CPI Playbook — Wednesday 13:30 London

KenMacro Desk · CPI Playbook

US CPI, Wednesday 13:30 London. Have the plan before the number, not after it.

The print itself is not the opportunity. The opportunity is that most of the market reacts to the headline, and the headline is the least informative part of the release. This playbook is the desk’s framework for reading the reaction: what actually matters in the data, the three ways it can go, and the specific thing that tells you whether the first move is going to hold.

No forecast, and no signal. A prediction that is wrong costs you money and a prediction that is right teaches you nothing. What you get is a decision structure that works whichever way it prints.

Free plan to your inbox and instant download. You do not need to fund anything to get the Lite plan. This is a preparation room, not a signal room.

Why core is the number to watch this month

Headline CPI includes food and energy. Both are volatile, and energy has been especially volatile recently — which means the headline can move sharply for reasons that say nothing about the underlying trend in prices.

Core strips both out. That is not a technicality: it is the difference between a number that reflects a temporary swing in the oil price and a number that reflects whether inflation is genuinely cooling. When energy has been moving hard, the headline tells you about energy and the core tells you about inflation. The Fed reacts to the second one.

So on Wednesday, the first place to look is not the big number on the screen. It is core, month on month, and whether the annualised trend is turning.

The three readings that matter, in order

1. Core, month on month. The cleanest read on the current trend. Annualise it mentally rather than waiting for the year-on-year figure to catch up — year-on-year is a rear-view mirror.

2. Shelter. The single largest component of core, and the slowest to turn. A cooling core that is being held up by shelter is a different story from one that is broad-based.

3. Core services excluding shelter. The part most tied to wages. If this is cooling, the disinflation has reached the stickiest place it needs to reach.

What just happened, and why Wednesday matters

Three things landed in the past two weeks. Each one on its own would be worth watching. Together they point the same way, and that is what makes this print worth preparing for properly.

1. The jobs report came in badly

Friday 7 August. Payrolls fell 23,000. The market had been looking for a gain of around 80,000. June was revised down to just 20,000. Strip out government, which shed 53,000 jobs, and private payrolls managed +30,000.

Unemployment did tick down to 4.1% from 4.2%, but not for a good reason. It fell because fewer people are counted as looking for work. A falling unemployment rate driven by people leaving the workforce is not a strong labour market. It is a shrinking one.

The number that matters most is the one nobody led with. Average hourly earnings rose 2 cents to $37.62. That drops the annual pace to 3.2%, against 3.5% expected, and it is the slowest since May 2021.

Wages feed straight into services inflation. When wage growth slows, the stickiest part of inflation loses its fuel. That is the link between a weak jobs report and Wednesday’s core number.

2. Oil has come down hard

Brent has dropped below $80 for the first time in nearly a month. WTI fell 5.43% to $75.98. Before that, when Iran signalled it would hold off while the US pause held, Brent fell 8.7% in a session to $88.36 and WTI dropped 7.5% to $82.61.

So the risk premium that had been sitting in the oil price is being taken out. Lower oil is disinflationary, and it works through prices quickly. It shows up at the pump, in freight, in anything that has to be moved.

One caveat worth keeping. The market has not fully priced out the chance of this flaring up again, and it is right not to. This is a de-escalation, not a resolution.

3. The last CPI already showed the split

June came in with headline at 3.5% on the year while core sat at 2.6%, down from 2.9% in May and below what was expected. Core was flat on the month. Energy fell 5.7% and was the single biggest reason the headline dropped 0.4% on the month.

That gap between 3.5% and 2.6% is almost entirely energy. Hold that thought, because it is about to work in the other direction.

The trap in Wednesday’s headline

Oil moved up sharply through July. Wednesday’s release covers July. So there is a real chance the headline number comes in firm, or even jumps, purely because energy rebounded in the month being measured.

If that happens, the first wave of reaction will treat a hot headline as a hot inflation print. It is not the same thing. Energy moving the headline around tells you about oil. It tells you very little about whether underlying prices are still cooling.

Core is where the answer is. With wage growth at its slowest in over four years and the labour market shedding jobs, the case for core continuing to cool is stronger than it has been in a while. If core confirms that, the headline is noise.

This is the whole reason to have a plan before 13:30 rather than react at 13:31.

What the rates market is actually pricing right now

Before the scenarios, the starting point. Current target rate 3.50–3.75%. These are CME FedWatch implied probabilities as at 9 August 2026:

16 Sep
no change 57.0%
hike 43.0%
ease 0.0%
→
28 Oct
no change 42.5%
hike 57.5%
ease 0.0%
→
9 Dec
no change 21.4%
hike 78.6%
ease 0.0%

Read that again: zero probability of easing at any of the three meetings. The market is not debating when cuts arrive. It is debating how quickly hikes do. By December the pricing is 78.6% for at least one hike, with 28.7% on 400–425 and 5.4% on 425–450.

This is the single most important thing to understand before Wednesday, and it is the opposite of the framing most commentary is still using. A cooling core print does not price in cuts from here. It prices hikes out. That is a different trade, with a different size, and a different invalidation.

Three scenarios, and what each one actually does

Core comes in cooler than expected

Mechanism: the 43% / 57.5% / 78.6% hike pricing above starts getting unwound. Nominal yields fall as that hiking premium comes out of the curve, and because inflation expectations fall by less than nominal yields do, real yields fall. That is the transmission that matters — and note it happens without a single cut being priced.

What tends to follow: pressure on the dollar; support for gold and silver, which pay no yield and therefore benefit most when the real return on cash falls; support for rate-sensitive equities. With hike pricing this elevated, there is more premium available to unwind than in a normal cycle — which is precisely why the move can be larger than the print seems to justify.

What would invalidate it: a cool print where the cooling is entirely one volatile component, or where core services excluding shelter is still firm. The first move happens anyway; it just does not last.

Core comes in hotter than expected

Mechanism: hike pricing extends. December is already at 78.6%; a hot core pulls that forward into October and puts the higher buckets in play. Nominal yields rise, real yields rise with them.

What tends to follow: a bid for the dollar; pressure on gold; pressure on the parts of the equity market that need cheap money.

What would invalidate it: a hot headline driven by energy while core is unchanged. That is a headline reaction, not a policy reaction, and it usually fades once the internals are read.

In line

The most common outcome and the one most people handle worst. When the number matches consensus, the reaction is decided by the internals and by the revisions to prior months — not by the print. This is where reading the release rather than the headline is worth the most.

What decides whether the first move holds

The initial spike is liquidity, not conviction. Three things separate a move that continues from one that reverses inside the hour.

Does the rates market agree? If the dollar is moving but front-end yields are not, the move has no engine behind it.

Are real yields confirming? Nominal yields alone are not enough. Gold in particular tracks real yields, not nominal ones, and this is the single most common mistake made on release day.

Did the revisions change the trend? A revision to previous months can quietly matter more than the current print, and it is routinely ignored in the first ten minutes.

Two things not to do at 13:30

Do not trade the first candle. Spreads widen, slippage is worst exactly then, and the algorithmic reaction to the headline frequently reverses once the internals are parsed.

Do not size as though this print is knowable. It is not. The edge on release day is in having a plan for each outcome, not in guessing which one arrives.

Where this leaves you

Jobs are weakening. Wages are growing at the slowest pace since May 2021. Oil has given back a chunk of its risk premium. Core was already cooling before any of that fed through.

And yet the rates market is pricing no cuts at all, and a 78.6% chance of a hike by December.

Those two pictures cannot both be right for long. Wednesday is the next piece of evidence. If core cools again, a lot of hiking premium has to come out of the curve, and that repricing is where the move is.

The playbook covers how the jobs data, oil and the release interact, and which combination gives the cleanest setup.

Mentorship
Waiting list for the next intake

Ken works with a small number of traders one to one on exactly this — reading the macro regime, understanding how it transmits into price, and sizing risk around it. Intakes are limited and run in cohorts.

The waiting list for the next sessions is open. Applications are reviewed individually, and being on the list is not a commitment.

Join the waiting list for the next intake

Source. Rate probabilities are CME FedWatch implied probabilities, read on 9 August 2026, and move continuously. Check them again before the release.

Risk. This page and the accompanying playbook are educational research, not investment advice and not a recommendation to buy or sell anything. Trading leveraged products carries a high risk of losing money rapidly. Nothing here is a forecast of any economic release or of any market’s response to one. Past performance is not a reliable indicator of future results. Consider taking independent advice if you are unsure.