Bank of Canada Rate Hike Bets: Macklem Speaks, and the Precedents Are Thin
Tiff Macklem speaks before the Halifax Partnership in Halifax this morning, his first public remarks since the August inflation report landed, and he does it with the market pricing a Bank of Canada rate hike next month at better than a coin flip. The Bank lists his topic as “Economic developments.” The text goes up on the Bank of Canada’s speech page at 11:05 ET, the Governor speaks at approximately 11:20 ET, and a moderated press conference follows at approximately 12:30 ET.
We are publishing before any of that, so nothing here is a preview of what he will say. What follows is the context worth having before 11:20. The Bank has already written down, in its own words, the one development that would force it to respond. And the configuration Canada is in right now, headline inflation running well above the Bank’s own core measure while that core sits at or below 2%, has occurred in nine of the 288 months since September 2002. We computed what the Bank did in each of them.
Where the rate decision stands
The policy rate is 2.25% and has been since October 30, 2025, and the next decision is October 28. Montreal Exchange one-month CORRA futures, as of the September 18, 2026 settlement, price a 54% chance of a 25 basis point increase, an implied rate of 2.39%, with the contracts quoted on the Montreal Exchange’s CORRA futures page. That would be the Bank’s first increase since July 13, 2023. We have argued both sides of that decision at length already, so rather than repeat it: our full read of the August inflation report carries the component breakdown, the core measures and the case each way.
The Bank wrote down its own trigger
On September 16 the Bank published the Summary of Governing Council deliberations for the September 2 decision, covering meetings held August 25 to September 2. It left the rate at 2.25%, and in explaining why, it named the condition under which it would not.
if higher energy prices did spill over into other components of the CPI, members agreed that it could require a monetary policy response to prevent broad-based inflation from setting in.
At the time of those meetings the answer to that test was no, “there was little evidence that higher gasoline prices were passing through broadly to the prices of other goods and services,” but it was a tolerance with an expiry date attached: “members agreed that the longer they were high, the more likely they would be passed through. This increased the upside risks to inflation.” The counterweight was a labour market Governing Council called “still soft, with the unemployment rate around 6½% and subdued wage growth.”
Now the sequencing, which is easy to get backwards. Those meetings ran August 25 to September 2, and Statistics Canada published the August CPI report on September 14. The deliberations came out two days after that report, but they describe a discussion that had only July’s data in front of it. The Bank has not publicly responded to the August numbers at all, which is most of why a scheduled speech on economic developments is worth an hour this morning.
What the data has done since
The August figures cut both ways. Headline inflation held at 3.0% year over year, unchanged from July, while CPI excluding gasoline, the measure that speaks most directly to the Bank’s spillover test, rose to 2.4% from 2.2%. That moved the wrong way for a hold. Gasoline was still up 22.8% year over year, easing from 25.7% in July, all of it in Statistics Canada’s August consumer price index release.
Against that, the Bank’s preferred gauges did not move: CPI-trim was 1.9% for the third consecutive month and CPI-median 2.0%, both at or below the midpoint of the target range. Unemployment was 6.4% in August and 6.4% in July, down from a 2026 peak of 6.9% in April, so the softness the Bank leaned on is real but is not deepening. Retail trade and payroll employment for July arrive Thursday September 24, where StatCan’s advance estimate says “sales decreased 0.8% in July,” and July GDP lands Tuesday September 29, with September CPI not due until October 19.
Nine months out of 288, and what the Bank did in them
Here is the work. We took Statistics Canada’s all-items CPI index (table 18-10-0004, vector v41690973), computed the year over year change ourselves, and subtracted the published CPI-trim series (table 18-10-0256, vector v108785715). That gives a monthly gap between headline and core for every month from September 2002 to August 2026, the full 288-month span over which the trim measure exists.
In August the gap was 1.1 percentage points, headline 3.0% against trim 1.9%, so we screened the whole series for months where the gap was at least 1.1 points and CPI-trim was 2.0% or lower. That threshold is set by the current month’s own two readings, so it is a test built to fit the present. What it returns is not a statistic but a list of dated episodes with known outcomes: nine months out of 288, about 3% of the record.
| Month | Headline | CPI-trim | Gap |
|---|---|---|---|
| September 2005 | 3.2% | 1.8% | 1.4 |
| March 2011 | 3.3% | 1.6% | 1.7 |
| April 2011 | 3.3% | 1.5% | 1.8 |
| May 2011 | 3.7% | 1.8% | 1.9 |
| June 2011 | 3.1% | 1.8% | 1.3 |
| August 2011 | 3.1% | 2.0% | 1.1 |
| May 2026 | 3.2% | 2.0% | 1.2 |
| July 2026 | 3.0% | 1.9% | 1.1 |
| August 2026 | 3.0% | 1.9% | 1.1 |
Source: Statistics Canada tables 18-10-0004 and 18-10-0256, monthly, September 2002 to August 2026. Gap computed in house.
Three of the nine are this year, which leaves six prior months in exactly two clusters: one month in 2005, and five consecutive months in 2011, March through August, with July 2011 just short of the threshold.
The 2011 cluster, five of the six, is the closer match on the surface: headline above 3%, trim as low as 1.5%. The target for the overnight rate was 1.00% throughout, where it had sat since September 8, 2010, and the Bank did not move it at all through the episode. Its next change in either direction was a cut to 0.75% on January 21, 2015, computed from the Bank’s own target for the overnight rate series (V39079) via its Valet API.
September 2005 goes the other way. The Bank was already tightening and kept tightening through a gasoline spike that followed Hurricane Katrina. Its Bank Rate series (V122530) runs 3.00% in September 2005, then 3.25% in October, 3.50% in December, 3.75% in January 2006, 4.00% in March, 4.25% in April and 4.50% in May 2006. With the overnight target 25 basis points below the Bank Rate, that is a move from 2.75% to 4.25%, 150 basis points in eight months.
Five to one toward holding, then, and the ratio is close to meaningless, because the five are one episode in one year. This is a sample of two, not six. What two cases can support is a distinction rather than a probability: in 2011 the Bank looked through an oil-driven spike with soft core, and in 2005 it was already mid-cycle for its own reasons when the spike arrived. Which makes the useful question not what history says the Bank will do on October 28, but whether it is currently inside a cycle or outside one. It has been on hold for almost eleven months and its last move was a cut. That is an observation about where the Bank is standing, not a forecast about where it steps next.
What that leaves for a Canadian portfolio
Last week the TSX split more or less along the line a rate decision splits it. Week over week to the September 18, 2026 close, real estate sector ETFs fell 1.53%, the weakest of the eight sectors, while financials rose 0.84% and utilities rose 1.12%.
Real estate is the most rate-sensitive corner of the index and it was the corner that fell. What matters more than one week’s move is which listed landlords can keep paying through a higher rate path, which is what our breakdown of Canadian REIT stocks weighs. Financials sit on the other side of the same decision, though a higher policy rate pushes funding costs, margins and credit quality in different directions at once, and the Big Six are not interchangeable on any of them; our rundown of Canadian bank stocks takes them one at a time.
One last caveat on the history above, because it is the kind that gets lost once a number is in a headline. Past central bank behaviour does not predict future policy decisions, and two episodes could not predict one even if it did.
None of that is a reason to reposition ahead of a speech. The Bank wrote down what would change its mind, the one measure that speaks to that test moved 0.2 points the wrong way in August, its own preferred gauges did not move at all, and the Governor has not spoken publicly since. That is why 11:20 this morning is worth the hour.
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. Policy rate, rate history and quotations from the Bank of Canada’s own publications: the Summary of Governing Council deliberations for the September 2, 2026 decision, the target for the overnight rate series V39079 and the Bank Rate series V122530 via the Bank’s Valet API. Inflation figures from Statistics Canada tables 18-10-0004 and 18-10-0256 and the August 2026 consumer price index release of September 14, 2026. Unemployment from Statistics Canada table 14-10-0287. Rate pricing from Montreal Exchange one-month CORRA futures, September 18, 2026 settlement. Sector and market figures as of the September 18, 2026 close.
