Friday’s CPI: the number that decides the Fed is not the headline
Hi trader,
yesterday afternoon the ECB raised rates.
Second hike of the year, deposit rate at 2.50%, effective from 16 September.
And the curious part is the reason why.
Not because the European economy is running hot, it is growing 0.9% this year, which is barely more than nothing.
It hiked because the war in the Middle East keeps pushing on prices, and because that increase has stopped being just energy.
Let me walk you through it properly, because it is exactly the same question America will be asking itself on Friday afternoon.
What comes out on Friday, and when
It is the US CPI for August, the consumer price index.
The Bureau of Labor Statistics publishes it on Friday 11 September at 8:30 am New York time, which is 1:30 pm here in the Canaries and in London, and 2:30 pm in central Europe.
There are two numbers inside it, and most people only look at one.
| What | August forecast | July (released 12 Aug) |
|---|---|---|
| Headline CPI, year on year | 3.3% | 3.4% |
| Headline CPI, month on month | +0.4% | +0.1% |
| Core CPI (ex food and energy), year on year | 2.4% | 2.5% |
| Core CPI, month on month | +0.2% | +0.2% |
The headline is the big number, the one that shows up in your phone notifications at 8:30 and one second.
The core is the same index without food and energy, which are the two components that swing the most.
It sounds like a detail for economists.
It is the detail that decides what the Fed does next Wednesday.
One clarification that matters, because almost nobody spells it out: the Fed does not measure its 2% on the CPI, it measures it on the PCE, which comes out two to three weeks later.
Friday’s CPI is the first taste, and it is the one the market moves on.
But keep the number in mind: in July the CPI was at 3.4% and the PCE, the index the Fed actually watches, at 3.7%.
Why the big number lies this year
Look at the forecasts in that table.
Headline monthly going from +0.1% to +0.4%, so four times as much.
And meanwhile the annual headline goes down, from 3.4% to 3.3%.
Core moving by one tenth instead, from 2.5% to 2.4%.
If Friday comes out like that, the message is not “US inflation is taking off again”.
The message is “US inflation is being made by the barrel”.
Brent above 100 dollars, a 4% jump yesterday that took it into the 105 area on yet another attack on commercial shipping, pump prices following in stages, half the increase within two weeks and the rest over the month and a half after that.
And it is not just a forecast any more.
Yesterday US producer prices for August came out at +5.4% year on year, and more than three quarters of the monthly increase is energy: +4.2% in thirty days, with diesel at +24%.
Producer prices are the floor above, the one the consumer sees a few months later.
This is called cost-push inflation, inflation that comes from the supply side.
It does not come from people buying too much, it comes from something costing more because getting it here has become complicated.
And there is very little a central bank can do about inflation that comes from the supply side.
Raising rates does not reopen a strait and does not make a tanker arrive any sooner.
The old textbook says a central bank should “look through” a supply shock, because it fades when the shock fades.
The problem is that the textbook holds only as long as that increase stays locked inside energy.
The detail the ECB put right in front of you yesterday
And this is where yesterday becomes useful, instead of just being an item on the evening news.
Look at the projections the ECB published yesterday, the ones from its own staff:
| Year | Headline inflation | Core inflation |
|---|---|---|
| 2026 | 3.0% | 2.5% |
| 2027 | 2.5% | 2.6% |
| 2028 | 2.1% | 2.3% |
Did you notice?
In 2027 and 2028 the ECB expects core to be higher than the headline.
And there is one more detail: it revised that 2027 and 2028 core up by a tenth compared with June.
It does not just expect the second round, it expects it a shade stronger than it did three months ago, and it puts that down to an economy holding up better than forecast and to wages growing a little faster.
Translated: it expects energy to come back down, and it expects that by then the increase will already have passed into everything else.
Transport costing more, so goods costing more, so services, so wages asking to catch up.
They are called second-round effects, and they are the only thing that justifies a hike in the middle of a war.
It did not hike because of the barrel.
It hiked because the barrel is getting into places it does not come out of.
Lagarde said two things in the press conference worth keeping: that this hike was “a no brainer”, an obvious call, and that she cannot anticipate what the next move will be.
No commitment to a rate path, decisions taken meeting by meeting.
The exact same words you will hear on Wednesday from the other side of the Atlantic, want to bet?
Why Friday carries more weight than usual
Because the Fed decides on 16 September.
Two-day meeting, the 15th and the 16th, decision on Wednesday afternoon, and this time there is also the dot plot.
The dot plot, let me explain, is that chart of dots where each of the people sitting at the table, including the ones who do not vote this year, marks where they see rates at the end of this year and in the years after.
Nobody signs their own dot, so nobody really commits, but the cloud of dots tells you which way the wind is blowing inside the room.
It comes out four times a year. The last one was in June. The next one is Wednesday.
And it lands at an unusual moment.
US rates have been sitting at 3.50-3.75% for five meetings.
In July three members voted against, and not to ask for a cut: to ask for a hike.
Then on 28 August, at Jackson Hole, Fed Chair Warsh talked about credibility and about a 2% target with no discounts, and the odds of a September hike almost doubled in three days, from 35% to 66%.
And yesterday the market changed its mind in half an hour.
That producer price print pushed the bet on Wednesday’s hike from about 60% to about 70% on futures, and prediction markets toward 63%.
That is why Friday is not just another data point.
It is the last inflation print before a decision that is already half priced, but not closed.
And a market at 70% is the worst place to be caught by a data release: the hike is in the price enough not to pay you if it lands, but not enough to protect you if it does not.
And afternoons where one number can still move a decision that is already half written are the afternoons when the chart does not forgive a lapse in attention.
The four roads the afternoon can take
I am not giving you entry signals, of course not.
I am giving you the map, so that whatever comes out at 8:30 does not catch you open-mouthed.
Hot headline on the month, cool core. The consensus scenario. This is the “it is all fuel” case: the Fed can say it is letting that increase pass through, the hike bet deflates a little, yields take a breath and equities get some air. Careful though, it is relief with an expiry date stamped on it, because with Brent up there the same question comes straight back next month.
Hot headline, hot core. The worst one if you hold bonds and equities. It means the barrel is already inside the rest of the basket, exactly the scenario the ECB put in writing for Europe yesterday. Here Wednesday’s hike becomes the main bet, the dot plot shifts up, the 10-year, which yesterday was trading around 4.9% and sits at its highest since late 2023, goes looking for the next threshold and equities pick up the bill.
Headline in line, core below forecast. The best one for equities. The hawks are left without ammunition for a week and the market celebrates. It is also the scenario where it is easiest to fall in love with the move and stay long into Wednesday, which with a dot plot on the way is an elegant way of giving the profits back.
Everything below forecast. The hike prices out, and the old “weak data, kind central banks, buy everything” reflex is back. But watch out: this Fed chair has built the first months of his term on the credibility argument. A good print does not change his problem, it postpones it.
How I handle it
Three practical things, all of them learned the expensive way.
I do not enter on the spike.
I do not trust the first candle at 8:30, and not for psychological reasons so much as technical ones: wider spreads, slippage, prices going one way and coming straight back in the time it takes you to read the headline.
The market reacts to the big number first, then reads the core, and only then really makes up its mind.
In my diary, the move that holds into the close is almost never the first one.
I look at where the walls are before the number is out.
On a day like this, part of the move is not made by the news, it is made by the levels where the market is already loaded with options and market makers have to hedge.
That is not black magic, it is gamma exposure, and knowing where those levels sit changes how you read a spike.
The macro calendar and the gamma on the S&P and the Nasdaq are on the Thunder Desk, they are free, and they spare you the classic scene of a trade left open by accident thirty seconds before the release.
Open it in the morning, not at 8:29.
I cut size before, not after.
If you have a position going into Friday afternoon alive, you do the maths now on what you actually lose if the stop goes with a double spread, with numbers, not by eye.
And I write everything down.
Not the result, the reaction: what I expected, what came out, what price did in the first five minutes and what it did at 10 am.
After three or four macro releases logged in the diary you start noticing that you, personally, have a recurring behaviour on days like these.
It is almost always the same mistake, and it is almost always an expensive one.
Three things to take home
- Friday’s big number and the number that moves the Fed are two different numbers. Monthly headline expected to accelerate hard on energy while the annual one falls. Core expected to sit still. It is the distance between the two that tells you whether this inflation is a barrel of oil or something deeper.
- The ECB showed you the criterion yesterday. It hiked looking not at the headline but at a core that in its own projections stays above the headline in 2027 and 2028. When a central bank fears second-round effects, the hike arrives even in the middle of a war and with growth at 0.9%.
- Wednesday brings the dot plot, and the hike is already 70% priced. An inflation print landing five days before a decision that is almost made but not signed is worth double the usual. You do not need to guess it. You need to not be surprised by it.
If you want the wider picture, the piece on what was at stake at the Fed at the end of July and the one on what a strong dollar does to your trades still hold up, the context has changed less than it looks.
The headline tells you what fuel costs. The core tells you whether the problem is the fuel or you.
Let me close with the thing I repeat to myself every time there is a big release on the calendar.
My job on Friday at 8:30 is not to guess the number.
It is to know in advance what I do in each of the cases, and to have already decided which of those cases keeps me out of the market.
The rest is noise.
Suerte Amigo!
Tiziano Brunno
Tradingblog
Want to put your market reading to the test?
Compete and challenge other traders inside an Arena in a demo environment, with no real capital at risk. Discover the Performance Arena Events by The Thunder Trader.
Disclaimer: purely informational and educational content. It does not constitute financial advice or an invitation to trade. Trading involves the risk of capital loss.
