On 23 July, Brent spot traded as high as $105 a barrel, and fed funds futures put the odds of a September rate hike at roughly 82%. On Tuesday, the same barrel settled at $88.58, down 3.9% on the day and more than 7% since Friday. The commodity that justified the hike is about 16% cheaper than it was at the peak. The hike is not. It has gotten more expensive.
That is not a mood reading. It is arithmetic. On 17 August, futures priced a September increase at 30.6%. On 25 August, they priced it at 36.6%, and prediction markets put it a little above even money. But the cleanest number is not in any probability tool at all. It is on the tape: the two-year Treasury note yielded 4.195% on Tuesday, down from 4.236% the session before. The federal funds target range is 3.50% to 3.75%.
The number nobody is quoting: 57 basis points
Work it through. The midpoint of the target range is 3.625%. The two-year yields 4.195%. The gap is 57 basis points. Measured against the top of the range rather than the middle, it is still 44.5 basis points.
A two-year note is, in essence, the average of every expected overnight rate between now and 2028, plus a modest term premium. When that average sits 57 basis points above today’s policy rate, the bond market is not saying a hike is possible. It is saying that across twenty-four months, the average policy rate clears today’s by more than two quarter-point steps — and that is a net figure, after subtracting every cut that might arrive in the same window. For a curve like that to make sense, you either hike early and more than once, or you hike once and then sit still for a very long time.
This is why the two-year is a better instrument than a probability readout. FedWatch-style percentages describe a single meeting and lurch with every data release. The two-year compresses two years of monetary policy into one price that real money has to defend. And that price barely blinked while crude fell out of bed.
What actually happened on Monday
The trigger for the slide was a sanctions package that was announced as an escalation and read as an all-clear. On Monday the Treasury rolled out a globally framed sanctions round aimed at Iran — with no deadline, no defined targeting parameters and, decisively, without touching the institutions that actually clear Iranian oil money. Treasury Secretary Scott Bessent instead offered trading partners an opportunity to fall in line with the new directive. Strategists at Brown Brothers Harriman called it more of a warning shot than a decisive blow.
Traders drew one conclusion from that: a government announcing economic coercion is not a government about to escalate militarily. Ole Hansen, head of commodity strategy at Saxo Bank, described the shift from military confrontation to economic pressure as something that helped calm nerves across energy markets. Meanwhile a Pakistani field marshal spent the day in Tehran reportedly carrying a proposal on sanctions relief, and Qatar said its mediation was continuing. Brent fell harder than WTI, which is precisely the signature of a Strait of Hormuz premium being taken out of the price — that premium attaches to the seaborne benchmark, not to the American inland grade.
The physical market disagrees with the price
Here is where a market commentary has to part ways with a news wire. On the same Monday that the price fell more than 3%, two vessels transited the Strait of Hormuz. Two. That is the lowest count since early May. The International Energy Agency now projects a global supply deficit of 1.8 million barrels a day for the third quarter of 2026 — more than double its previous estimate.
So the price fell while physical availability got worse. That is not a contradiction; it is a precise statement about what is being traded. For the past several weeks the oil market has not been pricing the barrels moving through the strait. It has been pricing the probability distribution of what might happen to the strait. Tim Waterer of KCM Trade named the residual: Iran retains the ability to disrupt shipping, which is why a residual premium stays in the oil price. What disappeared on Monday was not supply. It was the odds attached to the tail.
To a central banker, that distinction is everything. A committee cannot eat a risk premium. It has to decide whether the inflation it measures has broadened — and a falling war premium does not answer that question in either direction.
Why it was never the oil
Which is the resolution. The story that crude was driving the rate bet was always a shortcut. The causality ran through an intermediate step, and that step is holding.
July’s producer price report is the cleanest illustration. The headline was flat; the core rose 0.4%. A flat headline with a firm core means the energy component dragged the index down while everything else kept getting more expensive. That is the textbook definition of broadening. Core personal consumption expenditures inflation stands at 3.3%, and inflation has now run above the Fed’s 2% target for more than five years.
Which is why the 29 July vote came in at 9-3 — not close, but conspicuously split. Lorie Logan of Dallas, Beth Hammack of Cleveland and Neel Kashkari of Minneapolis all dissented, and all three dissented in favour of a quarter-point increase. Chair Kevin Warsh, presiding over his first meeting, summed it up at the press conference by saying he had asked for a good family fight and got one. The statement itself was far shorter than what had become the norm.
Three officials voting to hike while crude falls are not voting about crude. Their argument is the duration of the overshoot, not the width of a shipping lane. And that is exactly why the oil slide did not drag the rate bet down with it: the bet was never hanging on the hook the headlines had put it on.
The lag nobody prices
There is a second reason, purely mechanical, that is easy to lose in the noise: even if crude simply stayed where it closed on Tuesday, it could barely rescue the data that matters next.
The July personal consumption expenditures price index lands today at 8:30 a.m. Eastern. It measures July — the month in which Brent briefly printed $105. Consensus looks for core to rise 0.18% on the month after 0.13% in June, and 3.2% year over year after 3.3%. Headline is seen at 3.6% after 3.7%, and up 0.07% on the month after falling 0.11%.
None of Friday’s slide is in those numbers. Very little of it will be in the August figures either, since three-quarters of August was a month of materially higher prints. And even where energy does pass through, the route from the barrel to the core runs via freight, packaging, petrochemicals and services, and it takes quarters rather than days. For the September meeting, the current oil price is close to irrelevant. The bond market knows this. That is the second reason the two-year did not move.
One footnote that will matter in September: the statistical agency is introducing calculation changes designed to capture price dynamics in computer hardware, portfolio management and legal services more accurately. They arrive next month and can revise today’s July figures after the fact. Anyone trading the second decimal this afternoon is trading a number that is not final.
Where this lands in a portfolio
The direct effects are the obvious ones and mostly already in the price. Integrated producers — Exxon Mobil, Chevron, ConocoPhillips, and in Europe Shell and BP — take a 16% move in the benchmark straight through the cash flow statement, with the refining margin partially offsetting at the margin. On the other side sit the fuel buyers: Delta Air Lines, United Airlines and the freight complex, where jet fuel and diesel are the second-largest cost line after labour, though hedging programmes mean the benefit shows up with a delay of one to three quarters rather than immediately.
The more interesting effect is the one running through the two-year. A policy rate that stays at or above 3.625% for two years re-rates everything valued on a discount rate, and it does so unevenly. Long-duration equity — anything whose cash flows sit mostly beyond 2030 — carries the largest sensitivity, which is worth remembering on a day when the single most crowded position in the market reports after the close. Banks, insurers and money-market complexes sit on the other side of that trade: a higher-for-longer front end is net interest income.
For anyone actually deciding what to own, the practical read is narrower than it looks. If the two-year at 4.195% is correct, front-end Treasuries and short-duration credit are being handed a yield that already assumes the hike, which means the coupon is compensation for a risk the buyer no longer has to forecast. If it is wrong — and the labour data below is the reason it might be — the same instrument delivers a capital gain on top of the coupon. That asymmetry is the argument for the front end, and it does not require having a view on Iran at all.
The case against
There is a serious counterargument, and it is not in the oil market. It is in the 7 August labour report. Forecasters expected 83,000 new nonfarm jobs. The economy shed 23,000. Government payrolls fell by 53,000 while private payrolls rose 30,000. Unemployment ticked down to 4.1% — but it fell because fewer people were working or looking for work, not because more were hired. And the number a central bank actually watches: average hourly earnings grew just 3.2% year over year, the slowest since May 2021.
Wage growth of 3.2%, against any ordinary productivity assumption, is consistent with a 2% inflation target. Anyone arguing against the hike has to argue from exactly there: inflation that does not embed itself in wages is inflation that burns itself out. On that reading the two-year at 4.195% is simply too cheap, the bond market is too gloomy, and those 57 basis points are an opportunity rather than a warning.
That also defines the test cleanly. If core comes in meaningfully below consensus this afternoon and wage growth stays below 3.3% in September, then it was the oil after all, and the two-year has to come back down. If core confirms 3.2% or prints above it while Brent stays near $88, the proof is in: two variables that diverge this cleanly were never closely related.
Friday, 10 a.m.
The resolution is closer than it looks. The Kansas City Fed’s symposium runs Thursday through Saturday in Jackson Hole, and Kevin Warsh delivers his first keynote as Chair on Friday at around 10:00 a.m. Eastern. The official theme is financial innovation and its implications for payments and policy — explicitly not inflation. A Bank of America survey shows the market has priced a broadly neutral speech.
Both facts together make the event more dangerous, not less. Under Warsh the Fed no longer telegraphs decisions ahead of meetings, and the meeting statement has been shortened. Where there is no pre-announcement, a major speech carries genuine information value — and when neutrality is already in the price, the reaction function is asymmetric in both directions. A single sentence on Friday can move more than this afternoon’s July data.
What is worth taking from this week is less a forecast than a method. If an explanation for a market move is correct, it has to survive the explanatory variable reversing. Crude reversed, by about 16% from the high. The rate bet stayed — and on three independent gauges it went up. That falsifies the popular explanation and leaves the less comfortable one standing: inflation that has run above target for five years does not get decided by a barrel of oil. It gets decided by everything downstream of it. The next piece of evidence lands this morning at 8:30, and on Friday the man who has to interpret it steps up to a podium in Wyoming.
Try TradingView Free for 30 Days
Plus get a $15 discount on your first subscription through this link.
