Crude at 91.777 will not stall this Fed, and the market that thinks it might is asking the wrong question. The interesting number this week is not the 9.09% move in WTI. It is that the core CPI index rose 0.724 points in July against 0.245 for all items, which is to say the underlying series was accelerating relative to headline before oil did anything at all. The supply shock is arriving on top of a core impulse, not instead of one.

The obvious read is that a headline inflation impulse forces a pause. I do not believe it. The effective funds rate sits at 3.63 against a 10y breakeven of 2.35, payrolls added 162 thousand in August, high yield OAS is 2.65, and the Fed's balance sheet expanded again in the week to 2026-09-02. Nothing in that describes an economy that requires easing, and the easing is happening regardless. A committee that cuts into tight credit and positive payrolls is not going to be deterred by gasoline prices it can categorise as transitory. It has the vocabulary already. It used it in 2021 and never retracted it in a way that changed the reaction function.

So the 2y at 4.34, falling, is probably right about the Fed. It is wrong about everything else. The honest price on this board is the 30y at 5.25, which sits further above the 10y at 4.77 than the 10y sits above the 2y. That shape is the market saying policy will be loose and the consequence will be paid at the long end. I would fade the front-end rally and own the steepener.

The supply side deserves better than the scarcity story usually told about it. US LNG exports rose 23% in the first half of 2026, Permian wells are getting longer, eight liquids pipelines have been completed since the start of 2025, and Dangote is pushing Nigerian product volumes. The barrels are being found and moved. What is thin is not production, it is the buffer, and a 4,450 MBBL draw excluding the SPR to 424,460 with gasoline down 1,173 tells you the cushion absorbing any disruption is smaller than the price of Brent at 97.88 suggests. Distillate built 796, which is the only comfortable line in the release.

Breakevens at 2.35 are too low for a world where the policy rate is being cut, the balance sheet is growing, energy is repricing, and core is running ahead of headline. The gap closes through the long end, not the two-year.

Yen at 156.254 is the same trade wearing different clothes. Dollar liquidity is expanding and a broad dollar index at 118.7479 has not yet registered it.

The crude move is the least informative number on the screen. WTI at 91.777 and Brent at 97.88 after a week up roughly nine and eight percent looks like an inflation event and is mostly a refining event. Crude stocks excluding the SPR drew 4,450 MBBL, gasoline drew 1,173, and distillates built 796. That is a picture of refiners running hard and capturing margin, which is exactly what the EIA said when it pointed at elevated crack spreads as the driver of pump prices. A crack spread is a claim on the barrel, not a shortage of it. And the supply pipeline behind it is filling out: eight petroleum liquids pipeline projects completed since the start of 2025, longer wells lifting Permian output, Dangote pulling more product out of Nigeria, LNG exports up 23 percent in the first half of the year. This is not a scarcity regime. It is a bottleneck with a queue of new capacity behind it.

So the answer to whether the Fed should stall at 3.63 because gasoline got more expensive is no, and it is not a close call. The households who pay the pump tax are the same households who pay the unemployment tax if the Fed decides to answer a refining margin with a policy rate. You cannot drill with a funds rate and you cannot crack a barrel with one either. Whoever wins from crude at 91.777 is not the marginal worker, and the fix for a regressive price shock is not a second regressive shock stacked on top.

The bond market is halfway to agreeing. Breakevens sat at 2.35 and did not move. The 2y fell to 4.34. But the 30y is at 5.25 and the curve is at 0.41, which says the long end is pricing something the front end is not, and it is not oil.

What it might be pricing is the July CPI print, which is the datapoint actually worth arguing about. Core rose 0.724 index points against headline's 0.245. Core running that far above headline is not an energy story and it will not fade when cracks normalise. It was firming before the barrel moved. Payrolls added 162 thousand in August, high yield OAS is at 2.65, financial conditions are not tight in any way a credit desk would recognise. The case that underlying inflation is finished does not survive contact with that combination.

Which leaves a policy that most people will get backwards. The right move is to look straight through the crude spike and to take the core print seriously, and the temptation will be to do the opposite: to react to the headline everyone can see at the pump and to treat a hot core reading as noise because it arrived in a month when energy explained the story. If the Fed stalls here, let it stall on core. Stalling on oil would be the error, and it would be paid for by people who do not set crack spreads.

Crude up nine percent on the week and the front end rallied. That is the whole story, and the inflation framing is the wrong one.

Look at where the curve sits relative to policy. Effective funds at 3.63. Two-year at 4.34, down five basis points on the day. Ten-year 4.77, thirty-year 5.25. Every point on the curve trades above the overnight rate. A market that expects the Fed to cut does not price that shape. The "oil spike stalls the Fed" concern assumes there is an easing path to stall. There isn't one in the price. The stall is already in, and has been.

Breakevens confirm it: 10y at 2.35, unchanged. Gold down 0.96% to 4429.825, silver down 1.15%. If this were the front end of a headline inflation impulse, the metals would not be the worst thing on the board the same session. High yield OAS at 2.65, a basis point tighter. Credit sees a supply event, not a macro one.

The inventory mix says the same. Crude drew 4,450 MBBL to 424,460. Gasoline drew 1,173. Distillate built 796. That is a refining margin story, which is exactly what the EIA is describing when it points at crack spreads. Product tightness at the pump is a tax on the consumer, not a wage-price mechanism. Real yields do the tightening for you.

The relationship that actually broke this week is not in the oil complex. Japan imports every barrel it burns. Crude up nine percent should be a terms-of-trade hit to the yen. Instead USD/JPY is down 2.40% over five sessions, EUR/JPY down 2.14%, GBP/JPY down 2.52%. The yen is the strongest thing on the board into a crude rally that should be punishing it. That is not a fundamental move. That is a position coming off, and the cross-yen leading the spot pair tells you which one.

Meanwhile the petro currencies barely showed up. USD/CAD down 0.49% on the week, USD/NOK down 0.77%, USD/MXN down 0.86%, against a nine percent move in the underlying. Broad dollar index at 118.7479, higher. FX did not treat this as a demand impulse either.

I do not have the crude curve here, and without it I cannot separate a genuine inventory squeeze from a spot bid on one week of draws. That matters and I am not going to pretend around it.

Where the pain is: short front-end. If the crude move does not transmit, and the mix of drawn gasoline against built distillate says it partly won't, then the growth cost lands before the inflation does, and the two-year has a long way down from 4.34 before it prices a single cut. Ten-year-two-year at 0.41 and flattening is the early version of that trade. The barrel is not the risk. The position in the barrel's second-order effects is.