‘It will be easier to select the appropriate table for a particular pension plan, based on their socioeconomic factor,’ says Normandin Beaudry’s David Dupont
Canadian pension plan sponsors are facing a fresh reckoning on mortality assumptions, according to new mortality tables released by the Canadian Institute of Actuaries.
The research found life expectancy for 65-year-olds has increased by about one year compared to the previous set of assumptions, translating to a two to three per cent increase in pension liabilities and costs.
"It's good news that people live longer,” said David Dupont, principal in the pension practice at Normandin Beaudry. “But for plan sponsors specifically, if people live longer, it means that pension plans have to pay more. More annuity and more payments, so it's costlier.”
The Canadian Pensioner Mortality 2024 tables are only the second set of mortality tables built on Canadian-specific experience. Before 2014, actuaries relied on American tables to set mortality assumptions for Canadian pension plans, explained Lydia Audet, who's also principal in the pension practice at Normandin Beaudry.
According to Dupont, individual pension plans lack the data volume needed for credible analysis, so the CIA pools data from registered pension plans across the country to build tables all actuaries can use in their valuations.
When the first Canadian tables were published a decade ago, they revealed that Canadian pensioners lived longer than US tables had projected. That raised questions about whether the 2014 analysis overstated longevity.
Dupont said the 2024 data puts that debate to rest. Most of the liability increase stems not from the base mortality tables themselves but from the CIA's updated mortality improvement scale, also published in recent years.
When only the base table changes, male liabilities can actually decrease by about 0.5 per cent, depending on plan demographics.
"Most of the 2-3 per cent increase in IVT comes from the new mortality improvement, which is more optimistic about improvement in the future," said Dupont.
The 2014 tables drew a simple line between public and private sector plans, and whichever sector a plan belonged to determined the table it used, Audet explained, adding that framework has since been replaced.
According to both experts, the CIA's research group found that mortality differences have less to do with whether an employer is public or private and more to do with the type of work people perform and their socioeconomic standing, like a blue-collar worker in the public sector, has a different mortality profile than someone in a white-collar role.
"They try to differentiate their mortality based on their employment and on socioeconomic factor instead of only public and private," she said. "We think it could result in some challenges for employers that have all that kind of different employment in their unique pension plan. But it also makes sense," she said.
Dupont agrees that the new categories are an improvement, particularly as the old public-private split created awkward conversations. For example, a financial services firm technically fell under the private sector table, even though its employees' longevity profile tracked much closer to public sector workers.
Telling a sponsor from the private sector that the public sector table was more appropriate for their group never landed well. While the old system had simplicity on its side, sector determined the table, precision suffered, he noted.
"With the new names for the table, it will be easier to select the appropriate table for a particular pension plan, based on their socioeconomic factor," Dupont said. “Using the new analysis and updating them regularly also allows the proper funding of pension plans in advance, so we don't have a bad surprise in 10 or 20 years.”
Notably, sponsors who’ve already purchased annuities from insurers are insulated from the new tables entirely. For those still carrying longevity risk on their books, annuity purchases remain the most common hedge, noted Audet.
"Longevity swap is another solution. It's maybe something we see less in Canada. It's more popular in the UK. But we have seen some deals with longevity swap," she said, emphasizing however, beyond those two tools, the options for offloading longevity exposure are limited.
According to Dupont, the widely cited two to three per cent increase is based on a fairly typical pension plan, with active members averaging about 45 and retirees around 65. Once a plan moves away from that profile, either toward a much younger membership, a much older retiree group, or both, the effect can shift.
He also underscored that plan design still matters. Indexation pushes more value into payments due decades from now, which can lift the impact beyond the headline range.
Meanwhile, Audet underscored how the estimate is only a rough guide, noting the discount rate can also materially change the outcome, with a plan using a 6 per cent rate producing a different result from one using a 4 per cent rate. She believes sponsors want a usable approximation, but the actual effect depends on a range of demographic and actuarial assumptions.
For sponsors with December 31, 2025 funding valuations, the timing is tight but workable, as Dupont noted while the report only landed weeks ago, most legislative filing deadlines fall in September, leaving room to incorporate the updated tables.
He said adoption makes sense in most cases but not all. For example, if a plan already carries enough margin in its assumptions that reflecting the new mortality data would have no practical effect on funding, waiting for the next valuation cycle is defensible.
Audet added that the CIA is expected to publish further guidance in the coming weeks, which may give some actuaries reason to pause.
“It’ll be difficult to ignore this information but it should be at least considered, and then the actuary will determine their best estimate assumption if the table is appropriate to use or not," she said.
