The 2.25% Hold Buys Time, Not Relief, for Canada's Renewal Wave

What we observe. On September 2, the Bank of Canada held its overnight rate at 2.25%, leaving the benchmark where it has sat long enough that markets have begun treating it as semi-permanent. The problem is that the households most exposed to this rate have not yet felt it. The CPI sub-index for mortgage interest cost stood at 182.1 as of July, a level that reflects the rolling-in of renewals at rates far above what borrowers locked in during 2020 and 2021. Ontario's all-items CPI was 169.9 over the same period, meaning mortgage costs are running well above the general price level and squeezing real purchasing power in Canada's most populous province. The household debt service ratio reached 14.75% as of the first quarter of 2026, and average hourly wages were growing at only 2.8% year-over-year as of July. That gap between what households owe each month and what their paycheques are adding is the core tension. The labour market adds another layer. The national unemployment rate came in at 6.4% in August, with the labour force itself at roughly 22.6 million workers. Ontario carries about 38% of the national population and its unemployment rate feeds directly into that national figure, as does British Columbia's at roughly 13% weight. A 6.4% unemployment rate is not a crisis, but it is high enough that a meaningful share of the households facing mortgage renewals through 2027 will do so with less job security than they had when they originally signed. The non-bank mortgage share has climbed to nearly 24.5%, meaning roughly one-in-four mortgages now sits with lenders whose funding costs and risk appetite differ from the big chartered banks, and who may price renewals more aggressively or have less flexibility to restructure.

Why it matters. The institutional stakes here are straightforward. If the renewal wave passes without a significant rise in arrears, the Bank of Canada's gradual easing cycle will look well-calibrated. If wage growth stays below 3% while debt service ratios hold near 15%, a material share of renewing households will face a genuine cash-flow shock even at 2.25%, because they are rolling off rates that were often below 2%. The non-bank share amplifies this: those lenders are less subject to the regulatory forbearance tools that the Office of the Superintendent of Financial Institutions can deploy at the chartered banks, so stress in that segment could surface faster and with less cushion.

Three relationships are doing most of the work here. First, the national unemployment rate is heavily shaped by Ontario's labour market, which carries roughly 38% of the population and has a measured relationship strength of about 0.80 with the national figure. British Columbia adds another 13% of population weight, with a similarly strong relationship at 0.80. Together, these two provinces can move the national unemployment rate materially without any change in the rest of the country. If either province softens further, the 6.4% national rate will drift higher, and that matters directly for mortgage renewal cash flows: a household that loses income while facing a renewal at current rates is in a qualitatively different position than one that simply faces a higher payment on a stable income. Second, the non-bank mortgage share is mechanically driven by the volume of residential mortgage credit extended by non-bank lenders. As that volume has grown, the share has reached 24.5%. Non-bank lenders typically fund themselves at spreads above the overnight rate, so a hold at 2.25% does not necessarily mean their mortgage pricing holds flat. Third, the participation rate is the dominant driver of the size of the labour force itself, with a near-perfect relationship considerable strength. If participation falls as discouraged workers exit, the unemployment rate can stay flat or even fall while actual employment conditions deteriorate, which would give a misleadingly benign read on household income risk heading into the renewal wave.

Every major economic stress condition in the model is currently dormant: there is no active inflationary backdrop, no credit cycle contraction, no liquidity crisis, no asset bubble, and no deflationary pressure. That dormancy is actually the most important single fact for interpreting the hold. Under an active inflationary backdrop, the relationship between the policy rate and consumer demand would be roughly twice as powerful as in normal conditions. Under an active credit cycle contraction, the feedback from borrower distress to tighter lending standards would also be amplified. Because none of those conditions are active, the transmission from the 2.25% rate to household cash flow is running at baseline strength, not at a crisis multiplier. That is the good news. The bad news is that baseline transmission is still substantial when the debt service ratio is at 14.75% and wage growth is running at 2.8%. The absence of a crisis amplifier does not mean the underlying arithmetic is comfortable.

Only one relationship in this analysis has any track record against real outcomes: the link between total Canadian employment and the national unemployment rate. That relationship has been contradicted once and confirmed zero times in the data history available. What that one contradiction likely reflects is the participation-rate dynamic described above: employment can rise while the unemployment rate also rises if enough previously sidelined workers re-enter the labour force. This is a known feature of the relationship, not a flaw in the logic, but it does mean the model has not yet demonstrated that it calls this correctly in real time. All other relationships described here, including the provincial unemployment weights, the non-bank share mechanics, and the participation-rate-to-labour-force link, have no confirmed or contradicted history in the data. The mechanism rests on sound economic logic and the structural arithmetic of how national aggregates are constructed from provincial components, but it does not yet have a validated track record. Readers should weight the directional argument accordingly.

The counterargument starts with the labour market. A national unemployment rate of 6.4% is elevated but not alarming, and the competing drivers of that rate are genuinely diverse: Ontario, British Columbia, Quebec, and Montreal all contribute meaningfully, and they do not always move together. If Ontario's labour market holds while BC softens, the national rate could stay range-bound even as the renewal wave crests. More importantly, the retail price pressure spread (the gap between nominal and real retail prices) was 3.5 percentage points as of June, which suggests some nominal income support is still flowing through the economy. If wage growth accelerates even modestly toward 3.5% or 4%, the cash-flow squeeze on renewing households becomes much more manageable. The absence of any active stress condition in the broader economic backdrop supports the view that this is a slow grind, not a cliff.

What would falsify this. The thesis would be falsified if, by March 2027, the national mortgage arrears rate remains below its five-year average while the national unemployment rate falls back below 6.0% and average hourly wage growth accelerates above 3.5%. That combination would mean households absorbed the renewal shock without material distress, which would imply the debt service ratio overstated the actual cash-flow risk.

What to watch. The single most important thing to monitor is whether average hourly wage growth crosses 3.5% year-over-year before the bulk of the 2027 renewals price in. If wages accelerate to that level while the overnight rate stays at or below 2.25%, the cash-flow arithmetic shifts from uncomfortable to manageable for most renewing households. If wages stay below 3% and the unemployment rate drifts above 6.8%, the renewal wave will produce a measurable rise in arrears, particularly in the non-bank segment.