Cheap crude here is a refining story, not a price-level story, and treating it as cover for easing mistakes where the barrel actually went.

Look at the inventory line rather than the futures screen. Crude excluding the SPR built by 95.0 MBBL against a stock of 428,910. That is a rounding error. Gasoline drew 2,536 and distillate drew 2,228. Refiners pulled crude through and the product left the tank. Nothing in that pattern describes demand rolling over. What fell is the price of the input, while the price of the thing people actually buy held its ground.

The honest reading is supply, and I will give it its due because I have spent years underweighting exactly this. The EIA is describing longer Permian laterals, eight completed liquids pipeline projects since the start of 2025, more waterborne export capacity out of the Gulf Coast, Dangote pulling shipments out of Nigeria, and record domestic natural gas production in 2026. That is not a cycle. That is capacity arriving, and it arrived while everyone including me was narrating a decade of underinvestment. Brent at 90.876 and WTI at 84.065 after a week down roughly four percent is the market marking that to reality.

What it is not is a reason to cut. The July CPI moved the headline index by 0.245 and core by 0.724. Energy was doing the work at the top line while the part of the basket a central bank can plausibly influence went the other way. A committee that eases on Brent is easing on the component furthest from its control, and it will have to explain itself when the gasoline draw eventually shows up in the pump.

Credit where the facts demand it. The two-year at 4.19 sits well above the effective funds rate at 3.63, which is not the term structure of an easing cycle. The balance sheet fell again, to 6,730,912 million. On both counts the line is being held more firmly than I would have forecast a year ago, and I am not going to explain that away.

The disagreement is at the long end. Thirty-year at 5.18, ten-year at 4.66 against a breakeven of 2.33, and a ten-two spread of only 0.47. Meanwhile gold at 4,590.645 and silver at 69.1425 are not the prices of a world that expects real rates to stay where they are, and high yield at 2.67 says financial conditions are loose whatever the policy rate reads.

So: the next move is a cut, energy will be the stated justification, and the front end will get it while the thirty-year declines to follow. Fade the disinflation trade in the belly. The steepener is the position, and cheap crude is what pays for the entry.

Check the products before you write the disinflation note. Crude fell hard on the week, Brent to 90.876 and WTI to 84.065, but gasoline stocks drew 2536.0 MBBL and distillate drew 2228.0 MBBL while crude excluding SPR built a rounding error of 95.0 MBBL. That is not a market telling you demand has rolled over. That is a market telling you there is more oil than there is capacity to turn it into the things people actually buy. The barrel got cheaper. The gallon has to get through a refinery first, and the refinery is where the tightness sits.

So the pass-through to the household is slower and smaller than the futures screen implies, and anyone building a September cut on falling Brent is standing on the wrong plank.

The supply story itself is the more interesting one, and it settles an old argument. EIA has the US on track for record natural gas production this year, longer Permian wells lifting crude and gas output, eight liquids pipelines finished since the start of 2025, and Dangote pulling shipments out of Nigeria. Every one of those is a capital cycle that was dismissed as impossible during the last spike, when the permanent-shortage crowd had the floor. Prices are cyclical more often than structural. This is what that looks like when it resolves, and gas up 4.20% on the week while crude falls is exactly the sort of divergence that a single-narrative energy story cannot hold.

Now the part I have to say plainly, because it cuts against my own instinct. The last CPI print had core rising by more than the headline index. Energy is not where the remaining inflation lives, which means cheap crude is not an argument for easing, it is an argument about a component that was already going to behave. Doves who reach for it are borrowing a case they do not need and will not be able to defend in three months if core keeps grinding.

The case that does hold is the labor one. Payrolls fell on the month while the unemployment rate fell a tenth to 4.1%. Those two facts only reconcile through people leaving the count. A falling unemployment rate produced by shrinking participation is not a tight labor market, it is a discouraged one, and it is the low end of the wage distribution that exits first. Fed funds effective sits at 3.63 against a 2y at 4.19. The market is not pricing the deterioration it can already see in the establishment survey.

Ignore the long end as an objection. The 30y at 5.18 and the 10y at 4.66 against breakevens of 2.33 is a term premium and fiscal argument, not an inflation one, and high yield at 2.67 says credit is fine.

Cut in September. Not because oil is cheap. Because the employment mandate is not the junior mandate and the July data already broke.

Brent lost -4.56% on the week and US 10y breakevens moved +0.010. That is the whole story. If the crude selloff carried disinflationary information, the inflation market would have taken it. It did not, and the front end did not either: 2y at 4.19, effective funds at 3.63, 10s2s at 0.47 and unchanged on the day. A curve with the 2y that far above the funds rate is not a curve waiting to be talked into cuts by a lower oil print.

The supply-side read is right and the conclusion drawn from it is wrong. Crude stocks excluding SPR built +95.0 MBBL, which is rounding error, while gasoline drew -2536.0 and distillate -2228.0. That is not a market with slack in it. That is crude abundance sitting upstream of a tight product complex. The barrel is cheap; the refined molecule the consumer actually buys is not. Add natural gas up +4.20% at 2.906 into EIA's record production year, and the claim that energy is doing the Fed's work for it gets thin. Gas rising against record supply is a demand signal. Products drawing over two million barrels each while crude flatlines is a throughput signal. Neither is disinflation.

The core-versus-headline split makes it worse for the doves. Core CPI rose +0.724 on the index against +0.245 for all items. The energy complex was already suppressing the headline before this week's move. Take another leg out of Brent and you widen that gap without touching the part of the basket the Committee cares about. Unemployment at 4.1, down -0.100, with payrolls at 158858 thousand after a -23.000 revision. High yield OAS at 2.67, three basis points tighter. Credit is not pricing a demand accident, and credit is usually early to that.

So the pain trade is anyone long the front end or long steepeners on the theory that cheap crude buys a dovish Q4. It buys nothing. The 30y at 5.18 sits further above the 10y than the 10y sits above the 2y, which is a term premium story, not a policy story, and lower oil does not compress term premium.

One thing I will not force into this frame. Gold at 4590.645 with breakevens at 2.33, HY at 2.67, and the broad dollar at 118.0628 does not map to anything in my framework. It is not an inflation hedge at those breakevens and not a credit hedge at those spreads. Silver at 69.1425 alongside it. Either the metals complex is pricing something the rest of the curve has not seen, or the relationship between gold and real rates has stopped holding. I cannot tell which from here, and I would rather say that than invent a positioning story for it.

Everything else moved less than a tenth of a percent on the day. USD/JPY at 159.344, vol nowhere. That is the crowded part.