Trustees must carry out a data and benefit analysis before looking to complete a buy-in or buyout says Charlotte Moore
When transferring a closed defined benefit scheme to an insurance company, it’s likely the trustees will put considerable effort into putting the right investment strategy in place to achieve this goal. However, it’s worth also allocating some resource to ensure the scheme’s data is up-to-date and benefits are correctly calculated as this can make a significant difference to the premium paid.
If the scheme’s member data and benefit calculations are accurate, the insurance company can produce a precise quote for transferring the risk onto its balance sheet. If these checks have not been carried out, there is the risk the insurance company might have to pay out larger benefits and for longer than it initially assumed.
Matthew Crewe at Xafinity, says: “Trustees should carry out a full data and benefit analysis as part of the process of implementing a buy-out or buy-in from an insurance company to get the best possible price.”
Pension schemes should not underestimate just how much incorrect data can increase the cost of a de-risking exercise. Monica Cope, chief operating officer at Veratta, says: “It would be possible to misjudge the size of the scheme liabilities by up to 5%, and possibly up to 10%.”
Although 5%-10% does not sound that large, this has a significant impact on a scheme’s funding strategy. “This would make it much harder for a scheme to accurately manage their liabilities,” says Cope.
If a scheme does not carry out a thorough data audit, it could result in the insurance company not being prepared to give a quote for de-risking project. Tiziana Perrella, head of buyout at JLT Employee Benefits, says: “While this is unlikely to happen to a larger scheme, this can be a real problem for smaller schemes.”
It would be unrealistic, however, for a pension scheme to aim for perfect data – that would simply be too costly a task. Instead, trustees should focus on those areas which will make a significant difference to the cost of the insurance.
One of the key areas to get correct is equalisation. This can be a highly complex area as it involves equalising pension benefits between the genders. Schemes had to provide equal benefits from the early 1990s, even though the state retirement age for men was 65 and 60 for women.
Many schemes thought they dealt with these equalisation issues in the early 1990s when the initial ruling on gender equalisation was made, but this may not be the case. Up to date reviews by legal advisers often unearth problems which need to be resolved. “Problems could be created either by inadequate scheme documentation or changes to retirement calculations over a long period of time,” says Anna Smithson at Xafinity
Key risks of not carrying out a data and benefit analysis:
If trustees do not undertake a review or analysis of data and benefits they risk:
- Over or under paying member benefits.
- Claims from employees that can involve the Pensions Ombudsman and a considerable amount of time and expense.
- Under or overstating scheme funding liabilities,
- Having to pay a more expensive insurance premium in the case of buy-out or buy-in.
- Increased expense to cleanse data and correct benefits because the insurer requires a more detailed review - in order to be confident that they are receiving accurate information.
- Exacerbating existing calculation problems. In other words, the longer you leave a problem the worse it gets and the more expensive it becomes to fix it.
In general members are now treated much more generously than they were in the past. This means that previous retirement calculations need to be re-visited and benefit payments will probably be increased.
Some issues are complicated because data required to calculate them is not even contained on the electronic record. For example, the number of future beneficiaries and their benefit terms has a significant impact on the price of a buy-out or buy-in.
Phil Titchener, head of data solutions at Towers Watson, says: “The problem is that most schemes will not usually calculate these benefits until the member of the scheme or has died.” However, providers would like to know exactly which benefits it will have to pay.
“Effectively the schemes need to bring the work forward and calculate those future benefits for all the scheme members in bulk,” says Titchener. That would be a straightforward exercise if all the pieces of data were available to calculate the benefits. “But that’s not usually the case,” he adds.
Given the complexity of this task, some schemes may be tempted not to bother. But that would not be a sensible decision. Titchener says: “If the insurance scheme is not provided with this data, it will make prudent assumptions.” This could result in the scheme being charges a punitive premium.
For example, insurers will assume that 80% to 90% of the members of the scheme are married and there is an average age difference of three years between males and females. However, certain industrial sectors do not comply with these norms. For example, bankers tend to have younger spouses – in other words, spousal benefits will need to be paid for longer.
The number of female part-time workers can also create another curve ball that insurers need to take into account. These workers often tend to be affluent and therefore likely to live for longer which means a greater longevity risk.
Titchener says: “Insurers have been burnt in the past about making the wrong assumptions and so are much more likely to charge a higher premium if the scheme’s members are likely to differ from normal assumptions and accurate data is not available.”
These skews away from normal population samples are more likely to happen at smaller schemes where there are fewer members which could be concentrated in particularly expensive insurance cohorts.
Perrella says: “While an insurer will be happy to help them to cleanse their data, they are less likely to want to engage with a smaller scheme until this has been done.”
Other issues which should be checked include changes in scheme factors, retirement calculations, accrual rates, or key rule definitions. “Significant differences between member’s categories should also be examined,” says Smithson.
Adding to the complexity is the fact that Guaranteed Minimum Pensions are in the process of being reconciled as ‘contracting out’ will end in April 2016.
Once a pension scheme decides to go ahead with the data and benefit analysis, they can choose either to use their own administrator or to appoint an independent contractor.
There are significant benefits to using an independent administrator. Cope says: “It’s not always the best idea to use your own administrator. An independent administrator will be better able to test the validity of their calculations.”
Smithson agrees: “If the scheme decides to use their existing administrator, it’s possible that not all the issues will be spotted.” A benefit and data audit is more efficient when undertaken by a data and benefit analyst. It’s much easier for a specialist and independent analyst to be objective and spot data gaps and calculation errors.