[Skip to content]

Sign up for our daily newsletter
The Actuary The magazine of the Institute & Faculty of Actuaries

Meaningful mortality

With defined benefit (DB) schemes, numerous assumptions are made to help assess the cost of providing benefits. Essentially, this is all contained within the scheme valuation. While a great deal of work goes into these assumptions, they still remain assumptions. Does anyone know what inflation will be or what the investment returns will be? Obviously not, but everyone should make assumptions that are realistic and prudent.

When it comes to mortality studies, we are able to refine the assumptions much further but, in the vast majority of cases, still rely on national or regional average and various other models. By their very nature, these models will not exactly match many schemes — most will be either side of the average. In effect, this means many employers will be funding for benefits they are unlikely to ever pay and, in theory, some may actually be under-funding. Using an average indicator is an unreliable technique when far more advanced methods are available. Having the most accurate mortality understanding as possible is critical to establishing the specific requirements of each individual pension scheme.

Our true mortality is, without question, heavily linked to our lifestyle and demographics. To put this in context, imagine two very different individuals:

The first works for a large firm of city lawyers. They work hard, but have access to private healthcare, receive subsidised sports club membership and take regular family holidays. In addition, they may have a staff restaurant offering healthy option lunches. I can list more, but you get the picture. Naturally, there are negatives, such as long hours and stress, but these are far outweighed by the positive elements of the lifestyle and career.

You then compare this with the factory worker, who works long shifts with as much overtime as possible. While it is not a stressful environment, it is a hard working environment and a tough life. The sports club is more likely to be the local pub and linked with drinking and socialising, rather than sport. He does not have the perks associated with the professional role, but still qualifies for a defined benefit pension scheme.

We may see differences in their diet, standard of living, newspaper readership, education, social class and so on, but it is already immediately obvious that their lifestyle is very different. However, in most cases, virtually the same mortality assumptions are applied. In reality, their life expectancy could easily be five years apart or more. Multiply this by the hundreds or thousands of members in a pension scheme and it will change the potential liability dramatically.

For smaller pension schemes that have considered buyout of members’ benefits, this works well assuming members are happy to provide health information; surprisingly, most are. The overall cost of the actual buyout/in can be substantially cheaper than a standard bulk buyout/in.

For larger pension schemes, a mortality adjustment is more desirable in the short term, with individual buyout potentially a longer-term plan. Currently, the individual annuity market is only just discovering this opportunity and only a few providers have the technology to make this work. However, they are moving quickly, as this opportunity develops.

In a recent mortality exercise, ITV broadcaster Scottish TV secured a considerable saving by, in simple terms, obtaining medical evidence from the pensioners and benchmarking this with the annuity rates available at the time. This data was passed to the scheme administrators, who helped the employer negotiate with the scheme actuary and trustees. While this involved a lot of work, the saving of £5m was clearly very significant. In theory, many more schemes will be in the same position and should consider such an exercise.

To give an idea of how much can be saved, for every year that an individual’s life expectancy is reduced, the scheme liability can reduce by as much as 4%, so even modest changes can make a substantial difference. The medical conditions will reduce the liability by the following amounts:
• Smoking: 5-10%
• High blood pressure: 4-12%
• Asthma: 4-10%
• Diabetes: 5-15%
• Heart conditions: 5-50%
• Most cancers: 5-80%
• Stroke: 7-40%.

Naturally, the health of the members will affect the cost of providing benefits and the expected mortality. All mortality tables will include an allowance for unhealthily lives — does that reflect the reality?

The potential savings for each individual member can be varied, so it is critical to get the extent of the condition(s) to make sure the information is as accurate as possible. Clearly, in all schemes a proportion will be in good health. Based on the reviews we have already undertaken, those with health issues have outweighed those without. In fact, in one recent survey of a scheme based in the south of the country, we found that only 28% of the members were in good health.

What has become perfectly clear for every DB pension scheme is that relying on industry averages is not a sufficiently reliable and accurate method of valuing a scheme’s true cost. Real data is available and the time has come to stop relying on averages when determining the cost of providing benefits.


Nick FlynnNick Flynn is longevity director with LEBC Group and head of the retirement adviser division