Receipt Housing · Analysis
Mid-credit borrowers are entering serious delinquency faster than subprime
According to Globe and Mail reporting based on Equifax Canada data not publicly released in full, near-prime mortgage delinquency rates (credit scores 621–680) rose 31% from Q4 2024 to Q4 2025 — outpacing the weakest subprime borrowers (23%) and those in the 521–580 range (28%). In the five named high-cost markets, near-prime delinquencies rose 55.6%. Many of these borrowers would have passed the federal stress test at origination. The qualifying-rate overlay tested whether borrowers could absorb a rate increase. It did not test whether they could absorb a rate increase while grocery prices, property taxes, and insurance costs rose simultaneously. The middle is falling behind — and it's falling behind faster than the bottom.
That's the receipt. You're caught up.
You now know what the data shows. Everything below is the full sourced analysis.
Full analysis
Continue reading
Key Facts
Verified and sourced to primary documents
Context
What this analysis might be missing
Interpretation
Our analysis — labeled. Includes the counter-argument
Falsifiers
What evidence would change our view
Near-prime delinquency increase
+31%
Same cohort, 5 priciest markets
+55.6%
National near-prime delinquency
0.44%
Large-mortgage stress threshold
$800K+
Key Facts — Verified

According to Globe and Mail reporting based on non-public Equifax Canada data, near-prime mortgage borrowers (credit scores 621–680) saw their 90+ day delinquency rate increase by 31% from Q4 2024 to Q4 2025. This data has not been independently verified by The Receipts from a public Equifax release. [1]

The national mortgage delinquency rate was 0.26% as of December 2025, according to the Canadian Bankers Association. [2]

OSFI's mortgage stress test requires borrowers to qualify at the greater of their contract rate plus 2 percentage points or 5.25%. This rule has been in effect since June 1, 2021. [3] [4]

The Bank of Canada's Staff Analytical Note 2023-19 projected that the median monthly mortgage payment would rise from approximately $1,200 to $1,600 by end-2027 as borrowers renewed at higher rates. [5]

OSFI's 2025–2026 Annual Risk Outlook stated that delinquencies were expected to continue rising as mortgage holders renewed at higher rates, with borrowers who originated at historic lows before March 2022 most exposed. [6]

The household debt service ratio was 14.57% in Q4 2025, according to Statistics Canada — essentially flat across the year. [7]


The Inversion

The conventional understanding of mortgage stress follows a simple hierarchy: the weakest borrowers fall behind first, then the pressure migrates upward if conditions worsen. Subprime at the bottom, prime at the top, and near-prime in between as a buffer zone that absorbs some turbulence but generally holds.

Globe and Mail reporting on March 16, 2026, based on Equifax Canada data not publicly released in full, disrupts that model. Near-prime borrowers — those with credit scores between 621 and 680 — saw their 90+ day delinquency rate increase by 31% over the year ending Q4 2025. That pace of acceleration was faster than the weakest subprime borrowers (credit scores 320–520), whose delinquency rate rose 23%, and faster than the 521–580 cohort, which rose 28%. [1]

The absolute numbers are still low. Nationally, the near-prime delinquency rate was 0.44% in Q4 2025. For the most creditworthy borrowers, it was 0.01%. Across all borrowers, 0.26%. [1] [2]

But the rate of change is where the finding sits. The middle tier isn't just experiencing stress — it's experiencing stress at a pace that exceeds the tier below it. That inversion is the signal this article tracks.

In plain English

People with decent credit — not great, not terrible — are falling behind on their mortgages faster than people with poor credit. The overall numbers are still small. But the speed of deterioration is faster in the middle than at the bottom, and that's the opposite of what the system was designed to expect.


What the Stress Test Assumed

The federal mortgage stress test — Guideline B-20 for uninsured mortgages, with a parallel rule for insured mortgages — requires borrowers to demonstrate they can service their mortgage at a qualifying rate above their actual contract rate. Since June 1, 2021, that qualifying rate has been the greater of the borrower's contract rate plus 200 basis points or a floor of 5.25%. [3] [4]

During the pandemic lending window — roughly 2020 through early 2022 — five-year fixed mortgage rates fell as low as approximately 1.7%. Borrowers at those rates qualified under the stress test at the 5.25% floor, since their contract rate plus 200 basis points fell below the floor. [4]

Those borrowers are now renewing. Renewal rates remain materially above pandemic-era lows, even after recent Bank of Canada easing. That means the actual renewal rate for many pandemic-era originations is below the rate they were stress-tested at. On paper, the qualifying-rate overlay worked: many of these borrowers were qualified for a payment level higher than what they're now actually facing at renewal. [1]

The Bank of Canada's 2025 Financial Stability Report confirmed this arithmetic. It noted that about 60% of outstanding mortgages would renew in 2025 or 2026, and that more than 90% of five-year fixed borrowers would face payment increases smaller than what they were stress-tested for. [8]

The qualifying-rate overlay, in other words, appears to have reduced rate-risk vulnerability for the cohort it was designed to protect. The question is what it was not designed to absorb.

In plain English

Most people who took out mortgages during the pandemic were tested against a rate of about 5.25%. They're now renewing at rates that are still lower than what they qualified for. So on the mortgage math alone, they should be fine. The problem is that the mortgage payment isn't the only bill that went up.


The Gap the Test Didn't Model

The mortgage stress test is not a full household-resilience model. It is a qualifying-rate overlay applied within broader underwriting rules that include debt-service ratios, property taxes, heating costs, and in some cases condominium fees. The qualifying-rate component tests whether borrowers can carry the mortgage at a higher rate. It does not continuously re-underwrite borrowers against later increases in the wider cost of living — grocery inflation, insurance premium growth, utility rate changes, and childcare costs that arrive after origination. [3]

The Bank of Canada's Staff Analytical Note 2023-19, published in December 2023, projected that the median monthly mortgage payment would rise from approximately $1,200 to $1,600 by end-2027 for borrowers renewing at higher rates — a roughly 33% increase in shelter costs alone. [5]

But that 33% increase arrived alongside a period in which the consumer price index for shelter in Canada's major metropolitan areas rose significantly. Statistics Canada data shows sustained shelter-cost inflation in the Toronto and Vancouver CMAs across 2023–2025, compounding the payment shock for borrowers in those markets. [9]

The household debt service ratio — the share of disposable income devoted to debt payments — was 14.57% nationally in Q4 2025, according to Statistics Canada. That was essentially flat across 2025, ranging from 14.40% to 14.64% quarter over quarter. [7] [10] [11] [12]

The national stability of the debt service ratio masks geographic concentration. Equifax's data — and CMHC's independent reporting — show that the stress is not evenly distributed.


Where the Stress Concentrates

The Equifax data reported by the Globe identifies five markets where the deterioration is most acute: Toronto, Vancouver, Brampton, Markham, and Oshawa — all high-cost markets in the GTA corridor or Metro Vancouver. In these markets, the delinquency rate has increased for every type of mortgage borrower. [1]

Near-prime delinquencies in those five markets rose 55.6% year-over-year. Subprime delinquencies rose 50%. Prime borrowers increased 32.5%. Even borrowers with the highest credit scores saw their delinquency rate increase by 4.5%. [1]

This geographic concentration aligns with independent data. CMHC reported in November 2025 that the national mortgage delinquency rate for Q2 2025 was 0.22%, but Ontario was 0.23% and the Toronto CMA specifically was 0.24%. [13]

The Canadian Bankers Association's December 2025 arrears data showed Ontario at 0.27% — the highest of any major province except Saskatchewan at 0.52%. [2]

Equifax also reported that borrowers with the largest mortgages — loans greater than $800,000 — are experiencing the sharpest increases. In Ontario, the delinquency rate on those large loans rose 28% year-over-year. In British Columbia, 26%. [1]

This is consistent with the stress test's arithmetic: the same percentage-point increase in rates produces a larger absolute dollar increase on a larger mortgage. A borrower renewing an $800,000 mortgage from a pandemic-era low to a current market rate faces a monthly payment increase that is, in dollar terms, substantially larger than someone renewing a $400,000 mortgage at the same rate change — even though both passed the same qualifying-rate test.

In plain English

The worst-hit markets are the most expensive ones — Toronto, Vancouver, and three GTA suburbs. In those cities, every borrower tier is getting worse, but the mid-credit borrowers are deteriorating fastest. The bigger the mortgage, the bigger the payment shock in raw dollars, even if the rate increase is the same.


How They Got Here

This article is Part 3 of The Housing Trap, a series that follows the consequences of the decisions documented in The Housing Receipt. The connection between the earlier parts and this data is structural, not incidental.

Part 1 documented how the federal government's Home Buyers' Plan channelled retirement savings into the housing market during the 2020–2022 bubble window, then expanded the withdrawal limit from $35,000 to $60,000 during the correction. Those buyers entered the market with depleted retirement savings and mortgages originated at pandemic-era rates. [14]

Part 2 documented how parents pledged their borrowing capacity through co-signing arrangements that carry joint-and-several liability, and now face co-signer exit failure: the mortgage renews, but the parent cannot be removed because the child cannot independently qualify at current rates. [15]

The borrowers appearing in the Equifax near-prime delinquency data overlap significantly with the cohort those instruments produced, though the data does not disaggregate at that level. OSFI's 2025–2026 Annual Risk Outlook explicitly identified "mortgages originated at historic lows before March 2022" as the most exposed group. [6] The Bank of Canada's 2024 Financial Stability Report stated that borrowers who took mortgages in 2021 and early 2022 at very low rates and near-peak house prices would "generally see larger increases in payments at renewal." [16]

This article does not claim that every near-prime borrower now in delinquency used the HBP or had a parental co-signer. What the data does show is that the cohort most exposed to delinquency acceleration — mid-credit borrowers who originated during the pandemic at the lowest rates in Canadian history — is the same cohort that the government's fiscal instruments and co-signing norms were designed to serve. The overlap is a structural observation, not a proven compositional claim.


The Regulators Flagged It

This pattern was not invisible to regulators. OSFI explicitly warned about it — repeatedly.

In March 2024, OSFI issued guidance titled "Reinforcing residential mortgage risk management practices," stating that many borrowers would face "payment shock at renewal or sooner, especially variable-rate mortgages with fixed payments." [17]

OSFI's 2024–2025 Annual Risk Outlook, published May 2024, stated that 76% of mortgages outstanding as of February 2024 would renew by end-2026, and that those homeowners "could face payment shock, most significantly those who took mortgages in 2020 to 2022." [18]

The semi-annual update in October 2025 reiterated that "a subset of borrowers will still face significant payment increases" and identified mortgages originated before March 2022 as the most vulnerable. [19]

These are not retrospective assessments. OSFI was describing the risk in real time, using language that directly maps to the cohort now appearing in the Equifax delinquency data. The regulator that administers the stress test was simultaneously flagging payment-shock risk that persisted even for borrowers originated within the stress-tested system.

Context — What Both Sides Omit

What critics of the stress test omit: The test was never designed to model total household cost-of-living exposure. It is a mortgage-specific qualification tool, not a comprehensive financial resilience assessment. Asking it to account for grocery inflation and insurance premiums would require a different instrument entirely. More than 90% of five-year fixed borrowers are renewing below their stress-tested rate, which means the test prevented a substantially worse outcome.

What defenders of the current framework omit: The stress test's design assumption — that rate risk is the primary threat to mortgage serviceability — was tested and found incomplete. The instrument succeeded at its stated objective while failing to prevent the outcome it was meant to guard against. OSFI flagged the payment-shock risk that underlies this gap. The 90% figure describes borrowers who are renewing below their tested rate, but says nothing about whether those borrowers can absorb the renewal payment alongside everything else that increased during the same period.

Interpretation — Labeled

In our assessment, the Equifax data reported by the Globe and Mail points to a gap in Canada's mortgage protection framework. The qualifying-rate overlay correctly reduced rate-risk vulnerability. It did not model the environment in which that rate variable would operate. Borrowers who qualified at 5.25% and are renewing at materially lower rates are technically within the test's parameters — and are nonetheless falling behind because the qualifying-rate component assumed the primary threat was a rate increase, not a rate increase arriving alongside sustained cost-of-living pressure.

The near-prime acceleration is particularly significant because these borrowers were not flagged as high-risk at origination. They had established credit histories, stable payment records, and sufficient income to pass the qualifying threshold. Their delinquencies are not the product of financial recklessness — they are, in our assessment, the product of a protection mechanism that defined the primary risk too narrowly.

Combined with the findings from The HBP Retirement Trap and The Co-Sign Trap, a pattern emerges: the government created fiscal instruments that funnelled families into the housing market, regulators approved them through a qualifying-rate overlay that tested one dimension of resilience, and both the instruments and the test are now associated with the outcomes they were designed to prevent. Whether the instruments caused those outcomes or merely coincided with them is a question the current data cannot fully resolve.

Counter-interpretation: The stress test prevented a substantially worse outcome. Without B-20, borrowers would have qualified at their contract rates — as low as 1.7% — with no buffer at all. The 0.44% near-prime delinquency rate, while rising, is still historically low and well within the loss-absorption capacity of Canada's major lenders. The acceleration in delinquency rates may reflect a temporary adjustment period as pandemic-era mortgages reset, not a structural failure of the qualification framework. The Bank of Canada's own analysis shows that the stress test improved borrower resilience and that more than 90% of five-year fixed borrowers are renewing below their stress-tested rate. If delinquency rates plateau or decline after the renewal wave passes, the qualifying-rate overlay will have done its job — imperfectly, but adequately.

What Would Change This Assessment
  • If the near-prime delinquency rate stabilizes or declines in Q1–Q2 2026 despite the ongoing renewal wave, the "spreading stress" thesis weakens. The acceleration may represent a one-time adjustment rather than a structural failure of the qualification framework.
  • If Bank of Canada rate cuts accelerate and renewal-rate shock narrows significantly (e.g., five-year fixed rates fall below 3.0%), the cost-of-living compression argument loses its mortgage component. The gap the stress test didn't model would close from the rate side.
  • If the absolute delinquency level remains below 1.0% nationally for near-prime borrowers through the end of the renewal wave (approximately 2027), the systemic risk interpretation would be unsupported. The velocity of change would have mattered less than the level.
  • If OSFI or the federal government introduces a cost-of-living-adjusted qualification test or other supplementary resilience measure, that would confirm the gap identified here but also change the forward-looking assessment of the framework.
  • If data emerges showing that near-prime delinquencies are concentrated in borrowers who did not use the HBP or parental co-signing, the series connection between Parts 1, 2, and 3 would weaken, and the policy-instrument thesis would need to be revised.

Sources (18)

  1. Clare O'Hara, "Homeowners with stronger credit scores are increasingly defaulting on their mortgage payments." Globe and Mail, March 16, 2026. Cites Equifax Canada data on mortgage delinquency by credit tier. Note: the underlying Equifax dataset by credit tier has not been publicly released; this source represents reporting based on a media briefing or non-public data product. theglobeandmail.com
  2. Canadian Bankers Association, "Number of Residential Mortgages in Arrears," month ended December 31, 2025. Province-by-province 90+ day arrears counts and rates. cba.ca (PDF)
  3. OSFI, "Minimum qualifying rate for uninsured mortgages." Current page, accessed March 2026. Qualifying rate: greater of contract rate + 2% or 5.25%. osfi-bsif.gc.ca
  4. OSFI, "Amendments to the minimum qualifying rate for uninsured mortgages — Letter (2021)." May 20, 2021. Effective June 1, 2021. osfi-bsif.gc.ca
  5. Bank of Canada, "The impact of higher interest rates on mortgage payments." Staff Analytical Note 2023-19, December 2023. Median payment projections: ~$1,200 to ~$1,600 by end-2027. bankofcanada.ca
  6. OSFI, "Annual Risk Outlook — Fiscal Year 2025–2026." March 13, 2025. Identifies mortgages originated before March 2022 as most exposed. osfi-bsif.gc.ca
  7. Statistics Canada, "National balance sheet and financial flow accounts, fourth quarter 2025." March 16, 2026. Household debt service ratio: 14.57%. statcan.gc.ca (PDF)
  8. Bank of Canada, "Financial Stability Report — 2025." May 2025. 60% of mortgages renewing in 2025–2026; 90%+ of five-year fixed borrowers facing increases smaller than stress-tested rate. bankofcanada.ca
  9. Statistics Canada, "Table 18-10-0004-12: Consumer Price Index by product group, monthly, percentage change." Shelter component by CMA. statcan.gc.ca
  10. Statistics Canada, "National balance sheet and financial flow accounts, third quarter 2025." December 11, 2025. DSR: 14.64%. statcan.gc.ca
  11. Statistics Canada, "National balance sheet and financial flow accounts, second quarter 2025." September 11, 2025. DSR: 14.41%. statcan.gc.ca
  12. Statistics Canada, "National balance sheet and financial flow accounts, first quarter 2025." June 12, 2025. DSR: 14.40%. statcan.gc.ca
  13. CMHC, "National delinquency rate drops but continues to rise in Ontario and BC." November 19, 2025. National: 0.22%, Ontario: 0.23%, Toronto CMA: 0.24%. cmhc-schl.gc.ca
  14. Department of Finance Canada, "Statement by the Deputy Prime Minister and Minister of Finance on the Canadian housing market." May 20, 2021. Insured mortgage MQR aligned to greater of contract + 2% or 5.25%. canada.ca
  15. Bank of Canada, "Financial Stability Report — 2024." May 3, 2024. Identifies 2021–early 2022 originations as facing largest payment increases at renewal. bankofcanada.ca
  16. OSFI, "Reinforcing residential mortgage risk management practices." March 11, 2024. Payment shock warning for variable-rate borrowers. osfi-bsif.gc.ca
  17. OSFI, "Annual Risk Outlook — Fiscal Year 2024–2025." May 22, 2024. 76% of outstanding mortgages renewing by end-2026. osfi-bsif.gc.ca
  18. OSFI, "Annual Risk Outlook — Semi-annual update — Fiscal Year 2025–2026." October 9, 2025. Identifies pre-March 2022 originations as most vulnerable. osfi-bsif.gc.ca
Series context (not counted as evidence sources)

The HBP Retirement Trap — Documents HBP withdrawal limit expansion and the fiscal pipeline from retirement savings into housing.

The Co-Sign Trap — Documents co-signer exit failure and joint-and-several liability exposure.

Equifax Canada, "Stable versus Struggling: Canada's Financial Divide Widens." February 25, 2025. Q4 2024 mortgage delinquency data by province. Independent corroboration of delinquency trend direction.

No corrections at time of publication — March 2026.
Reader Prompt

This article relies on Equifax data reported through the Globe and Mail. If you have access to primary Equifax mortgage delinquency data by credit tier, CMHC origination data by loan size, or any other primary-source evidence that supports or contradicts the claims made here, we want to see it. Contact: [email protected]