42% of Borrowers Now Choose Variable as CMHC Says Renewal Wave Has Peaked

The 42% figure is striking less because of where it sits and more because of how fast it got there. Variable-rate mortgages had been a minority product through most of 2023 and 2024, when the Bank of Canada's rate-tightening cycle left variable rates well above 5-year fixed offers. By February 2026, that pricing inversion had flipped. Variable rates traded at a discount to comparable fixed terms for the first time since 2022, and the term-selection share moved with the math.

The corresponding decline in 5-year fixed share is the other half of the same story. CMHC reports that the traditional 5-year fixed — long the default product for Canadian homeowners — accounted for only about 11% of new extensions at chartered banks. Shorter-term fixed products and variable-rate mortgages absorbed almost all of the volume that used to go to the 5-year. That migration changes the structural sensitivity of Canada's mortgage book: a larger share of the outstanding stock now re-prices on either Bank of Canada moves (variable) or shorter renewal cycles (1-, 2-, and 3-year fixed). The system is more rate-sensitive, in both directions, than it was three years ago.

The rate environment has been moving down in parallel. Average mortgage interest rates declined from roughly 4.8% in January 2025 to about 4.2% in January 2026, modestly reducing the size of the payment increase many renewers face. TD Economics has also reported that the average renewal payment shock is easing — to about 6% in 2026 from 10% in 2025 — reflecting the same cumulative-cut effect from a different angle. The CMHC framing and the bank-economist framing line up.

The term-choice shift is rational. CMHC's own explanation is that lower variable rates and uncertainty about the future rate path made borrowers hesitant to lock into long terms. Two design features of variable-rate products explain why they win in this environment. First, variable rates are priced off prime, which moves with the Bank of Canada overnight rate. Second, variable products carry no penalty exposure to bond-market repricing during the term — fixed-rate mortgages, by contrast, are tied to 5-year Government of Canada bond yields, which can move independently of the policy rate.

The Bank of Canada's recent posture has made both features more attractive. The April 29 rate hold at 2.25% was accompanied by language acknowledging two-sided risk: a US trade restriction shock could push the next move to a cut, while a sustained oil shock could push it to consecutive increases. A homeowner trying to choose between a 1-year fixed at one number and a variable at another is, in effect, taking a position on which branch of that scenario tree plays out. A 5-year fixed, by contrast, asks the borrower to commit to a single rate across both branches and several more years besides. In an environment where the central bank itself is publishing decision trees, the willingness to lock in a long single number has fallen — and the data shows it.

The shorter-term fixed products — 3-year and below — have absorbed much of what would historically have gone to the 5-year. CMHC's data shows borrowers have steadily migrated toward shorter terms over the past several years. The cumulative effect is a mortgage book where more loans renew more often, which makes the next rate cycle, in either direction, transmit faster to household cash flow.

CMHC is direct about the renewal trajectory. The mortgage renewal wave peaked in 2025. The number of borrowers renewing in 2026 is expected to be about 13% lower than the 2025 cohort. That language matters because the previous two years of public commentary treated the renewal cliff as a continuous, accelerating event. The CMHC framing repositions it: the most concentrated cohort has now moved through, and 2026 is the year the volume starts to ease.

The qualifier the report attaches is equally important. While fewer people are renewing in 2026, those who do renew still face significant payment increases at today's higher rates. The average renewal payment shock has eased relative to 2025, but it has not disappeared. The combination of lower volume and similar-per-borrower stress changes the system-level dynamic. Aggregate renewal pressure is dropping. Individual renewal pressure is not.

For homeowners, that distinction is the practical one. A system-level easing does not change what arrives in the renewal package. The lender's offer is governed by the same forces — bond yields, prime rate, the homeowner's loan-to-value position, the discounts available at switch — that drove the 2025 cohort's outcomes. CMHC's framing is helpful for understanding Canada's macro stability. It is not a personal forecast of what any given renewal will look like. Survey data running parallel to this — TD's finding that two-thirds of Canadian homeowners are anxious about mortgage renewal and 56% plan to cut spending — captures the same gap between system-level stability and household-level pressure.

The 90-plus-day delinquency data tells a careful story. The national rate rose to 0.24% in Q4 2025 from 0.21% a year earlier — the highest since early 2021 — but still low in absolute terms. The aggregate looks small. The decomposition is the part that matters. Ontario's delinquency rate rose about 35% year-over-year, with Toronto's 90-plus-day rate climbing from 0.20% to 0.29% — roughly a 45% jump. Other markets with elevated stress include Barrie (0.23% to 0.34%), Windsor (0.15% to 0.22%), Vancouver (0.16% to 0.21%), and Abbotsford-Mission (0.22% to 0.29%). At the same time, several markets — including Québec City and St. John's — saw arrears decline.

CMHC's deputy chief economist Aled ab Iorwerth has been clear that national arrears remain low by historical standards and the system as a whole is stable, while warning that pockets of significant stress are emerging — particularly in Toronto and Vancouver. That framing matters for how to read the table. A 45% year-over-year jump in Toronto arrears sounds like a system-level emergency until you note that the absolute level is still 0.29%. The signal is concentration, not collapse.

The stress test sits behind that distinction. The mortgage stress test, administered by the Office of the Superintendent of Financial Institutions through the minimum qualifying rate, requires uninsured borrowers at federally regulated lenders to qualify at the higher of their contract rate plus 2 percentage points or 5.25%. The intent is to ensure borrowers can keep paying if rates, income, or expenses move against them. CMHC explicitly credits the stress test with helping 2025 renewers absorb higher payments — most managed the jump in monthly payments relatively well, even as rates climbed from pandemic lows.

OSFI has also adjusted the rule at the margins. The stress test is no longer required for straight switches of uninsured mortgages at renewal, as long as the loan amount and amortization don't increase. CMHC reports that uninsured switches rose by about 34% between the second half of 2024 and the second half of 2025 after that change took effect — direct evidence that easing the switch friction changed renewal behaviour. For homeowners shopping a renewal, the practical effect is that moving lenders is now mechanically easier than it was two years ago, at least for borrowers with stable balance sheets. The recent Morningstar DBRS analysis validates OSFI's decision to keep both the stress test and loan-to-income limits in place as a dual safeguard — the existing tools are doing the work the report describes.

Federal policy changes have also shifted the composition of the new lending book. About 54% of mortgages extended to first-time homebuyers by chartered banks in Q4 2025 were insured, up from the mid-40% share typically seen before recent rule changes. Late-2024 federal updates allowed eligible first-time buyers of newly built homes to access 30-year insured mortgages — later expanded to all first-time buyers and all new-build purchases — and raised the maximum insured home price from $1 million to $1.5 million. The effect is to shift more new mortgages into insured products, which moves default risk away from bank balance sheets and onto the federal insurer. For renewal-cohort homeowners, this is background context rather than a direct change to their renewal options. For the system, it changes where risk lives.

Three data threads will tell homeowners whether the patterns CMHC describes are stable or shifting. First, the variable-fixed spread. The 42% variable share is grounded in a specific price differential between variable and 5-year fixed rates. If bond yields move materially — and the 5-year Government of Canada yield closed at 3.22% after the April BoC pause as big-bank fixed mortgage rates spiked — the variable advantage compresses or widens, and term-share data will follow. Second, the next CMHC RMIR release will show whether the Toronto and Vancouver arrears concentrations have stabilized, broadened, or reversed. The lag between renewal payment strain and formal 90-plus-day delinquency is several quarters, so the Q4 2025 reading is the leading edge of the 2025 renewal cohort, not the full story. Third, OSFI's posture on the stress test and switch exemption. Both rules are doing observable work in the data. A change to either would reshape the renewal-market choice set materially.

The Spring 2026 RMIR is best read as a status check, not a forecast. It confirms that Canadian homeowners are voting with their term choice for variable in numbers that would have been unusual three years ago. It says the most concentrated wave of renewals is past. And it shows that the stress in the system is real, but concentrated, with the policy framework around it still functioning. For homeowners with a renewal date on the horizon, the report doesn't answer which product to choose. It does describe the market the choice is being made in.