Four data-cleansing steps to improve your scheme

Poor quality data can frustrate a scheme’s business plans. It can also increase costs. Anecdotally, poor data can add 5 per cent to the cost of buyout. It can also attract unwanted attention from the Pensions Regulator.

Here are four key steps you should take to cleansing your scheme’s data:

1. Assess the quality of your data and identify areas of deficiency.
2. Prioritise work on data-cleansing, bearing in mind your scheme’s business plans.
3. Identify the resources required both to carry out the cleanse and to manage the project.
4. Once complete, obtain an independent endorsement of the quality of your data.

Why is this an issue now? Given the passage of time, scheme mergers, changing administrators and benefit changes, the quality of data has become eroded and deficient for some pension arrangements. 

Derisking, and all the different options this entails, plus the cessation of contracting-out in April 2016, are two key drivers for the need to review your data. Poor data can be a barrier to future action. 

Many trustee boards and sponsors now treat their pension arrangement like a business in run-off – they need to take action at appropriate times and if they do not have the right data to do this, this may prolong the lifespan of that arrangement.  

The regulator has for some time been trying to raise the standards of record-keeping.

In its July 2008 publication, Record-keeping: A Consultation Document, it noted that “the problems [with record-keeping] often crystallise at events such as wind-up, assessment for entry into the Pension Protection Fund and buyout. In the last of these, poor data can add 5 per cent to the cost of buyout”. 

Reporting on the quality of pension scheme data is now more commonplace in trustee meetings. This was not always the case. 

Action plan

It is proven that good data saves you money and time. Many pension arrangements have made a good start in considering whether they have good data. This has not only been driven by the regulator and its guidance, but also by the numerous derisking exercises over the past few years. 

For instance, it is not uncommon for buyout prices to be impacted if it can be proven that data is accurate and up to date. Having undertaken a data cleanse also demonstrates to insurers that pension scheme trustees are giving serious consideration to derisking exercises. 

As a starting point you should produce a data plan, which should be aligned with your business plan. Your business plan will incorporate the objectives for the pension arrangement, often agreed with the sponsor because of the strategic and financial nature of some trustee decisions and actions. 

The data plan should contain the following steps:

1. Confirm the quality of the data with details of any deficiencies. Consider reasonableness testing, bearing in mind that the presence of data does not automatically mean that it is accurate;
2. Consider the major exercises likely to take place over the coming years, eg cessation of contracting-out, and the action that should be taken in these specific areas to ensure data is correct to support future operation;
3. Identify the resources required, such as from your pensions administrators, who may in turn use a tracing bureau, and the timescales for the completion of the different tasks. Confirm how the plan will be managed and all stakeholders kept informed of progress and issues.
4. Consider getting an independent endorsement of the data quality, though appreciating there will be a cost to obtaining this.

You cannot manage without data, so measure its quality. Consider what data you need to support the business plan objectives, and develop a data plan and take action in accordance with this.

By taking these simple steps, your data will support your management of your pension arrangement and not detract from it.

Aiden Coloe is a partner, and Chris Holly a consultant, at LCP