The short version: Wednesday's PCE report settled the inflation question the wrong way — headline surprised hot at 3.7% and core spent a second straight month parked at 3.3%, which is a plateau, not progress. The market's answer was to raise September hike odds toward 40% while pricing a cut at almost nothing. Meanwhile the long end of the bond market is writing its own policy — the 30-year touched levels last seen 19 years ago — and the oil market quietly deleted the entire Iran sanctions premium it built just one week ago, even as the confrontation escalated. That last part is the one to sit with. The fuse did not go out. The market just stopped listening to it tick.
Fri Aug 28, ~10am ET: Kevin Warsh's Jackson Hole keynote — his first as Chair, on a stated theme of financial innovation and payments, not the policy path. Early Sep: one more full round of jobs and inflation data. Wed Sep 16: the FOMC decision that futures still price as a genuine hold-or-hike coin weighted about 60/40.
Two Months at 3.3% Is a Message, Not a Pause
The July PCE report landed Wednesday morning: headline up 0.2% on the month and 3.7% on the year — above the 3.6% consensus. Core, the number this Committee actually steers by, rose 0.2% and held at 3.3% year over year — exactly in line, and exactly where June left it. Step back and the trend is the story: core ran 3.4% in May, 3.3% in June, 3.3% in July. That is not acceleration, and it is not disinflation. It is a plateau sitting a full 1.3 points above the Fed's 2% target, and plateaus at 3.3% do not earn rate cuts.
The spending detail matters too. Personal income rose 0.4% and spending rose 0.2%, but the composition flipped: services spending added $86.2 billion while goods spending fell $49.9 billion. The American consumer has not stopped — they have rotated. That mix keeps the services-inflation engine warm, which is precisely the part of the basket a central bank finds hardest to cool without breaking something.
A Committee Priced for a Hike, Chaired by a Man Who Won't Tell You
Context for anyone catching up: the fed funds target has sat at 3.50%–3.75% for five straight meetings, and the July hold came on a 9-3 vote in which all three dissents — Hammack, Kashkari, Logan — wanted a hike, the most dissents in one direction since September 2016. The Committee's own year-end projections run 3.6%–4.1%: a range that leaves the door open to tightening and promises nothing about easing. After Wednesday's hot headline print, futures moved the September 16 odds to roughly 40% hike, ~60% hold, and a cut priced at almost zero. Read that again if your portfolio is still positioned for the 2024 playbook: the live debate at this Fed is hold versus hike.
Tomorrow morning, Kevin Warsh steps to the Jackson Hole podium for his first keynote as Chair. The honest preview is that his pattern since May has been short, deliberately non-committal statements — a chair who wants markets to do the tightening rather than his forward guidance. Anyone promising you a clear September signal from this speech is selling certainty that is not on offer.
There is a third Fed story running underneath both, and it prices in the bond market rather than the fed funds strip: independence. The Supreme Court ruled 5-4 in June that Governor Lisa Cook could not be removed — narrowly, and without defining what "for cause" actually requires. This month the White House signaled it may try again, with a response deadline that has just passed as we publish. Whatever your politics, a live question mark over who sits on the FOMC is one more reason lenders demand extra yield to hold 30-year promises — which brings us to the long end.
The Long End Is Writing Its Own Policy
The Treasury curve right now: the 2-year near 4.18% — notably above the 3.625% midpoint of the funds range, which is the front end telling you it takes hike risk seriously. The 10-year near 4.65%, off last week's 20-month high of 4.75%. And the 30-year around 5.3% — a 19-year high.
Here is the distinction that matters for a retail portfolio: this is not the inverted, recession-warning curve of 2023. The curve is positively sloped and steepening at the long end, and the drivers are fiscal, not just monetary — soft demand at recent 10- and 30-year auctions, national debt past $40 trillion, deficits running above $2 trillion a year, and a wall of competing corporate bond supply as the AI buildout gets debt-financed. Even the Treasury's buyback program, which sparked a rally two weeks ago, has faded as a support. When long yields rise for those reasons, it is lenders charging more for time itself — term premium — not simply a bet on the Fed. Practical translation: the "safe" long-duration bond in your portfolio is currently the risk asset, and it can lose money even on days the Fed news is friendly.
The Bomb Still Ticks — the Market Took Its Headphones Off
One week ago, a sweeping sanctions package took WTI to a four-week high near $87.50 with Brent through $94 — we wrote it up in real time. As of this week: WTI trades near $81 and Brent near $87. The entire spike has been given back. And note what happened while it was being given back: Washington escalated again — a campaign the administration itself branded an economic "D-Day," with Treasury designating more than 60 entities, individuals, and vessels tied to Tehran's procurement networks. The confrontation is deeper than it was a week ago. The price of its principal transmission channel is lower.
Two readings, and both can be true. The generous one: the market has watched this conflict for months, the Strait of Hormuz has stayed open, other producers have offset lost barrels, and softer demand is doing the rest. The less generous one: attention decay — after enough headlines, each new escalation moves price less, not because risk fell but because repetition numbed it. The June thesis — "a relief rally, not a peace" — has never stopped being the desk's frame. What changed this week is that the market removed a premium it may have to rebuild in hours, not weeks, the moment any response touches shipping rather than rhetoric. A premium that decays on schedule and rebuilds on surprise is the definition of an underpriced tail.
The same data-sensitivity shows in gold: a run to roughly $4,677 — its strongest monthly pace in decades — then a pullback toward $4,594 within a day of the in-line core print. And the second fuse, further from the front page: Russia-Ukraine talks remain suspended, with reporting that Moscow is unlikely to negotiate seriously before 2027. None of this is resolved. All of it is cheap to ignore — until it isn't.
Set the full picture side by side: the S&P 500 printed a record on August 14 and closed near 7,677 this week — records against a 30-year at 19-year highs, core inflation stuck at 3.3%, a Fed priced 40% for a hike, and an active great-power confrontation in the world's most important oil chokepoint. Equities are pricing calm; bonds are pricing stress. They are rarely both right for long, and knowing which divergence you are betting on is the actual job this month.
What the Desk Actually Does With a Week Like This
The standing discipline, on the record: no new size the day before the Warsh keynote — a first speech from a low-guidance chair is a repricing event with no edge for outsiders. No selling volatility just because the war premium faded; that is exactly the premium that gaps back. Duration kept short and deliberate — the long end's fiscal story does not resolve on one Fed meeting. Gold sized so a data-driven whipsaw is survivable in both directions. And the honest admission: nobody on this desk knows what Warsh says tomorrow or what Tehran does next week — the edge is not prediction, it is having written down in advance what gets cut, at what level, before the headline forces the decision. Members get this compressed into The Daily Brief every trading morning — the calendar, the odds, the levels — and the live positioning happens in the private members' channel with timestamps.
Nothing here is a recommendation. The framework: hedge when it is quiet and cheap, not after the bang when it is expensive; treat long-duration bonds as a position, not a pillow, while term premium is the driver; and write the exit before the entry — a geopolitical premium you are holding with no invalidation level is a donation waiting to be collected. What we own and when we cut happens live, in the channel.