What The August Consumer Price Index Actually Says To Owners Renewing In The Next Six To Twelve Months
Credit: Shutterstock
OTTAWA — Statistics Canada reported on September 14, 2026 that the Consumer Price Index rose 3.0% year over year in August, matching July's pace. That number will be read as bad news. For a homeowner with a renewal date inside the next year, it is close to the least informative figure in the release.
The reason is composition. A headline built on gasoline is a different economic event from a headline built on rent, wages or services, and it carries different consequences for what borrowing costs do next. In August the gap between those two stories was unusually wide: strip out gasoline and the same index rises 2.4%. Strip out food and energy and it rises 2.1%. The measures the central bank actually steers by sat at or just above its 2% target, exactly where they sat in July, and in one case exactly where they have sat since June.
This piece does three things. It separates the headline from the underlying measures and shows which number the 1.9% belongs to. It decomposes shelter, where owners and renters are currently living in two different inflation environments. And it traces the plumbing that connects an inflation print to a mortgage quote — plumbing that runs through the bond market for fixed rates and through the policy rate for variable ones, which is why the two react to different news on different days.
Stay on top of Canadian homeowner news
The all-items index rose 3.0% year over year and actually fell 0.1% month over month, though on a seasonally adjusted basis it rose 0.2%. Gasoline was the story: prices at the pump were 22.8% above year-ago levels in August, down from a 25.7% annual gain in July, and Statistics Canada attributed the continued elevation to the ongoing conflict in the Middle East. Transportation as a whole rose 7.5%.
The second-largest upward contributor was arithmetic rather than economics. Travel tour prices rose 26.1% year over year against 15.2% in July, and the release explains the jump as partly a base-year effect: Canadian travel to the United States fell sharply in 2025, which depressed airfare and travel tour prices, and that decline has now dropped out of the twelve-month comparison. Nothing new happened to travel prices in August. The comparison point moved.
That distinction matters more than it sounds, because it is the whole reason central banks publish a second set of numbers. The Bank of Canada aims to keep inflation at the 2 per cent midpoint of an inflation-control target range of 1 to 3 per cent, and it monitors a set of core measures that, in the Bank's own words, allow it to "look through" temporary changes in total CPI. A 3.0% headline is not a breach of the target. It is the top of the range, driven by the one input monetary policy cannot influence.
Readers tracking this series will recognize the pattern from a month earlier, when July's print showed the same 3% headline against a 2.2% ex-gasoline rate. August widened that gap rather than closing it.
Here is where most coverage of this release goes wrong, and where the headline on this article originally did too. "Core inflation fell to 1.9%" is a sentence that requires you to pick one of three official measures and then not check its history.
Statistics Canada's table of Bank of Canada core measures shows what actually happened across the past six months:
CPI-trim did not fall in August. It held, for the third month running. CPI-median held at 2.0%. The only measure that moved down was CPI-common, from 2.7% to 2.6%. A three-month plateau at or below target is a more durable signal than any single monthly step, and it is the more useful fact for someone deciding when to lock a rate — but it is not a decline, and reporting it as one overstates what the data did.
The three are not interchangeable. CPI-trim excludes the components whose monthly price changes sit in the tails of the distribution, which is how a 22.8% gasoline move gets removed automatically rather than by judgment. CPI-median takes the price change at the 50th percentile of the basket by weight — the middle item, immune to extremes at either end. CPI-common isolates the price movements that show up across categories at once, which makes it the slowest to respond and the highest of the three right now at 2.6%.
RBC Economics reached the same reading independently on release day, noting that CPI-trim was 1.9% and CPI-median 2%, leaving their average unchanged from the previous month. The bank's economists also reported that inflation breadth narrowed: the share of basket components growing faster than 3% over the past three months slipped to 25.9% from 30.6%, and the share growing faster than 5% fell to 17.7% from 25.4%.
Narrowing breadth is the technical version of the argument this article is making. Fewer categories are rising quickly. The pressure is concentrated, not general.
For the July baseline, the Nova Scotia Department of Finance's parallel analysis confirms the same starting point from a second government source: CPI-trim at 1.9% in July, unchanged from June.
This is consistent with the picture that emerged when business inflation expectations eased to 3.3% and forecasters converged on a hold through 2026.
Here is where most coverage of this release goes wrong, and where the headline on this article originally did too. "Core inflation fell to 1.9%" is a sentence that requires you to pick one of three official measures and then not check its history.
Statistics Canada's table of Bank of Canada core measures shows what actually happened across the past six months:
CPI-trim did not fall in August. It held, for the third month running. CPI-median held at 2.0%. The only measure that moved down was CPI-common, from 2.7% to 2.6%. A three-month plateau at or below target is a more durable signal than any single monthly step, and it is the more useful fact for someone deciding when to lock a rate — but it is not a decline, and reporting it as one overstates what the data did.
The three are not interchangeable. CPI-trim excludes the components whose monthly price changes sit in the tails of the distribution, which is how a 22.8% gasoline move gets removed automatically rather than by judgment. CPI-median takes the price change at the 50th percentile of the basket by weight — the middle item, immune to extremes at either end. CPI-common isolates the price movements that show up across categories at once, which makes it the slowest to respond and the highest of the three right now at 2.6%.
RBC Economics reached the same reading independently on release day, noting that CPI-trim was 1.9% and CPI-median 2%, leaving their average unchanged from the previous month. The bank's economists also reported that inflation breadth narrowed: the share of basket components growing faster than 3% over the past three months slipped to 25.9% from 30.6%, and the share growing faster than 5% fell to 17.7% from 25.4%.
Narrowing breadth is the technical version of the argument this article is making. Fewer categories are rising quickly. The pressure is concentrated, not general.
For the July baseline, the Nova Scotia Department of Finance's parallel analysis confirms the same starting point from a second government source: CPI-trim at 1.9% in July, unchanged from June.
This is consistent with the picture that emerged when business inflation expectations eased to 3.3% and forecasters converged on a hold through 2026.
Shelter — the largest line in most household budgets — rose 1.5% year over year in August, up from 1.3% in July. Against a 3.0% headline and a services aggregate running at 2.8%, that is remarkably cool. In Statistics Canada's major-components table the shelter index rose just 0.3% on the month.
But a single national shelter number conceals two components moving in opposite directions. Rent accelerated to 2.8% year over year from 2.5% in July, driven by Manitoba at 4.3% and Ontario at 2.4%, and rent was among the largest upward contributors to the annual change. The costs specific to owning, meanwhile, went the other way.
"Shelter" in the CPI covers both renting and owning. When the national figure looks calm, check which half of it is producing that calm — for owners and renters in the same city, the lived experience of shelter inflation right now is not the same number.
The provincial dimension of this has been visible for months, most sharply where Ontario's homeowners' replacement cost fell 4.6% in July, making it the only province where inflation did not accelerate.
Two owned-accommodation components appear on the downward side of the ledger. Statistics Canada's contributors table lists homeowners' replacement cost at -1.9% and other owned accommodation expenses at -1.6% among the main downward contributors to the twelve-month change, alongside household appliances at -5.5% and natural gas at -6.7%.
Homeowners' replacement cost tracks the depreciating structure rather than the land, so it moves with new-home prices. When it falls, the CPI records owning a home as getting cheaper — regardless of what any individual owner's payment is doing.
Which brings up the genuinely counterintuitive part of this release, and the part almost nobody explains. Mortgage interest enters the CPI directly, and during the tightening cycle it became one of the largest single drivers of Canadian inflation. Mortgage interest costs ran at 31% year over year in August 2023, and by January 2026 that had decelerated to 1.2%, according to TD Economics analysis reported by STOREYS. The same analysis projected that modest payment increases would persist early in 2026 before declines became the dominant renewal outcome in the second half of the year.
The mechanism runs backwards from intuition. As pandemic-era contracts reset, the measured mortgage interest cost index decelerates — so the renewal wave that felt like a shock to households has been pushing the national inflation number down rather than up. That is a statement about the index, not about anyone's cash flow. A household whose payment rises on renewal still pays more; the index simply records a smaller year-over-year increase than it did when rates were climbing.
It is also not a straight line, as second-quarter data showing household mortgage interest costs growing at their fastest pace since 2024 made clear, even with the policy rate pinned at 2.25%.
This is the piece of machinery worth understanding before the next print, because it explains how a rate quote can move in a week when the Bank of Canada does nothing at all.
Lenders fund five-year fixed mortgages against five-year Government of Canada bonds, so the yield on that bond — not the overnight rate — is the input that moves fixed pricing. On the Bank of Canada's published benchmark series, the five-year yield went from 3.44% on September 8 to 3.48% on September 9, 3.63% on September 10 and 3.65% on September 11 — 21 basis points in four sessions, all of it before this release landed.
CBC News put the transmission plainly earlier in the month, reporting that "the higher the yields for those bonds are, the higher the banks set their interest rates for those loans" and that fixed mortgage rates will not drop substantially until yields do. Worth noting: several widely read bank explainers describe this relationship the other way around, telling readers that the policy rate drives bond yields which drive fixed rates. That ordering makes the events of early September impossible to explain.
Variable-rate mortgages are priced as a discount or premium to prime, and prime moves with the policy rate. Prime sat at 4.45% as of September 11, 2026, 2.20 percentage points above the Bank's 2.25% overnight rate, and has been unchanged since October 2025.
So the two products are wired to different switches. Variable moves eight times a year, on announcement dates, and only when the Bank acts. Fixed moves every trading day, on whatever the bond market decides about inflation five years out. An inflation print like this one is read first by the bond market — which is where a renewing borrower should look for the immediate answer.
The most useful fact about the August print is that it surprised no one. A Reuters poll of analysts had forecast annual inflation of exactly 3.0% with no monthly change, and that is precisely what arrived. An on-consensus release is, by construction, already reflected in bond pricing.
The reaction was correspondingly small. The Canadian dollar weakened 0.29% to C$1.3909 per U.S. dollar and two-year Government of Canada bond yields firmed by 1.4 basis points to 2.703%, a rounding error next to the 21-basis-point move the five-year had already made in the preceding week.
The practical translation: this release, on its own, is not the thing that repriced anyone's fixed-rate quote. The week before it was.
Two independent reads published within hours of the release landed in the same place. RBC Economics called the report broadly consistent with its base case of a hold through the remainder of 2026 followed by gradual increases in 2027, while noting that risks are increasingly tilted toward an earlier hike if elevated energy prices persist. RSM Canada economist Tu Nguyen kept a baseline hold at 2.25% until the end of 2026 before a first-quarter 2027 increase, while writing that "the odds of a hike in December rose slightly".
Note the shape of that agreement. Both see a hold now. Both see the risk skewed toward tightening sooner rather than later, and neither sees a cut. For a renewing owner, that asymmetry is the real message of this print — the case for waiting is not that conditions are likely to improve.
The Bank of Canada has already said the quiet part out loud. When it held the policy rate at 2.25% on September 2, it observed in its own announcement that CPI inflation has been hovering around 3% in recent months mainly because of persistently higher gasoline prices, and flagged that the longer high oil prices persist, the greater the risk of spillover into other prices.
That framing was reinforced when the Bank warned that consecutive rate hikes were possible if oil kept pushing inflation higher. The direction of travel in the Bank's own language has been consistent for months: patient now, conditional later.
None of the above is a forecast this publication endorses, and no rate outcome is certain. Forecaster views are included because their agreement on direction — and their disagreement on timing — is itself the information a renewing borrower can act on.
Between now and the end of October there are exactly two scheduled events that can move the numbers behind a mortgage quote. Everything else is noise or market drift.
A reader holding a renewal date inside that window can use a short checklist rather than a rate prediction:
When comparing offers, ask what a quoted fixed rate is priced against and when it was last repriced. A quote set before the five-year yield moved in the week of September 8 may not survive contact with the current bond market.
The honest limit on all of this is time. As RSM Canada put it, inflation is mostly contained right now — but the longer the energy shock runs, the more likely it becomes broad-based. The composition argument in this article holds exactly as long as high fuel prices stay confined to fuel. The October 19 print is the next place to check whether they have.
Stay on top of Canadian homeowner news
About the Author
Angela Nightingale is the Senior Editor at Homeowner.ca with two decades of experience in digital publishing and content strategy. She has owned two homes, taken on her share of DIY projects, and learned what most guides fail to mention. She writes from the belief that the best home guidance comes from people who have lived through the decisions — and her goal is always to leave readers feeling confident, not overwhelmed.
