3.7% Headline, 0.2% Core Expected (US CPI Preview, 14 October 2026): What to Expect From September Inflation and What It Means for the Dollar
September CPI lands 14 October at 8:30 ET. Forecasts see headline at 3.6–3.7% on a 7% gasoline jump and core near 0.2%. The scenario map for the dollar.
3.7% Headline, 0.2% Core Expected (US CPI Preview, 14 October 2026): What to Expect From September Inflation and What It Means for the Dollar
September's Consumer Price Index lands at 8:30 a.m. Eastern on Wednesday 14 October, and the forecasts describe two different reports stapled together: a headline expected to jump to 3.6–3.7% on a 7% rise in pump prices, and a core index expected to rise only about 0.2% on the month. The headline is the number on every screen; the core is the one the Federal Reserve will read two weeks before its 28 October decision. For the dollar, the question is not whether inflation is high — it is whether this print moves the timing of the Fed's next hike from December back to October.
- When. Wednesday 14 October, 8:30 a.m. ET (12:30 UTC, 20:30 HKT). The last CPI before the 28 October FOMC decision.
- The forecasts. RBC: headline +0.6% m/m, 3.7% y/y; core +0.2%, 2.5%. Cleveland Fed nowcast: headline +0.53%, 3.60%; core +0.20%, 2.39%.
- The headline is a gasoline story. EIA's weekly survey averaged $4.355 in September against $4.058 in August — about +7.3% on the month and +37.6% on the year.
- The core annual rate is a coin-flip on rounding. A seasonally adjusted core print of roughly 0.22% is the line between "2.4%" and "2.5%". Read the monthly figure, not the rounded annual one.
- The Fed has already told you its direction. The September minutes: "most participants" saw another increase as "likely" appropriate "by year end." The open question is timing — and FedWatch priced about an 82% chance of an October hold.
- See how the interest-rate factor is scoring the dollar against seven other currencies on the live meter.
When the report lands and what it covers
The Bureau of Labor Statistics releases the September index at 8:30 a.m. Eastern — 12:30 UTC, 13:30 in London, 20:30 in Hong Kong. It covers prices collected through September, which makes it the month in which pump prices did most of their climbing and before the early-October retreat in crude.
The calendar around it matters more than usual. Producer prices and retail sales follow on Thursday 15 October. The Fed's communications blackout before the 27–28 October meeting starts on Saturday 17 October, which means — unlike August's CPI, which landed inside a blackout — officials have three days to say what they make of this print. The bank earnings season opens the same week, so the US500 is digesting rate-sensitive income statements while the rate path itself is being repriced.
What the forecasts expect
| September 2026 CPI | August actual | RBC forecast | Cleveland Fed nowcast (9 Oct) |
|---|---|---|---|
| Headline m/m, SA | +0.4% | +0.6% | +0.53% |
| Headline y/y | 3.4% | 3.7% | 3.60% |
| Core m/m, SA | +0.29% | +0.2% | +0.20% |
| Core y/y | 2.4% | 2.5% | 2.39% |
Sources: August actuals calculated from index levels published by the Bureau of Labor Statistics (seasonally adjusted all-items 334.131, core 337.765). Nowcast from the Cleveland Fed, updated 9 October. RBC Economics' forecasts are from its 9 October week-ahead note; Natixis, separately, expects 0.6% on the headline and 0.19% on core.
The forecasters agree on the shape: a headline roughly three times the core. Where they disagree is entirely in the second decimal — and, as the next section shows, that disagreement is about rounding, not economics.
We wrote up August's release in our August CPI post. Core came in at 0.29% against a 0.2% consensus, the first firm reading after three soft months. The single most useful thing September can tell you is whether that was the start of something or a one-off.
Why the headline is a gasoline story
The Energy Information Administration's weekly retail survey is published long before the CPI, so the energy leg of this report is close to arithmetic. Regular gasoline averaged $4.058 a gallon across August's five weekly readings and $4.355 across September's four, peaking at $4.478 in the week of 21 September. That is a rise of about 7.3% in one month. A year earlier, September 2025 averaged $3.166, so the annual gain is about 37.6% — well above the 27.4% the CPI's gasoline index showed in August.
Gasoline carries a weight of a few percent in the basket. A 7% move in a few-percent weight is worth roughly a quarter of a percentage point on the headline by itself, before diesel feeds into freight and before airfares, which rose 2.68% in August, are counted in core. That is the gap between a 0.2% core and a 0.6% headline.
The rounding knife-edge in core
The headline figure for core — the "2.4%" or "2.5%" that leads every summary — is a twelve-month comparison, and this month it sits almost exactly on a rounding boundary.
The arithmetic, from BLS unadjusted index levels: August 2026 core stood 2.446% above August 2025. September 2025's unadjusted core index rose 0.253%. So the annual rate rises above 2.45% — and rounds to 2.5% — only if September 2026's unadjusted core rises by about 0.257% or more. Using last September's gap between adjusted and unadjusted readings as a guide, that is roughly a 0.22% seasonally adjusted print.
That is why RBC says 2.5% and the Cleveland Fed says 2.39% while both expect about 0.2% on the month. The same logic applies to the headline: whether it rounds to 3.6% or 3.7% depends on a seasonally adjusted gain somewhere around 0.53–0.54%, which is exactly where the forecasts cluster.
The lesson for reading the release is simple. A "2.5%" core headline is not necessarily hotter than a "2.4%" one; it may be the same month rounded the other way. The monthly core figure to two decimals — 0.20% versus 0.29% versus 0.35% — is the number that carries information.
Where the Fed stands going in
The direction is not in doubt. On 16 September the Committee raised the target range by a quarter point to 3.75–4.00%, its first hike in three years. The minutes of that meeting, released 7 October, state that "most participants assessed that another increase in the target range for the federal funds rate would likely be appropriate by year end." The same document records that several participants saw short-term inflation expectations as elevated — the channel through which an energy shock becomes a policy problem.
The timing is what moved. After the 2 October jobs report showed payrolls up just 29,000, the market pushed the next hike from October toward December. On 5 October CNBC reported that CME FedWatch priced a nearly 82% chance of a hold on 28 October, even as the 10-year yield hit its highest level since 2002. The run-up is covered in our Fed minutes preview.
So the market already expects a hot headline and a December hike. Wednesday's report can only move the dollar by changing one of those two assumptions — and only the core can realistically change the second.
Three scenarios and what each means for the dollar
The dollar connects to this report through the interest-rate factor, one of the five factors the meter scores across the eight majors. What it reacts to is the change in the expected policy path — and the two-year yield is where that shows up first.
| Scenario | Core m/m | What it changes | Dollar channel |
|---|---|---|---|
| Soft core | 0.1–0.2% | Confirms the "energy, not broadening" reading; December stays the base case | Little new for the rate factor; any dollar strength from the headline tends to fade |
| In line | ~0.2% with a 0.6% headline | Nothing about timing; rounding drives the headlines | Muted; attention shifts to Thursday's PPI and retail sales |
| Second firm month | 0.3% or above | Two firm core months in a row, after the minutes' "by year end"; October hike odds rise from roughly one in five | Front-end yields reprice first; the dollar's carry advantage widens against low-yielders |
Two asymmetries matter. First, a headline surprise driven by gasoline is weaker evidence than a core one, because central banks conventionally look through supply-driven energy swings unless they start to feed inflation expectations — which is exactly why the minutes' note on elevated short-term expectations matters. Second, the dollar is not starting from neutral. The euro hit a 17-month low of 1.1204 on 5 October on European politics, as we covered in our euro note. When a pair has already moved a long way, a confirming print has less room to push it further than a contradicting one has to pull it back.
For the yen, the channel runs the other way: a firm core print widens the US–Japan rate gap the Bank of Japan has only partly narrowed with its September hike to 1.25%. The meter's dollar page, euro page and yen page show how each is scoring now.
What else to watch inside the report
- Shelter. Rent and owners' equivalent rent rose 0.17–0.19% in August and annual shelter fell to 3.04%. If that keeps slowing, it can offset firmness elsewhere in core.
- Transport services and airfares. These lines carry energy into core. August's 2.68% airfare jump was the clearest second-round effect so far.
- Core goods. Their three-month annualised rate rose from roughly zero to 0.87% in August. Diesel-driven freight costs are the route by which an oil shock reaches items that never touch a pump.
- Next month's base. The October 2025 CPI was never published — the Cleveland Fed notes it uses its own nowcast in place of that release — so the annual rates in next month's report rest on an estimated base. That is a reason to put more weight than usual on monthly figures from here.
Bottom line
Wednesday's headline will almost certainly look alarming, and almost all of the alarm is gasoline that has already started to reverse. The information is in the monthly core figure to two decimals. Around 0.2% leaves the Fed's December-not-October timing intact. A second consecutive 0.3% puts October back on the table, and that is the outcome that would move front-end yields and the dollar. More on how we build these notes is on our about page.
Educational macro context only — not investment advice.