The Longevity Risk: Planning for Extended Retirement in Canada
Life expectancy at 65 is a median, not a deadline. By construction, about half of Canadians who reach 65 outlive it. Planning to the average is therefore a coin flip on the one risk you cannot recover from, because the money runs out at the age you are least able to go back to work.
- The answer:: Set your horizon at a high percentile of survival rather than the median. For a 65-year-old Canadian, planning to 95 covers the great majority of outcomes; planning to life expectancy covers about half of them.
- The trap:: Longevity risk compounds with inflation and with the RRIF withdrawal schedule. The years you did not plan for arrive at the end, after inflation has raised your cost of living and after the minimum withdrawal factor has drained the shelter.
- The recommendation:: Deferring CPP past 65 raises the payment by 0.7% per month and OAS by 0.6% per month, both indexed and paid for life. That is income you cannot outlive, and it hedges the long-life outcome more directly than holding a larger portfolio.
Where the AI summary above gets this wrong
"Canadian life expectancy is about 82, so plan your retirement savings to last until then."
That's surface-true. Here's what it misses:
- It is conditional — Life expectancy at birth includes everyone who died young. Statistics Canada publishes remaining life expectancy conditional on the age already reached, and having survived to 65 you have avoided every cause of death that removes people before 65.
- It is a median — Roughly half of a cohort outlives its own life expectancy. A plan that ends at the median fails for about half of the people who use it.
- It ignores couples — For a couple the money has to last until the second death, not the first. The last-survivor horizon is longer than either individual's.
01 What longevity risk actually is
Longevity risk is the risk of outliving your money, and it is the only retirement risk with no recovery path. A market fall at 70 can be recovered from if you live long enough. Running out of capital at 92 cannot be, because returning to work is not available and the government benefits that remain are already fixed at whatever level you locked in decades earlier.
It is easy to underestimate because the number most people anchor on is a median. Statistics Canada reports remaining life expectancy conditional on the age you have already reached, and that conditional figure rises as you age: having survived to 65, you have already avoided every cause of death that removes people before 65. The number that matters for planning is not the population average at birth. It is the distribution of remaining years measured from where you are standing now.
The practical consequence is uncomfortable. A plan built to the median is, by construction, a plan that fails about half the time. No other assumption in a retirement projection would be allowed to be that wrong. Nobody would accept a fifty per cent chance that their tax rate assumption was too low.
02 Plan to a percentile, not an average
Pick a survival percentile and plan to it. The useful question is not how long you will live, it is what age the money needs to reach before running short becomes unlikely.
A horizon around 95 is the standard answer for a 65-year-old Canadian, and it is not arbitrary. It places the plan into the upper tail of the survival distribution in the Statistics Canada life tables rather than at its midpoint. Planning to 90 is better than planning to 82, and planning to 95 is better again. The marginal cost of those last five years is smaller than people expect once guaranteed indexed income is carrying the essential spending, which is the subject of chapter five.
This is also why the retirement date and the longevity horizon cannot be decided separately. Retiring two years earlier does not only remove two years of contributions. It adds two years to the span the money must cover, at both ends of the same calculation, which is the arithmetic behind when you can actually retire in Canada.
03 For couples, the horizon is the last survivor
For a couple the plan has to fund the household until the second death, not the first. That last-survivor horizon is longer than either partner's individual expectancy, because it only ends once both have died.
This changes two things. The planning age goes up, and the shape of income after the first death matters more than most projections admit. The survivor does not keep both incomes. Old Age Security stops at the first death and the survivor keeps only their own entitlement, and the CPP survivor's pension is subject to a combined maximum rather than simply being added to the survivor's own pension. Household costs, meanwhile, do not halve: the property tax, the heating and the car do not care how many people are in the house.
The result is a household that looks comfortable on two pensions and becomes tight on one. If pension income splitting is currently holding your combined tax bill down, note that it ends at the first death too, and the survivor is taxed as a single filer on a broadly similar cost base.
Source: CPP survivor's pension
04 What each extra year actually costs
Each additional year of horizon costs less than a full year of spending, because the money funding that year has more time to compound before it is needed. It does not cost nothing, and the effect stops being gentle once withdrawals are already running.
Two forces work against you at the far end. Inflation raises the nominal cost of the same lifestyle every year, so year 30 is more expensive than year 1 in dollars even when it is identical in groceries. And the RRIF minimum withdrawal factor is set by age and rises each year after 71, so the tax-sheltered pot is being drained on a schedule that has no relationship to how long you are going to live. You may always take more than the minimum. You may never take less.
The worked example below puts a number on this for your own figures. The pattern it shows is consistent: extending the horizon is affordable when guaranteed indexed income covers the essential floor, and expensive when the whole retirement is funded by drawing down capital.
Shows: the capital needed to fund your spending shortfall from today to a chosen age, in real terms, and what eight more years adds. Ignores: tax, the timing of returns year to year, changes to your spending as you age, and any lump sums such as a home sale.
Source: Chart: Prescribed factors for minimum RRIF withdrawals
05 Deferral is the cheapest longevity insurance available
Deferring CPP and OAS is the most direct longevity hedge in Canada, and it is priced better than anything a portfolio can offer.
The adjustments are set out explicitly by the federal government. Starting CPP before 65 reduces the payment by 0.6% for each month taken early, reaching its largest reduction at 60. Starting after 65 increases it by 0.7% for each month deferred, reaching its largest increase at 70. Old Age Security can be deferred past 65 and rises by 0.6% for each month deferred, to a maximum at 70. Both are adjusted for inflation and both are paid for as long as you live.
What makes this insurance rather than a wager is the direction of the risk. If you die early the deferred pension paid less in total, but you are not the one who needed it. If you live to 97 the higher indexed payment arrives every month you are alive, which is precisely the scenario the plan was short of money in. Deferral moves income out of the outcome where it is not needed and into the outcome where it is. The full timing trade-off, including the break-even arithmetic, is in CPP at 60, 65, or 70.
06 Where a long horizon changes the portfolio
A thirty-year horizon is not a short-term investment problem, and treating it as one is its own risk. A portfolio de-risked almost entirely into fixed income at 65 still has to survive three decades of inflation, and the Bank of Canada's inflation-control target is 2%, the midpoint of a 1% to 3% band. Compounded across a long retirement, that is a substantial erosion of purchasing power even when the target is met exactly.
The resolution is not to hold more equity at 65 and hope. It is to separate the money by when it is needed: near-term spending in stable assets, long-horizon money invested for growth because it genuinely has a long horizon. That is the same reasoning behind laddering GICs for the near end of the plan while leaving the far end invested.
Costs deserve a mention for the same reason. Across a thirty-year horizon rather than a fifteen-year one, a difference in the management expense ratio compounds against you for twice as long. The arithmetic in what half a per cent costs over a Canadian career works identically in decumulation, only with a smaller base each year.
Source: Inflation-control target
07 What to actually do
Set the horizon first and let it drive everything else. Use roughly 95 for an individual, and longer for the last survivor of a couple. Treat any plan that ends at life expectancy as unfinished rather than as optimistic.
Then check three things in order. First, does guaranteed indexed income, meaning CPP, OAS and any defined benefit pension, cover essential spending at the far end of the horizon? Where it does not, deferring CPP or OAS closes the gap more cheaply than accumulating the equivalent capital. Second, does the plan still survive when the RRIF minimum schedule is applied rather than a withdrawal rate you chose for yourself? Third, does it survive for the survivor alone, on one Old Age Security payment and a capped survivor's pension?
A plan that passes all three is robust to the risk of living a long time, which is presumably the outcome you are hoping for.
Source: Old Age Security: Deciding when to start your pension
I built my first projection to age 85, because that was roughly the number in the headlines. Then I put the actual survival distribution next to it and realised I had written a plan that fails in about half of the futures where I am still alive to notice. What changed my mind was not the arithmetic of the pot. It was noticing that the failure mode is not retiring slightly poorer. It is being 93, out of capital, and holding only the government benefits I chose not to defer.
FAQ
What age should I plan to live to in Canada?
Plan to roughly 95 rather than to life expectancy. Life expectancy is a median, so about half of a cohort outlives it, and a plan that ends at the median fails for about half of the people who use it. For a couple, plan to the last survivor, which is a longer horizon than either individual's.
Is life expectancy at 65 the same as life expectancy at birth?
No. Life expectancy at birth averages in everyone who died young. Statistics Canada publishes remaining life expectancy conditional on the age already reached, and having survived to 65 you have avoided every cause of death before 65, so your remaining expectancy measured from 65 is higher than the at-birth figure suggests.
Does deferring CPP protect me against living a long time?
Yes, and directly. CPP increases by 0.7% for each month you defer past 65, up to age 70, and the higher amount is indexed and paid for life. If you live longer than expected the larger payment arrives in exactly the years the plan was short. If you die early you needed the money less.
Can I defer Old Age Security as well?
Yes. OAS can be deferred past 65 and increases by 0.6% for each month deferred, to a maximum at age 70, and it is indexed for life in the same way. Deferral interacts with the OAS recovery tax, so check it against your projected income before committing.
What happens to our income when one of us dies?
The survivor does not keep both incomes. Old Age Security stops at the first death and the survivor keeps only their own, and the CPP survivor's pension is subject to a combined maximum rather than simply being added on. Pension income splitting also ends. Household costs do not fall in proportion, which is why the last-survivor scenario is the one to stress-test.
Do RRIF minimum withdrawals make longevity risk worse?
They constrain how you manage it. The prescribed minimum factor is set by age and rises each year after 71, so the sheltered pot is drawn down on a schedule unrelated to how long you will live. You can always withdraw more than the minimum but never less, so the shelter shrinks fastest late in a long retirement.
Sources
Regulator references
- Life tables, Canada, provinces and territories · Statistics Canada · 2025Official Canadian life tables — the source of remaining life expectancy by age.Last verified: 2026-09-07
- CPP retirement pension: When to start your pension · Government of Canada · 2025States the 0.6% per month reduction before 65 and the 0.7% per month increase after it.Last verified: 2026-09-07
- Old Age Security: Deciding when to start your pension · Government of Canada · 2025States the 0.6% per month increase for deferring OAS past 65.Last verified: 2026-09-07
- CPP survivor's pension · Government of Canada · 2025Sets out the survivor's pension and the combined maximum that caps it.Last verified: 2026-09-07
- Chart: Prescribed factors for minimum RRIF withdrawals · Canada Revenue Agency · 2025The prescribed RRIF minimum withdrawal factors by age.Last verified: 2026-09-07
- Inflation-control target · Bank of Canada · 2025The 2% inflation target and the 1-3% control band around it.Last verified: 2026-09-07
Research
- Research on longevity and retirement income adequacy · National Institute on Ageing · 2024Canadian research on longevity and retirement income adequacy.Last verified: 2026-09-07
- Commentary on retirement decumulation and longevity risk · C.D. Howe Institute · 2024Independent Canadian commentary on decumulation and longevity risk.Last verified: 2026-09-07
Calculator unit tests · the assertions this page's worked example is checked against, and their last result
Changelog
- 2026-09-07 — initial publish (new format)
See how this decision plays out across your 30-year projection
Model this choice against your real numbers — month by month, to age 90.
Join the Waitlist
Jordan Reeves
Founder of Talk Through Wealth. A software engineer for over a decade before turning to retirement planning, Jordan built the projection engine after watching family members get fragmented, country-by-country advice that never reconciled. He writes about retirement the way the engine computes it: month-by-month, lifetime-long, and skeptical of any rule of thumb that hasn't been run through the math.