The customer did not choose you. That is a customer the system delivered to you.
There is a number that does not appear on any renewal campaign brief, any retention dashboard, or any product roadmap currently sitting in a Canadian bank's Q3 planning cycle. It is 179.6. That is the household debt-to-income ratio for Canada as of Q1 2026 — meaning that for every dollar a Canadian household earns after tax, it carries $1.79 in debt. It has been climbing, with two brief interruptions, since 1990. It has never meaningfully come down.
The mortgage renewal wave everyone is calling the great opportunity of 2025 and 2026 is real. Roughly $900 billion in Canadian mortgages are due for renewal spanning late 2024 through end of 2026 — encompassing approximately 60% of all outstanding residential mortgage debt in the country, per TD Economics. Of that, $300 to $350 billion falls in 2026 alone, at an average mortgage balance of $537,000 to $560,000 per household, repricing at an average increase of $622 per month. That is $7,464 in added annual cost, per household, before anything else in their life has moved. Every product team, every sales floor, every retention strategist in the country is working on this. The question being asked, in every one of those rooms, is: how do we keep our customers, and how do we take someone else's?
That is the right commercial instinct. It is being applied to the wrong customer model.
The lever and what it does
The Bank of Canada and the US Federal Reserve have moved interest rates in almost perfect lockstep for thirty-six years. The same cuts, the same hikes, separated by weeks rather than months, responding to the same global cycles. If you plot both rate lines on a single chart, they are nearly indistinguishable.
The household debt lines beneath them are not.
American household debt-to-income peaked at 128% in 2007, then fell — hard, fast, and permanently — to roughly 81% by 2025, where it has stayed. Canadian household debt-to-income peaked at 184.5% in 2022, dipped slightly, and sits at 179.6% today. The same rate signal, received by two different populations, produced two completely different balance sheet outcomes over the same period.
The reason is not behaviour. It is architecture.
The clearing mechanism Canada does not have
When an American homeowner in 2009 owed $400,000 on a house worth $250,000, the system gave them an exit. Hand back the keys. Declare bankruptcy. The debt discharged. The household lost the asset, but it shed the liability. Millions of them did exactly this between 2008 and 2012. It was brutal. It was also structural: the loss moved from the household to the institution to the market, cleared, and stopped compounding. By 2014, the American household balance sheet was genuinely lighter.
In Canada, that exit does not exist in the same form. Mortgages are predominantly recourse — the bank can pursue you beyond the asset. CMHC insurance meant lenders were indemnified against default, removing any incentive to negotiate a writedown. The social and practical cost of bankruptcy in a smaller, relationship-based credit market is higher. So Canadian households who were underwater after 2008 mostly stayed in their houses and kept paying. The debt did not discharge. It rolled into refinances as rates fell, transferred into new purchase transactions as people moved, and accumulated quietly through every subsequent easing cycle.
The result is a household balance sheet that has absorbed thirty-five years of rate-driven borrowing incentives with no systemic clearing event. Not once. The 179.6% figure is not the product of recklessness. It is the product of a system that was designed to hold debt in place.
What this means for the renewal customer
The product and sales conversation about the great renewal is, almost universally, a conversation about rate and friction. What rate do we offer? How easy is the switch? What bundle do we attach to keep the customer? How do we make our digital experience faster than the competitor's?
These are real levers. They are also levers being pulled on a customer the model does not fully see.
The renewal customer arriving at a Canadian bank between now and the end of 2026 is not the same customer who signed the original mortgage in 2020 or 2021. That customer signed at a rate that made the payment feel manageable — sometimes barely manageable — against an income and a cost-of-living that have both changed materially since. Grocery bills, insurance premiums, childcare costs, variable-rate debt on credit cards and home equity lines have all moved. The household has been paying into a higher-rate environment since 2022. Some have absorbed it. Some are absorbing it by running down savings, extending amortisations, or carrying balances they did not carry before.
The renewal signature, when it comes, will look like retention. It will register on the dashboard as a win. But a meaningful portion of those signatures will be signed by households renewing not from preference but from necessity — because the Canadian system makes exit costly enough that staying, even on difficult terms, is still the rational choice. That is not a customer who chose you. That is a customer the system delivered to you. The distinction matters for everything that happens next.
How this shows up in the numbers — and what it hides
Canadian bank financial statements show a clean residential mortgage book. Provisions for credit losses on mortgages have been historically low for the entire post-2008 period. Loan-to-value ratios look conservative, typically disclosed in the 50–65% range on uninsured books. Net interest margin has run structurally higher than comparable US institutions, partly because recourse lending and switching friction mean the bank earns its spread for longer on the same relationship.
These numbers are accurate. They are also working from a history that has never included a genuine stress event in the Canadian housing market. The PCL model sees a clean record and projects forward. The LTV is calculated against current valuations in a market that has appreciated nearly continuously since 1998. The NIM looks like a product quality story; it is also partly a captive-borrower story.
What the statements do not show is where the risk actually sits. A large portion of the insured mortgage book transfers credit risk to CMHC — ultimately the federal government, ultimately the taxpayer. OSFI's domestic stability buffer exists precisely because the regulator understands that the system is holding risk that the income statement does not price. The buffer is the regulator's acknowledgement, written in capital requirements, that the PCL model is unproven at the scale of a genuine correction.
The risk has not been eliminated. It has been distributed to places that do not show up in a quarterly earnings release.
What I would do
This is the part the renewal brief is missing — not the rate, not the friction reduction, not the bundle. The customer insight.
Redefine the renewal segment. The standard segmentation for a renewal campaign is product-based: fixed versus variable, high-balance versus standard, broker-originated versus direct. None of these tell you what you actually need to know, which is: which customers are renewing from strength and which are renewing from constraint? These two populations need completely different offers, different conversations, and different success metrics. A customer renewing from strength is a cross-sell opportunity — wealth products, investment accounts, life stage planning. A customer renewing from constraint needs a conversation about amortisation extension, payment flexibility, and financial planning support before they need it, not after they miss a payment. Conflating them into a single retention campaign is how you optimise the wrong number.
Change the retention metric. Retention rate at renewal is the wrong measure when exit is structurally constrained. A 92% retention rate in the Canadian mortgage market is not the same signal as a 92% retention rate in a market with genuine switching optionality. The metric that matters is retention quality: what is the debt service coverage on the renewed book? What proportion of renewed customers extended their amortisation to manage the payment? What is the six-month delinquency rate on the 2025 renewal cohort versus the 2019 cohort? These numbers tell you whether you retained a customer or inherited a problem with a signature on it.
Design for the conversation, not just the transaction. The digital renewal journey every major bank is investing in right now is optimised for speed — fewest clicks, fastest close, least friction. This is correct for the customer renewing from strength who has already decided and just wants it done. It is exactly wrong for the customer renewing from constraint, who needs a human being, a real options conversation, and a reason to believe the institution sees them as something other than a balance on a book. Routing logic that identifies constrained renewers early and moves them to a different channel — proactive outreach, financial health conversation, a genuine product review — is the campaign design that builds long-term book quality, not just this quarter's retention number.
Build the growth strategy on the US comparison, not the domestic one. Every Canadian bank is benchmarking renewal performance against its own prior cycles and against domestic competitors. This is a closed loop — everyone is working from the same uncleared history, the same PCL models, the same LTV disclosures. The more instructive benchmark is the American household in 2013: what did recovery look like after a genuine clearing event? What products, what channels, what customer relationships proved durable when the balance sheet was actually tested? The institutions that understand this comparison are building for a Canadian housing cycle that has not happened yet. The ones that do not are optimising for a cycle that has not ended.
Make the channel economics reflect the actual cost. The broker channel originates a significant proportion of Canadian mortgage volume, and broker-originated mortgages renew at lower rates than direct relationships — the customer's loyalty is to the broker, not the bank. At renewal, the acquisition cost of retaining a broker-originated mortgage is higher, the relationship depth is lower, and the cross-sell rate is structurally limited. In a renewal wave of this scale, the channel economics conversation is not just about margin — it is about which part of the renewed book actually belongs to the bank in any meaningful sense, and which part will walk to whoever the broker recommends next time.
The three questions worth asking before the next planning cycle
Whether the household can absorb the renewed payment is a credit question. Whether the household will absorb it quietly, carry the strain invisibly, and renew again in five years is a customer behaviour question. These are not the same question, and most renewal strategy conflates them.
The great renewal is a genuine commercial opportunity. It is also the first real test of what thirty-five years of accumulated, uncleared household debt looks like when it reprices. The institutions that hold both of those things in mind simultaneously — the opportunity and the mechanism underneath it — will build something that lasts past this cycle.
The ones that treat it purely as a retention campaign will hit their numbers this year.