• • • •
Alan Clay Data Strategy Director, LexisNexis® Risk Solutions
Back to top
The dangers of bad data on the journey to buyout
Economics have shifted to a buyers’ market: Despite nearly half of schemes (45%) pursuing a buyout, the majority of these (82%) have yet to secure an insurer. This highlights the dominance insurers have in this industry and how competition has shifted. Scheme overconfidence is rife: Though most respondents believed their schemes’ data to be in a good position – with an average score of 8.3 (out of 10) for completeness and 8.2 for accuracy – contradictions emerged in the survey among other findings. Common concerns cited included low member engagement and high instances of returned post. There is a stark lack of realism: Of those targeting a buyout within two years, 59% have yet to find an insurer. Qualitative analysis suggests many respondents are underestimating the endgame journey ahead of them. Data poses a real cost – and barrier to endgame: Among the barriers cited by firms to improving data quality were uncertainty around the level of upfront investment required. This issue ranked highly for almost a third (31%) of respondents with 16% citing a lack of internal capability.
In addition to securing the best endgame terms, there has also been increased regulatory focus on data in the pensions industry. The Pensions Regulator's director of trusteeship, administration and defined benefit supervision, David Walmsley, told Professional Pensions' inaugural Professional Trustee Focus that the regulator will be placing greater focus on ensuring schemes are well governed and have high quality data as well as focussing on decision making. "The future is knocking at the door," he told schemes and with the imminent advent of pensions dashboards, urged them to "undertake regular data reviews".
Select a chapter to explore
In this report
Data quality and completeness underpin every aspect of running a pension scheme.
Clean, accurate and regularly maintained data not only has the potential to improve a scheme’s position for buyout but it can also reduce fraud risk and even improve outcomes for members along the way. Yet, the data-related challenges highlighted in this report echo some of those we identified in research previously carried out with Professional Pensions in 2021 – revealing how persistent issues remain, despite increased industry awareness. Getting the fundamentals right – correcting errors at source, using complete information, and keeping records up to date through regular screening – is essential to build resilience and secure the best terms for members.
Executive summary
Pensions: The digital dilemma
“Whilst schemes are set to benefit from improved endgame options, barriers remain in place – particularly when it comes to data”
Scroll to explore
Foreword
Funding levels in the Defined Benefit (DB) pension sector have hit a record high, with three in four schemes now in surplus, and deficit payments down by over £10bn per year.* As a result, more DB schemes are now reconsidering their endgame plans, particularly insurer-led buyouts as well as how to take advantage of the UK Government’s plans to relax rules on how schemes use their surplus. Whilst schemes are set to benefit from improved endgame options, barriers remain in place – particularly when it comes to data. As more schemes think about their endgame, the volume of transactions coming to market has increased significantly. This shifting balance of supply and demand means insurers can be more selective, intensifying competition among schemes. As a result, the onus is now firmly on trustees to ensure their data is fit for purpose. The latest Professional Pensions survey, produced in association with LexisNexis® Risk Solutions, surveyed over 100 pension scheme professionals and found:
For the first time in decades, pension schemes find themselves in a new and unusual situation: that of a surplus.
This report examines how data-related challenges are impacting pension scheme resilience, endgame outcomes, and the journey to buyout.
Next page
*Data readiness for buy-in and buyout Guidance, PASA Data WG, 2023
The role of data in endgame
The pensions landscape has evolved significantly over the past five years, with schemes having grappled with buyout deficits, low long-term interest rates and rising volumes of demand for insurer buyouts.
With many schemes now in surplus, demand for insurers has grown threefold – yet many schemes remain unprepared, our research shows. It quickly became apparent in the results that, despite the positives of being in surplus, many schemes are in a difficult position as a result. Though almost half of schemes (45%) are looking for a buyout, the majority (82%) are yet to find an insurer.
Has your scheme already secured an insurer for this transaction?
No - not yet
Yes
Despite schemes’ apparent confidence at a top level, some trustees question how realistic endgame plans may be due to data concerns. BESTrustees president Alan Pickering points to the percentage of schemes pursuing buyout within 24 months as “not realistic at all” and questions if the level of work required on data quality is fully comprehended. “Often it will take you nearly two years to get to buy-in and it may take you another year to reach the buyout phase,” explains Pickering. “[These findings] will be an underestimate but it’s unlikely that you're going to get there in two years.”
18
%
82
Back to home screen
Chapter One A pension timebomb
This was broadly the same for schemes targeting buyout within the coming 24 months, with 59% yet to find an insurer.
A lack of realism
This highlights a significant supply and demand mismatch, with insurers becoming more discerning with regards to the schemes they choose to transact with. The result of this shift? A closer focus on schemes, their surplus levels and data than ever before. This shifts the onus onto scheme quality with data issues having the potential to derail deals before they’ve even begun. “Some insurers will simply refuse to quote schemes if the quality is not there,” says Michael Clark, a professional trustee. “If the trustees cannot offer some guarantee about the quality of the data, [insurers] simply will not quote because there are better opportunities available to them. But whilst it is a competitive market for the schemes, it is also a competitive market for the insurers – there is more business than they can cope with.” Alan Clay, Data Strategy Director at LexisNexis® Risk Solutions, agrees that without high quality data, schemes may not get a quote at all, but also notes the quality and completeness of the data will have a significant influence on the accuracy of any quote offered. “If there's any missing data, assumptions will need to be made which will mean there is a higher variability in the actual quote received from the insurer,” he says. “It isn’t just about completeness either – correcting errors in full datasets is essential,” adds Clay, “as any inaccuracies will flow through into partial data and undermine confidence. Using complete personally identifiable information enables higher match rates and ultimately better accuracy.”
“What most schemes do is to try and get the best price and then insurers will give them a couple of years to clean up the data on agreed financial terms”
Peter Rennalls, Head of New Business Delivery & Solutions at Pension Insurance Corporation
Peter Rennalls, Head of New Business Delivery & Solutions at Pension Insurance Corporation, says that “you can transact with missing data, but you may get a better price if you’ve got good experience data to share regarding your scheme.” “Schemes rarely come to the market with perfect data,” says Rennalls. “What most schemes do is try and get the best price and then insurers will give them a couple of years to clean up the data on agreed financial terms.” He notes that the key obstacles are often the data relating to the “big ticket items”, such as GMP equalisation and contingent spouse pensions. “These are likely to have the biggest impact on pricing.”
294
number of transactions completed in 2024
226
number of transactions completed in 2023
Previous page
A buyers’ market
The need for robust, accurate data goes beyond buyouts: good data underpins scheme governance and helps reduce risks across the entire lifecycle.
However, to stand out in an increasingly crowded marketplace for buyouts, schemes must do more to ensure they are attractive propositions for insurers already spoiled for choice. This has made data quality a key differentiator. According to TPR, common data such as National Insurance number, surname and date of birth needs to be accurate so that a member can be uniquely identified. Schemes also need to be working with their administrator to ensure that scheme-specific date, such as the structure, member status and salary records are clearly defined and recorded. For pension schemes, data is often thought of as the foundation of endgame success. Inaccuracies in common or scheme-specific data could impact insurer valuations. Despite this, many schemes admit their data can be unreliable, out of date or lack clarity. According to PIC's Peter Rennalls, “insurers won’t necessarily turn a scheme with missing data away – we can price with assumptions – what we can’t do with bad data is a buyout.”
said they’d be more likely to engage digitally if it provided education on how to save for retirement
63
3 /3
said that they’d be more likely to engage digitally if it gave them an understanding of whether they are saving enough for retirement
72
2 /3
Chapter Two Educating to engage
Given the increasing importance of scheme data in securing an endgame solution, respondents were asked about the quality of their data. They gave an average rating of 8.3 for completeness and 8.2 for accuracy (both out of 10). Though these appear to be good scores, they are at odds with other findings on how schemes view their data. Though nearly half (46%) viewed their data as being already of sufficient quality, a contradictory picture emerged when 16% of respondents said they lacked the internal capability to improve data quality. The same proportion admitted that data is only ever reviewed as part of the broader business, while 12% stated a lack of clarity around regulatory expectations was a barrier for improvement work.
Overconfident and underprepared: Schemes and their data
Schemes pursuing endgame may assume they are in good stead because their data is extensive and appears to be complete, but this can hide issues that need to be addressed. Inaccuracies and gaps could be lurking in a scheme’s data, which may also undermine member engagement. This was the top selected challenge for schemes in the quality of their member data, with nearly half (45%) citing poor engagement as the cause. A similar proportion (40%) had reported mail being returned after communication attempts, with 19% also singling out “inappropriate contact attempts.” This means that scheme data is at risk of becoming outdated. Recent LexisNexis Risk Solutions analysis suggest that data quality, on average, declines by 11% per annum as people’s circumstances change and given typically low engagement with their pensions, they do not inform schemes of their change in details. This means the scheme themselves – i.e. the data custodians – fail to keep up. “Like all statistics, the basic Pensions Regulator data score can be misleading,” Graham Jung, professional trustee at Pi Partnership says. “Don't be lulled by the fact that the number is good, look behind it.”
45
of repondents cited poor engagement as the cause of poor data quality
46
of respondents viewed their data as being of sufficient quality
40
of respondents reported mail being returned after communication attempts
According to PASA’s Data Readiness for Buy-in and Buyout Guidance, published in 2023, “High standards of data quality enable smooth, efficient and preferential pricing terms when pursuing de-risking projects with insurers.” However, despite an increased industry focus on data, the survey results remain largely unchanged from separate research carried out by Professional Pensions and LexisNexis Risk Solutions in 2021 – underlining how persistent the data challenge is and suggesting an ongoing lack of sufficient action thus far. Not only can poor data present security risks and potential liabilities in the future, but it can also compromise schemes’ ability to secure the best terms for their members. As the digital landscape evolves and as the role of the Pensions Dashboard comes into focus, ensuring schemes have a well-crafted data strategy is becoming crucial alongside data quality. “Keeping on top of the timeliness of data is also crucial, making sure that it's fit for purpose when you use it rather than when it's first collected – out-of-date data is going to be a huge issue when the pensions dashboard comes in. In the past 12 months we have audited data from over 100 schemes and whilst the data scores highly from a completeness perspective, comparison to our reference data highlights only 87% are living at the address held for the member,” says Alan Clay from LexisNexis® Risk Solutions.
Tackling scheme data issues
Discussions around data are increasingly on the agenda – with some schemes reporting them at every meeting. Yet the findings show that practical action often lags behind, and significant gaps remain in how schemes prepare their data for endgame.
Implementing a regular data cleansing plan for our scheme
2
Chapter Three An appetite for apps
What steps has your scheme taken towards that end goal? Please state how far along your scheme is in implementing each of the measures below
It became clear among respondents that much can still be achieved in terms of data governance. Nearly half of respondents said they discuss data at every trustee meeting, yet only 35% have commissioned a formal data readiness report, and just 26% of buyout-aiming schemes have completed a buyout-specific audit.
On average, how often does data upkeep feature in the agenda of your scheme’s trustee meetings?
This attitude was further evident in what schemes identified as the drivers to tackling potential data shortcomings. The forthcoming pension dashboard and working towards a buyout were the top selected drivers – at 66% and 65% respectively – and came ahead of internal issues like operational inefficiency or change in management processes.
Which of the following critical events would prompt your scheme to invest to improve its data?
“You do have to tackle this head on, get your hands dirty and find out exactly what's going on under the surface,” says Roger Mattingly, trustee director at the Independent Governance Group who argues against cutting corners with data quality. “It's the difference between cost and investment. To take short cuts on this is not economic; the way to create efficiencies is to get the data as clean as is humanly possible.” Here, Mattingly advocates a more proactive approach to data and for schemes to ensure they are engaging with members and that all data is accurate. This is about moving scheme data up the agenda and ensuring it gets the attention it demands. This may mean more expense, but BESTrustees Alan Pickering argues there is a potential disjoint between data completeness and data accuracy. “Data completeness is quite reassuring, but that reassurance is superficial,” says Pickering, who points out the temptation for some parties to avoid data’s inconvenient details. “One of the things that has created the gap between completeness and accuracy is a weakness in the relationship between the client and the ongoing administrator. The administrator is, in some ways, incentivised to sweep problems under the carpet in the hope these problems will never surface.”
Such problems will invariably surface, however and the journey to endgame should force schemes to confront any shortcomings in their databases. Doing so is not cheap, but there is an argument that this expense should not be negatively viewed as a cost, but as a worthwhile investment in order to secure the right insurer for buyout. “There's no point in stating it's going to cost (say) £300 to trace a member – who cares? It is absolutely an investment in the Scheme data” states Professional Trustee, Michael Clark. “If you've got data with holes in it, you'll end up having to insure the member because you cannot take the risk of not insuring them. Or if you're being brave and you don’t, then the member comes out of the woodwork a few years later after the trust is brought to a close, then the liability falls to the employer, who is not going to be happy.” At some point, issues and inaccuracies within scheme data will need to be tackled. Procrastinating could cause some to wait longer for their endgame. Trustees agree these issues need to be tackled head-on, but this report’s findings suggest many schemes are unaware of the extent of the work required on this.
Alan Pickering, president at BESTrustees
“One of the things that has created the gap between completeness and accuracy is a weakness in the relationship between the client and the ongoing administrator”
Currently formulating a plan
No plan in place
Plan in place, but no action yet taken
Plan in place and already being actioned
Commissioning data cleansing plan from scheme administrator
Conducting a scheme data audit
Commissioning data readiness report from our scheme’s administrator
Updating our benefit specification
Conducting a comprehensive buyout data audit (buyout schemes only)
0%
25%
50%
75%
100%
13%
18%
11%
58%
19%
12%
15%
49%
32%
35%
31%
21%
14%
34%
27%
29%
22%
Discussed in every trustee meeting
Discussed in 4 of every 5 meetings
Discussed in 3 of every 5 meetings
Discussed in 2 of every 5 meetings
Discussed in 1 of every 5 meetings
Never discussed
45%
5%
3%
Introduction of pensions dashboard
A regulatory focus on member data quality (not just data completeness)
Working towards a buyout
Data issues that directly impair operational efficiencies
A change management process such as preparation for a move to a different administrator
Data to improve member engagement and the effectiveness of communications/personalisation
Occurrence of a major data breach in the industry
A liability reduction exercise (improving data quality in advance)
Digitalising processes to deploy a new digital pension experience for members
Entering a master trust
Other (please specify)
20%
40%
60%
80%
66%
65%
61%
54%
46%
38%
33%
9%
Data on top: What insurers want
As schemes approach buyout, data quality is emerging as a key differentiator. With insurer capacity stretched, schemes must present themselves as clean, accurate and ready to transact. Should an insurer find any red flags it can easily move onto the next scheme, as so many are in surplus and queuing up to exit.
“Undertaking a data quality audit is win-win. It improves your efficiency in the short term and will increase your buying power and reduce running costs and reduce adviser fees when the time comes to transfer the risk”
Simply put, insurers will not want to take a risk on schemes with data that is either incomplete or hiding potential issues. Alan Clay, of LexisNexis® Risk Solutions, stresses the importance of schemes ensuring their data is presentable and dependable from the outset. “If you take one extreme – a scheme with very poor quality data – they're not going to get a quote from an insurer,” he says. “Clearly the better the data is, the more accurate the quote is going to be and fewer assumptions will need to be made. After all, if there's any missing data, the insurer will need to make assumptions about what the missing data might be hiding. Missing data also leads to delays whilst insurers determine how to best fill the gaps. In a buyer’s market they could just move onto the next opportunity.” “So, the better the data in the first place, the faster the process and the more precise the quote from the insurer.”
Data red flags
Some of the issues that will cause concern for insurers include:
Missing or incomplete data especially dates of birth and presence of spouse Outdated records or data that is clearly wrong Evidence of poor member engagement Reliance on paper-based records Generic stopgaps – i.e. people using 1st January as a DOB to fill in missing DOBs Members being obviously too old (115 + over) or too young (<18) Lack of consistent formats for fields like DOBs and postcodes
Getting to grips with data
Chapter Four The data dilemma
The road to endgame can be long and complex, with the consensus among trustees that schemes need to use their time carefully and work on getting data in order before meeting with insurers. “You need to give the administrator at least 12 months to get their data in line,” advises professional trustee Michael Clark. “You need to have those conversations in good time - you don't go to market on a Friday having just told your administrator the plan on Monday. You need reports from the administrator every month if not every week towards the deadline date with updates on what data cleansing is being carried out.” In agreement is BESTrustee’s Alan Pickering who warns that it is imperative to be thorough and to go beyond superficial metrics, ensuring all records are accurate and of value.
• • • • • • •
“The system incentivises us to just look at superficial completeness; supplier is happy, the customer is happy but they're deluding themselves,” says Pickering. “Undertaking a data quality audit is win-win. It improves your efficiency in the short term and will increase your buying power and reduce running costs and reduce adviser fees when the time comes to transfer the risk.” Regular maintenance is equally critical. Ongoing screening helps keep records current and significantly reduce the risk of overpayments. The benefits of better quality data are clear, and any shortcomings will be harder to ignore when in the crucial period of endgame pursuit. Schemes with proactive data strategies are more likely to secure better pricing and terms and this can be the first thing insurers ask about. “There might be some uncertainty from the insurer when they ask when the scheme last audited their data, because if it was three years ago and data degrades between 8% and 15% a year, that could mean close to a quarter of it being unusable or inaccurate,” points out Clay. Those schemes that don’t tackle their data with the right strategies and approach, risk being left behind and therefore are likely to pay more and wait longer.
