A dip or a blip: What do falling life expectancies mean for DB schemes?
At a glance:
- Reduction in mortality improvements can no longer be called a blip
- DB members have different experience to rest of the population
- Longevity swap pricing may come down, presenting opportunities
Mortality improvements have declined for yet another year, ducking previous estimates. James Phillips explores what this means for pension schemes
Life expectancy is unpredictable, yet defined benefit (DB) schemes rely on long-term estimates and short-term experiences to calculate their liabilities across decades.
Many schemes use data compiled by the Continuous Mortality Investigation (CMI), which produces annual tables calculating the average life expectancy for male and females at ages 45 and 65.
The latest edition of this table, published 28 March, shows yet another year of high mortality rates, suggesting a trend that has persevered over the last six years may not be a ‘blip'.
The average life expectancy of a man aged 65 is now 22.2 years, down from 22.5 in last year's table. For a woman, this sits higher at 24.1 years, although this is also down from last year, when it sat at 24.6 years.
Since 2013, these have fallen from 22.8 years for a man aged 65 and 25.1 years for a woman, despite life expectancy having improved between 2000 and 2010.
With this decline, how should DB trustees react, and what does this mean for the bulk annuity market in the short term?
Dip not blip
After significant improvements in life expectancy in the noughties, many expected the trend to continue into this decade. Yet, the last six years has seen some reduction in mortality improvements.
The period had been called a ‘blip' but this long-term stagnation has caused some to arrive at the realisation that this is perhaps a misnomer.
Rather, Barnett Waddingham believes that other factors may have been at play during the noughties, such as health and social care spending, accelerating mortality improvements, which make that decade the blip instead.
Accordingly, a debate continues on how much weight the industry should put on recent experience.
Lane Clark and Peacock (LCP) partner Charlie Finch says schemes will always need to err on the side of caution.
"The big question is: how much weight do you put on the recent experience?" he says. "If you set yourself a benchmark to determine value for money, should you put lots of weight on the heavy experience or less weight? That makes quite a material difference.
"Ultimately, if you're a scheme looking at this, you need to recognise that there is uncertainty in what longevity will be."
Furthermore, underlying data suggests members of DB schemes have higher life expectancies than those who are not. Indeed, both the CMI and Aon Hewitt note that DB members have seen a slower decline in mortality improvement between 2011 and 2015.
Aon Hewitt head of longevity Tim Gordon adds the data is further complicated within DB schemes, making it more difficult to analyse longevity risk.
"The curious thing is that the analysis suggests that mortality is skewed and the less well-off members of DB pension plans have had higher mortality improvements than the more well-off," he observes.
"You can't go from those numbers to deducing exactly what the drivers are - maybe they are cuts in government spending in terms of healthcare and welfare. Being a member of a DB plan may have insulated you against that.
"At the same time, maybe the better-off had some of the high improvements earlier and they are levelling off earlier than the rest of the population. Actually attributing this to a cause is really difficult and unreliable."
Prudence
These conclusions may mean trustees will have to be more prudent in their longevity assumptions, taking into account the potential differences between their members and the general population.
Punter Southall longevity specialist Dan Auton states schemes will need to consider their members' characteristics and circumstances.
"While the large dataset means the CMI model is statistically very credible, trustees and sponsors need to consider how well the characteristics of their pension scheme members match those of the general population," he says.
"Since the recent increase in deaths is less apparent among DB members, trustees and sponsors need to be wary before reducing their valuation assumptions to reflect the higher than expected mortality rates felt by the general population.
"In particular, it is important that the socio-economic characteristics of their schemes are taken into account when considering future improvements in longevity of their members."
Gordon says Aon Hewitt will take a more prudent view when advising its clients.
"We are going to take a tempered view, which means we are going to adopt the CMI 2016 projections but a less reactive version, which will be a little more prudent than the raw tables," he states.
This may particularly be the case for schemes undergoing triennial valuations over the next year, as the updated CMI figures could provide a reprieve from soaring liabilities.
In fact, an LCP analysis suggests the revised estimates could reduce DB schemes' liabilities by between 2% and 3% compared to last year. This is a significant reduction which would be welcomed by DB trustees after a year which saw deficits soar on the back of falling discount rates as a result of lower gilt yields.
Similarly, Mercer predicts the change could wipe off around £28bn of UK private sector DB liabilities.
Longevity swaps
Altogether, these changes may have an impact on the pension insurance and reinsurance market. With mortality improvements down, the market may be more willing to take on the longevity risk. Meanwhile, a reduction in liabilities could lead to schemes reassessing the value of deals.
Nevertheless, with the recent data suggesting the last few years of lower mortality improvements has not just been a blip, insurers and reinsurers may now start baking this into their pricing.
Finch says insurers may have been taking a more cautious approach over the last few years.
"Most insurers and reinsurers follow and use the CMI tables and projections, but they adjust them heavily," he says. "They adjust them for their own experience and the lives they have insured, and they also reflect their views of where improvements will go in the future.
"What we've found is they've factored in this recent experience, but they've been more cautious about projecting that."
Consequently, Finch argues that longevity swap pricing may improve this year, and trustees may want to keep on top of the market particularly if competitive pressures act in the insurance market.
Gordon agrees, stating the market remains uncertain but competition always adds an extra dimension.
"The market is in a state of flux; it had not seen a huge fall off in mortality improvements as a likely scenario, and it is still adapting to this emerging data," he says. "Is what's happening in the national population representative of the pension scheme population?
"Reinsurance markets are struggling, in the same way as pension schemes, to make sense of that, so the reinsurers are being very cautious about reducing their pricing. Yet, they are looking over their shoulders to see what their competitors are doing."
Despite life expectancy having fallen since the start of the decade, the underlying data of DB members' experiences adds a significant complication to how DB schemes make assumptions and calculate liabilities. Therefore, schemes will need to act with some prudence.
However, these latest CMI figures do present an opportunity for transferring longevity risk; if trustees see value in a transaction, now might be the time to get on with it.
More on Defined Benefit
Trustees have 'false confidence' in their scheme strategies
Nearly every trustee is confident of the next stage in their scheme’s strategy, despite almost an equal number being forced to consider replacing plans within the prior 12 months, according to research by Barnett Waddingham.
Companies overstate liabilities by £60bn amid Covid deaths
Companies could be overstating their pension liabilities by up to £60bn due to their life expectancy assumptions, according to XPS Pensions Group.
Updated: Landmark GMP ruling 'significant challenge' for schemes
Defined benefit (DB) schemes that provide GMPs must revisit and, where necessary, top-up historic cash equivalent transfer values (CETVs) that have been calculated on an unequal basis, a landmark court judgment said last week.
FCA must stop DB transfer history repeating
Regulators must act now to impose some "proper regulation" to stop another defined benefit (DB) transfer advice disaster, saysTim Sargisson.
DB funds as an economic lever could derail schemes from responsibilities
Opportunities for defined benefit (DB) schemes to pursue investment approaches that help repair the UK’s economy cannot stand in the way of improving member outcomes, Aegon says.
