Inflation and prices
US CPI measures how fast consumer prices are rising. Markets do not trade the level, they trade the surprise against what was expected, because that is what moves Fed pricing. A print can be high and still be bullish for bonds if it came in lower than the market feared.
Also known as US consumer price index, US inflation
US headline CPI rose 3.7 per cent in the year to September 2026, against 3.4 per cent in the year to August 2026. Core CPI, which strips out food and energy, was 2.6 per cent against 2.4 per cent.
On the month, seasonally adjusted, headline was +0.6 per cent and core was +0.3 per cent. The unadjusted all items index stood at 336.940.
Release schedule
The next US Consumer Price Index release is scheduled for 10 November 2026 at 08:30 New York time, which is 13:30 in London. That date is read from the BLS Consumer Price Index release schedule each time this page is built, so it follows the publisher rather than being stored here.
The Bureau of Labor Statistics publishes its CPI schedule a year in advance, and that schedule states which reference month each release covers, so nothing here is inferred from the release date. The figures on this page are refreshed from the BLS series once the release has run, and until they are the snapshot says so rather than showing the previous month as though it were current.
Because CPI is the single release most likely to reprice the Fed, and everything else is downstream of that.
Inflation is the variable the central bank is explicitly targeting, so an inflation surprise changes the expected policy path more directly than growth or employment data. The market does not wait for the Fed to respond. It reprices the next two years of policy within seconds, and the front end of the Treasury curve moves first.
That is why the reaction runs in a fixed order. Fed pricing moves, the 2 year yield follows, the dollar follows the rate differential, and gold and equities follow the discount rate. A trader who understands that sequence knows where to look in the first minute rather than guessing.
It is also why CPI day is one of the few scheduled events where the timing is known to the second. The release is 08:30 New York, and liquidity thins deliberately in the minutes before it.
The market cares about the surprise, not the number.
A three per cent print is not hawkish or dovish on its own. It is hawkish if the market expected two point eight and dovish if it expected three point two. This is the single most common mistake in reading inflation data, and it is why a hot headline can be followed by falling yields.
Within the release, core matters more than headline for policy. Food and energy are volatile and largely outside the central bank's influence, so core is treated as the better read on underlying pressure. A hot headline driven entirely by petrol prices moves the Fed far less than a hot core.
The monthly pace usually matters more than the annual rate for the immediate reaction. The annual figure carries a base effect from twelve months ago that everyone already knows about, so it is the monthly change, seasonally adjusted, that tells you what is happening now.
Composition is the last filter. Shelter, services excluding shelter and goods behave differently and are read differently, because a rise driven by lagging shelter data means something quite different from one driven by services inflation reaccelerating.
Scenario analysis
Hotter scenario
A hotter print takes cuts out of the curve before it puts hikes in.
The first move is in Fed pricing, and the first ten to fifteen basis points on the 2 year are usually the market removing expected easing rather than adding tightening. The dollar tends to firm on the widening rate differential and gold tends to struggle as the real return on the alternative rises.
What matters next is whether the long end follows. If the 2 year rises and the 10 year does not, the market believes the Fed will hold tighter for longer without changing where rates end up. If both rise, the market has changed its view of inflation itself, which is a much bigger move.
In line
A print in line with expectations is not a non event, and treating it as one is a mistake.
If the number lands where the market expected, Fed pricing should not move, and if yields move anyway the information is in what else changed. Positioning going into the print, a surprising composition beneath an in line headline, or a revision to the previous month can all move the curve on a number that looked unremarkable.
The useful habit is to check whether the front end actually did nothing. An in line print that still lifts the 2 year tells you the market was leaning the other way and is being forced to unwind.
Softer scenario
A softer print adds cuts back into the curve, and the front end takes most of the move.
Because the 2 year spans the window the market is repricing, it falls further than the 10 year, the curve steepens, and the dollar usually softens as the rate differential narrows. Gold typically gets its cleanest tailwind here, since the real return on the alternative is falling.
The version that catches people out is a soft print the Fed publicly ignores. If officials push back within days, the move reverses, and anyone who traded the first hour is left holding a position the curve no longer supports.
Market impact
Bond yields
The front end moves first and moves most.
CPI changes the expected policy path, and the 2 year Treasury is the price of that path, so it carries the bulk of the repricing. The 10 year moves less, because a single monthly print rarely changes the market's view of where rates settle over a decade.
That difference is visible in the curve. A hot print usually flattens it and a soft print usually steepens it, and the size of that change tells you whether the market read the release as a shift in policy or a shift in the long run inflation outlook.
Currencies
The dollar reacts through the rate differential, not through inflation itself.
Higher US inflation is not mechanically dollar positive. What moves the currency is the change in expected US short rates relative to expected rates elsewhere. A hot US print that lifts the US front end while European and Japanese pricing sits still widens the differential and pulls capital toward the dollar.
This is why a hot print can leave the dollar flat. If the market decides the central bank will look through the number, the policy path does not move, the differential does not move, and neither does the currency, however alarming the headline looked.
Gold
Gold responds to what CPI does to real yields, not to inflation as a headline.
This is the point most retail commentary gets backwards. Gold is not a simple inflation hedge. A hot print that raises expected policy tightening more than it raises expected inflation pushes the real yield up, and gold falls on inflation news. A hot print the Fed is expected to tolerate pushes the real yield down, and gold rises.
So the number to watch after CPI is not the CPI. It is the real yield, which is the nominal yield minus the breakeven, and the direction it takes in the hour after the release.
Equities
Equities feel CPI through the discount rate and through the cost of leverage.
A hot print lifts the near part of the discount curve and raises the return on cash, which is the alternative to holding equity risk at all. Long duration growth companies, whose value sits in distant cash flows, feel it first and hardest.
A soft print does the reverse, and the rally is usually led by exactly the names that suffered most on the way up. That asymmetry is why index level moves on CPI day often understate what happened underneath, where the rotation is far larger than the headline suggests.
Three mistakes come up on every CPI day.
The first is trading the level instead of the surprise. High inflation that is lower than feared is a dovish event, and traders who react to the headline number alone repeatedly find themselves on the wrong side of the first move.
The second is treating gold as an inflation hedge. Gold follows the real yield, and a hot print that raises the expected policy response harder than it raises expected inflation is bearish gold, whatever the inflation headline says.
The third is ignoring revisions. The previous month can be revised at the same moment the new month is published, and a revision large enough to change the trend will move the curve even when the current month lands exactly where everyone expected.
The framework breaks in two identifiable situations.
The first is when the central bank has told the market in advance that it will look through the data. Under an explicit commitment to hold, an inflation surprise can arrive and the policy path barely moves, which severs the usual link from CPI to the front end and therefore to everything downstream.
The second is a supply shock large enough to change growth expectations at the same time. When an inflation surprise is also a demand destruction signal, the front end can price tightening while the long end prices weaker growth, and the curve flattens for reasons that have nothing to do with the usual mechanism.
If CPI and the front end are moving in opposite directions for more than a few sessions, one of those two conditions is usually present.
Official reference
The US Bureau of Labor Statistics publishes the Consumer Price Index for All Urban Consumers, known as CPI-U, at 08:30 New York on a scheduled date roughly two weeks after the reference month ends. The figures here are read directly from the BLS public API.
Year on year rates are calculated from the NOT seasonally adjusted index and month on month rates from the seasonally adjusted index. That is the convention BLS and market participants use, and mixing the two produces figures that will not match anything published elsewhere. Core excludes food and energy.
Seasonally adjusted CPI is subject to revision. BLS recalculates seasonal factors each year, which can change previously published monthly rates for the preceding five years. The values shown here are re derived from the source on every refresh rather than stored once, so a revision is reflected automatically.
No market consensus or forecast figure is shown on this page. Expectations are not published by the BLS and this desk has no licensed consensus provider, so quoting one would mean sourcing a number we cannot stand behind. Everything above is official data.
One quick step
Drop your email and the desk will send you the details.
KenMacro
Free community
Thousands of traders staying ahead of the macro, together. It's free, and you're one click away.
Instant access. Free, always.
Three thousand traders already get it. The full week ahead, in your inbox before the London open.
Free, weekly, no spam. Unsubscribe in one click.
The next Sunday brief lands in your inbox at 12:00 BST. In the meantime, head to macro insights for the desk reactive coverage.