portfolio, can benefit from a common approach. Compared to a fragmented, fund-specific approach, this can lead to significant acceleration and cost savings.

But also consider starting cooperation early with the external auditor (or IT auditor) who has to issue the AUP statement. Building Audit-ready documentation from the start of the data analysis prevents hassle and re-work at the end of the process.

In short, give data quality the attention it deserves, put the subject on the policy agenda and get to work on it.