Canada’s July GDP Was Flat. Revisions Moved the Quarter More
Canada’s real GDP by industry was “essentially unchanged” in July, Statistics Canada said in this morning’s release. The number that moved more was not July’s. Revisions published alongside it, to months already in the books, shifted the third quarter tracking estimate further than the month everyone had been waiting for. That quarter now tracks somewhere between roughly +1.43% and +1.97% annualized on our arithmetic, and which end you take depends on an August figure StatCan has estimated but not yet measured. Data as of September 29, 2026.
One thing to fix in place first. Everything here is the monthly industry-based series, the value added by each industry. The quarterly expenditure-based series measures spending, it is the basis the Bank of Canada projects on, and the two do not have to agree.
The level June reached is doing the work
July came in at 2,371.771 billion chained 2017 dollars at annual rates against June’s 2,371.886 billion, a change of -0.005%, which is why it publishes as 0.0%. Year over year, July is up 1.42%.
The quarter is a different question, because a quarterly rate compares an average of three months against the average of the previous three. Q2 averaged 2,363.393 billion and was strong in its own right, up 0.94% on the quarter and 3.83% annualized on this basis. July sits 0.35% above that Q2 average. That gap is the carry-over, and it is already banked.
Run it forward on our arithmetic on Statistics Canada table 36-10-0434, vector v65201210. If August and September did nothing at all and simply held July’s level, Q3 would still print +1.43% annualized. Put in StatCan’s +0.2% advance estimate for August, hold September flat, and it becomes +1.97%. Those are illustrations under stated assumptions, not forecasts, and the entire gap between them is an August estimate that will be replaced by a measurement on October 30.
Source: Statistics Canada table 36-10-0434, vector v65201210, monthly real GDP by industry to July 2026, and the August advance estimate in the September 29 release.
The revisions moved the quarter more than July did
This release revised the series back to January 2025, and that is the more interesting half of the morning.
Two days ago we ran this same carry-over calculation on the vintage that existed then, in the arithmetic we published before the release. On that data the Q2 average was 2,362.5 billion, June was 2,369.3 billion, and a completely flat third quarter came to +1.15% annualized.
This morning June was revised up 0.11% to 2,371.886 billion and the Q2 average rose 0.04%. The same flat-quarter scenario now prints +1.45%. Adding July’s own change takes it to +1.43%.
So the tracking estimate moved 0.27 of a point between our preview and today. About 0.29 of that came from revisions to months already published, and about -0.02 came from July, the month everyone was waiting for. A reader watching only the headline saw a zero and would fairly conclude the quarter had weakened. On the levels it strengthened, and none of that came from the month in the headline.
The second quarter itself was rewritten in the same pass. On the earlier vintage, a Q1 average of 2,341.6 billion against a Q2 average of 2,362.5 billion gave +0.89% on the quarter, +3.62% annualized. On this morning’s vintage, Q1 is 2,341.300 billion and Q2 is 2,363.393 billion, which is the +0.94% and +3.83% quoted above. Q1 was revised down and Q2 up, and a quarter that was already three months old got visibly better without a single new month of data.
Which is the honest frame for that +1.97%. It is not a point estimate. It is the top of a range whose bottom is +1.43%, sitting on a series whose flat-quarter reading just moved 0.29 of a point in one ordinary revision. The August release on October 30 revises these months again, and it arrives two days after the Bank has already decided.
What moved underneath the zero
Ten of twenty industrial sectors expanded. Goods-producing industries were essentially unchanged and services-producing industries essentially flat, which is the shape of a month where the gains and the losses were close to the same size.
| Sector | July change |
|---|---|
| Utilities | +1.7% |
| Construction | +1.3% |
| Accommodation and food services | +0.8% |
| Professional, scientific and technical services | +0.3% |
| Real estate and rental and leasing | +0.2% |
| Wholesale trade | -0.4% |
| Mining, quarrying and oil and gas extraction | -0.5% |
| Manufacturing | -0.9% |
| Retail trade | -1.0% |
Source: Statistics Canada, gross domestic product by industry, July 2026.
Construction rose 1.3% in a fourth straight monthly increase with every subsector expanding. StatCan names two as contributing the most: engineering and other construction activities at +1.5%, and non-residential building construction at +2.9%, the latter its largest growth rate since January 2022, which the agency attributes to institutional building activity associated with construction of a new hospital in Ontario. Utilities rose 1.7%, more than offsetting June’s contraction, with electric power generation, transmission and distribution up 1.7% on what StatCan links to a July heat wave.
Against that, retail trade fell 1.0% and largely gave back June’s increase, every subsector down except building material and garden equipment dealers, with gasoline stations and fuel vendors off 3.5%. Manufacturing fell 0.9%, its first decline in four months, and petroleum and coal product manufacturing contributed most at -5.7% as activity at petroleum refineries dropped 6.2% on unplanned downtime at a refinery in southwestern Ontario. Mining, quarrying and oil and gas extraction fell 0.5%, declining across all subsectors, potash down 6.4% on lower Saskatchewan output.
What September would have to do to change the quarter
Holding August at the +0.2% advance estimate, the quarter is close to settled by what has already happened. Every row below is an illustration under that assumption.
| September change | Q3 annualized |
|---|---|
| -1.5% | -0.06% |
| -1.2% | +0.34% |
| -0.9% | +0.75% |
| -0.6% | +1.15% |
| -0.3% | +1.56% |
| 0.0% | +1.97% |
| +0.3% | +2.38% |
Our arithmetic on Statistics Canada table 36-10-0434, vector v65201210, with August held at StatCan’s advance estimate.
To wipe the quarter out entirely, September would have to fall 1.46%. Over the 89 monthly changes in this series from March 2019 to July 2026, only two were that bad or worse: March 2020 at -7.54% and April 2020 at -10.66%. The worst month since January 2021 is April 2021 at -1.03%. A 1.5% quarter needs September at -0.34%, and a 2.0% quarter needs +0.02%.
The advance estimate called July correctly, and has now put 0.2% on August
Published with the June figures on August 28, StatCan’s advance estimate said real GDP was “essentially unchanged” in July, with increases in real estate and rental and leasing and in professional, scientific and technical services offset by decreases in retail trade and manufacturing. All four moved in the named direction, at +0.2%, +0.3%, -1.0% and -0.9%, and the headline matched.
That makes eight observations in the tally we keep against first-published figures, reference months December 2025 through July 2026. The direction is worth stating, because it runs one way: the first published figure came in above the advance estimate five times, equalled it twice and came in below it once, a mean difference of +0.06 points on our arithmetic. On this record the advance estimate has erred low. The limits matter as much as the direction. Eight observations is a short record, it measures each estimate against the first published figure rather than the number left standing after revision, and “essentially unchanged” has now appeared three times in the run. So treat it as a scorecard, not a reason to bank the +0.2% now attached to August, where StatCan sees mining and quarrying and retail trade up and oil and gas extraction down. That estimate gets updated with the full August release on October 30.
What this leaves for October 28
The policy rate is 2.25% and has been since October 30, 2025. The next decision is October 28 and it carries a Monetary Policy Report. July GDP is the last GDP reading of any kind the Bank gets beforehand, because August GDP lands October 30, two days after.
Going in, one-month CORRA futures were not pricing a hold. On our reading of the Montreal Exchange’s CORRA futures quotes, the October contract implies an average CORRA of 2.31% against September’s 2.2925%. Only the four days from October 28 to month end would carry a new rate, so that 1.75 basis point gap in the monthly average scales to 13.56 basis points, or 54% of a 25 basis point move, implying 2.39%. Spot CORRA is 2.30% against the 2.25% target. Those odds have firmed: the same calculation on the September 25 settlement gave 47%, which we published two days ago as an easing from 54%, and the most recent settlement puts it back at 54%. Every one of those settlements predates this morning’s release. That is a fact about the futures, not a forecast, and it raises the question of whether the Bank has ever stepped straight out of a hold like this one, which is what the precedent for a Bank of Canada rate increase takes apart.
Now both halves of what today’s data says about that decision. A third quarter tracking +1.97% on the industry basis runs above the 1.5% the Bank wrote down in the July Monetary Policy Report Overview, where growth was “expected to have averaged just above 1% in the first half of 2026 and to average around 1.5% in the second half of the year.” Read alone, that is an economy running hotter than the Bank planned for.
It does not stand alone. That 1.5% is the average of the third and fourth quarters together, and the Bank projects on the quarterly expenditure basis. In a speech published September 21, Governor Tiff Macklem said that “if these new tariffs remain in place, growth could be roughly halved in the fourth quarter, to below 1%,” sizing the exposure at “about 5% of Canada’s goods exports to the United States” and adding that “federal government support programs should mitigate some of the harm.” A quarter near 2% paired with a fourth quarter below 1% averages close to the 1.5% the Bank already published.
That is a reversal of our own conclusion, so we will say so. Two days ago, on the vintage that existed then, we wrote that the second half averaged below the 1.5% the Bank published, that the July projection path “looks stale against the Bank’s own later words”, and that this was “a reason to expect the projection arriving with the October 28 decision to be a lower one”. Two things changed the answer, and neither was July: the revisions, and StatCan’s August estimate. On the flat-from-July path, the one that gives +1.43%, that earlier conclusion still holds.
So the growth side of the October 28 decision is not a number the Bank can lean on. It is a range that moves with revisions of the size this release just delivered, and on the upper end of it growth is landing roughly where the Bank expected, which argues for neither alarm nor a hike by itself. Inflation is the other half of the decision and it is not in this release. Before the Bank meets, Statistics Canada publishes the September Labour Force Survey on October 9 and the September Consumer Price Index on October 19, both at 08:30 ET.
Nothing repriced much on the print
At 09:30 ET, about an hour after the release, the S&P/TSX Composite was at 35,463.71 against Monday’s close of 35,489.90, the Canadian dollar was at 70.51 US cents against 70.65, and the S&P 500 was at 7,699.60 against 7,683.69. Those are live intraday readings taken at 13:30 UTC, not closes.
Where the decision shows up in a portfolio
A rate decision is not felt evenly across a portfolio, and listed real estate is one of the places it is felt directly, whichever way October 28 goes. The question either way is which landlords can keep paying their distributions on the rate path that actually arrives, and that is the work in our breakdown of Canadian REITs.
The macro read for the next month is narrow. The quarter is close to decided by arithmetic that has already happened, the next GDP print arrives two days after the Bank has spoken, and today’s release moved the quarter mostly through months that were published long ago.
Disclaimer: The content on bestcanadianstocks.ca is for informational and entertainment purposes only and does not constitute financial advice. Past performance is not indicative of future results. Always consult a qualified financial advisor before making investment decisions. Real GDP figures are from Statistics Canada’s July 2026 gross domestic product by industry release of September 29, 2026, and from table 36-10-0434, vector v65201210, via StatCan’s Web Data Service. Quarterly rates, carry-over and the September scenarios are our arithmetic on that vector. Bank of Canada rate expectations are computed from Montreal Exchange one-month CORRA futures settlements with spot CORRA from the Bank of Canada Valet API, from settlements that predate the release. Index and currency levels are live intraday readings taken at 13:30 UTC on September 29, 2026, not closes.
