Having learnt the hard way, accountants know that changes to longevity
assumptions mean only one thing: higher disclosed pension liabilities. The
‘issue’ of increasing life expectancy is now highly significant when it comes to
financial statements for companies with final salary pension schemes.
But when you ask your scheme actuary what changes have been made to longevity
assumptions and how they affect the scheme, you soon hit a brick wall of jargon
that only the most ardent FD stands a realistic chance of being able to
While actuaries may reassure you that their longevity assumptions are based
on standard mortality tables, it is clear that different companies use
significantly different assumptions. Far from there being a standard assumption,
our research has found that FTSE 350 companies use life expectancies ranging
from 82 to 88 years old for males currently aged 65. This alone equates to a
difference in pension liability values of around 20% between the top and bottom
of the range.
Recently there have been a number of high-profile corporate deals, such as
Alliance Boots, where different views on mortality assumptions can appear to add
hundreds of millions to scheme liabilities. When this is the case, FDs need to
have a firm grasp on life (and death), actuarially speaking, as the assumptions
can bring into question the whole financial position of a scheme and as a result
can have a major impact on the deal itself.
Greater understanding of mortality assumptions used in pension disclosures is
clearly indispensable, so here is a guide to the jargon.
Longevity vs mortality
These two terms are often used interchangeably but they are really opposites
of each other. Longevity measures how long you might live whereas mortality
assesses when you are likely to die. If you have a 1% chance of dying in the
next year, then you have a 99% chance of surviving that year.
Pensioner life expectancy is a major determinant of the cost of pensions paid
by pension schemes. This is true even for members who are years away from
retirement because the offsetting effect of their dying before they reach
retirement is small. Life expectancy for pensioners is calculated from the
probabilities of their surviving in future years.
To set a mortality assumption for a pension scheme you first need to analyse
the mortality experience of its pensioners. For very large schemes, an actuary
can construct a table of observed death rates, at different ages, for
pensioners. For smaller schemes the analysis is restricted to a broad comparison
of the experience against mortality rates in published ‘standard’ tables. From
here, the best-fitting table needs to be selected. Mortality tables selected in
this way provide a base only because of the need to adjust for improvement
factors and, in some case, for rating factors too.
Where a pension scheme is too small to have enough meaningful mortality data
with which to construct mortality rates, your actuary may be able to do little
more than select as a base an up-to-date standard published table of death
rates. Such published standard tables are typically constructed from large
populations, such as national statistics, or from data pooled by insurance
In reality, no standard tables for occupational pension schemes exist,
although a major study by the UK actuarial profession is nearing completion.
Relevant factors include industry sector, earnings levels and geographical
region, and reflect the socio-economic make-up of the scheme’s members, which is
believed to be the major determinant of health and longevity. Postcodes or
amounts of pension are commonly used as the basis for rating factors for
socio-economic class. These rating factors are then used to adjust the base
Evidence continues to show that people are living longer. Therefore base
tables selected with reference to the mortality experience of current pensioners
are unlikely to accurately reflect the mortality of current workers when they
Accordingly, mortality tables attempt in a number of ways to build in
allowances for greater longevity in the future. Improvement factors may be based
on observed past improvements, but they remain subjective all the same.
More and more pension disclosures are based on standard tables with a ‘cohort
adjustment’. These adjustments are simply patterns of improvement factors
introduced by the actuarial profession in 2002. They take their name from the
phenomenon of the group, or ‘cohort’, of pensioners born around 1926 whose
longevity seems to be improving at a faster rate than for those born earlier.
The reasons for this are not fully understood. The actuarial profession
created ‘short’, ‘medium’ and ‘long’ versions of the cohort adjustment,
referring to the length of time over which we might expect longevity to continue
to improve at the rates now being observed.
There are alternative approaches to allowing for longevity improvement
factors but with any approach it is important to understand the effect of the
adjustments and not just to accept the extra jargon.
Armed with this knowledge, what should FDs do to challenge the mortality
assumptions used in pensions disclosures?
There is a wealth of mortality information available to actuaries to advise
on the setting of mortality assumptions. Because of the financial significance
of these assumptions, there should be plenty of debate about how mortality
assumptions are selected and this should be fully disclosed in company accounts.
The trick is to ask the right questions (see ‘question the assumptions’ below).
There is no one-size-fits-all solution and we expect more not less
differentiation in future. However, this needs to be backed up by
scheme-specific analysis combined with sound judgement. It falls on the FD’s
shoulders to challenge the pension scheme actuary and ensure that appropriate
allowance is made for longevity in the company’s financial statements.
Question the assumptions
You can expose the assumptions used by your pension scheme by getting answers
to the following questions:
• When was a mortality study last carried out?
• What other evidence or judgement has been used instead?
• What base table has been used and why?
• What rating factors have been used and why?
• What improvement factors have been used and why?
Martin Potter is a partner at pensions
consultant Hymans Robertson
Dennis Layton takes up the position on April 1 and will contribute to the firm’s goal of becoming the leading global professional services organisation by 2020
Richard Cartwright becomes the new head, taking over from incumbent head of office David Lemon
Brian Burke, business development director, has moved within the firm to 'develop Quantuma’s networks with Sussex professional firms'
Stephen Mills joins the Manchester office from IBM, where he spent 12 years as an associate partner in the data, analytics and cognitive consulting group