The FRED spot WTI series reached $84.25 on July 27 against $70.56 on June 30, a gain of 19.4%. Brent marked $90.75 early Thursday, up 4.04% over 24 hours, and the USO oil fund gained 7.32% to $129.31. Ten-year inflation compensation closed Wednesday at 2.26% against 2.24% on June 30.
Two basis points.
Take the other available measure and it reads worse. Across the matched June 30 to July 28 window the nominal ten-year yield went to 4.61% from 4.44%, while the ten-year inflation-indexed yield went to 2.41% from 2.20%. The breakeven implied by that pair narrowed 4 basis points to 2.20%. One series has compensation almost flat on the month. The other has it tighter, while the crude input gained close to a fifth of its value.
A market that pays two basis points for that has taken a view. I think it is the wrong one, and the arithmetic on the entry is what makes the case rather than the number of ways it can win.
June's cooling was measured with $70 crude in it
The June inflation data were soft in a way that deserves respect. Headline CPI printed 3.53% year over year against 4.25% in May, a drop of 72 basis points. The index level itself fell 0.42% to 332.568 from 333.979, an outright decline rather than a slower climb. That combination is what gave the bond market permission to stop paying for inflation risk.
It was also recorded with WTI at $70.56 on June 30.
WTI now sits between $83.43 and $84.25 depending on which series you take, and Brent is above $90. July and August headline prints face the same arithmetic running the other way, through energy alone, before any judgment about the rest of the basket. The honest limit on that claim is that there is no core CPI, no core PCE and no shelter series in front of me, so whether June's cooling was broad or narrow cannot be checked here. That caps this at a moderate-conviction view rather than a confident one.
The compensation leg sat still because the real leg took everything
The month in the ten-year happened in the inflation-adjusted leg. Its yield added 21 basis points to 2.41%, a 9.5% rate of change, against 17 basis points on the nominal. Real yields therefore did more than all of the work, and the residual compensation went backwards.
That same 2.41% is the reason equity multiples did not expand on improving inflation data. The S&P 500 closed Wednesday at 7,294.6 against 7,482 on July 21, down 2.5% across eight days in which the claims drop and the CPI deceleration were already public. Better data bought nothing.
Gold is the one price that broke the pattern. It rose 1.16% to $4,129.2 over the same eight days, absorbing 6 basis points of real-yield rise from its July 21 marker and 21 from June 30, and gaining ground anyway. That is the first sign in weeks that something other than the real rate is setting the gold price, and an inflation-hedge bid is the most direct explanation available.
The crude move is in the waterborne grade
Brent added 9.4% from its $81.62 July 21 marker. Live WTI at $83.43 sits 1.3% below its own $84.54 marker from the same date. The fresh leg is entirely in the seaborne barrel.
Matched marks from Wednesday at 23:30 put Brent at $89.27 and WTI at $83.43, a spread of 5.84. On July 21 the same spread sat roughly 0.20 below 5.00. My reading is a supply-risk premium on waterborne crude rather than a signal about US demand, which fits the July 7 tanker strike in the Strait of Hormuz region that the news feed scored 7 and classified as a major oil-shock signal. That event is three weeks old now and works as a background condition rather than a catalyst.
The character of the move is what matters for the inflation question. A supply-led rise in crude lifts the energy line without the offsetting weakness elsewhere in the basket that a demand-led rise would bring.
The steepening that cannot be taken apart
The gap between ten- and two-year yields printed 0.45 on Wednesday. July 28's pair, 4.61% and 4.26%, implies 0.35. Ten basis points of steepening in one session.
There is no July 29 observation for either individual leg. That means the move cannot be decomposed, and the two candidate explanations point in opposite directions: a two-year rally would say the front end has begun pricing a cut, while a long-end selloff would say something quite different about how policy is being read. The ten-year spot marker of 4.622% on Wednesday against 4.61% on July 28 leans toward the former, but one basis point across two different sources settles nothing.
So I am not going to tell you which end moved, and neither should anyone else working from these series. The next two-year print against the ten-year answers it. A reading below 4.26% with the ten-year holding near 4.61% would show the front end did the steepening; a ten-year above 4.70% with the two-year flat would put it in the long end.
Nobody is questioning the hold over the next fortnight. Effective fed funds was 3.63% at both the June 30 and July 28 observations. The guidance is a coin flip.
The case against, in full
A long position in inflation compensation loses in half the distribution, and that belongs on the record rather than in a footnote. Weight the disinflation branch at 30, where June turns out to be the trend, energy fails to pass through and compensation narrows as the realized data cool. Add the growth-crack branch at 20, where demand destruction compresses compensation from 2.26% for an entirely different reason. Against them sit energy-led reflation at 35 and an outright energy shock at 15. Fifty against fifty.
One correction on how to hold it. Long inflation-indexed bonds against nominals, duration-matched, is long compensation and nothing else. Falling real yields alone do not pay it. Pairing it with a short in nominal duration and calling the pair balanced is wrong, because both legs are exposed to the same disinflation outcome.
Wednesday's equity tape is the best evidence the other side has. Caterpillar marked $782.71, down 6.91% over 24 hours. Semiconductors came off 4.79% on the SMH basket to $504.22, homebuilders 4.52% to $104.68 on XHB, Nvidia 3.55% to $190.01, and JPMorgan 3.53% to $344.71. VIX rose 7.32% to 19.5. High-yield spreads stood at 2.84% on July 28, wider than 2.69% on July 21 and 2.75% on June 30. Cyclicals and credit deteriorating together while crude gains is what the demand-destruction story looks like on a screen, and it is why the growth branch carries 20 rather than 5.
GDPNow supports that side too. The nowcast reads 1.3% as of July 8, down 56.7% from 3.0% on June 17, and it refreshes today alongside the Q2 advance print. A sub-1.0% advance takes weight straight out of the reflation case.
The labor data disagree, and they are fresher. Initial claims fell 13.8% to 187,000 on July 18 from 217,000 on June 27. Unemployment ticked down to 4.2% in June from 4.3% in May. Financial conditions eased over the same month, the Chicago Fed index to -0.554 on July 24 from -0.510 on June 26, though that observation predates the credit and volatility deterioration by several days and the next release is August 5.
What settles it
Friday's PPI is the most informative release in the next two weeks. Hot on energy components, with compensation moving above 2.35% while the real yield stalls at or below 2.41%, is the confirmation. Cool, with compensation at or below 2.26%, kills the position outright and thins the gold case within days. Today's Personal Income and Outlays release and the Trimmed Mean PCE series come first, and August 12 CPI is where the energy arithmetic either reaches the headline or does not.
A decline in real yields on its own is not this position working. That is the easiest way to misread the next fortnight, and the most expensive.
Explore these indicators together: Chart WTI Crude Oil, Brent Crude Oil, and 3 more on the Indicators Dashboard