macro
research.

June payrolls added just 57,000 against a consensus of 110,000, the first miss after three straight months of beating, and the labour crack the desk named as the one risk to the floor. But hiring cracked while wages accelerated, and a cut forced into four-percent inflation is a stagflationary cut, bearish the long end, not bullish. The crack arrives. The floor holds.

May PCE printed a four-handle headline and the hottest core since 2023. The Fed's own preferred gauge has now confirmed what the desk called in May: this is a services-led pipeline, not an energy spike, and it pins the Fed against the floor. The cut case is gone.

The shorthand: the Fed is running the brake and the accelerator at once. The brake is the rate hand, hawkish dots and a guidance withdrawal that lifts term premium. The accelerator is the capital hand, a leverage relief already in force and a Basel re-proposal that cuts rather than raises capital, freeing a domestic bank bid and a wave of balance-sheet capacity. Banks are rising into higher rates because equity already prices it. The honest limit: the returning bid skews short, so it eases the front end, the plumbing and credit more than the 30-year. The floor still needs term premium to turn. Same floor, new cross-current.

The shorthand: forward guidance was a term-premium suppressant, and the new chair is removing it. The 30-year sits near 4.94%, the front end anchored near 4.21%, the curve steep at +70bp, and the 10-year breakeven still near 2.21%. Firm long yields with flat breakevens is a premium story, not an inflation one. The floor under long rates does not fall when the Fed goes quiet. It gains a new leg. Same floor, higher premium.

The Iran ceasefire is signed and Brent has fallen into the high 70s, yet the 30-year has not followed it down. The financial press is now asking why. The desk answered it before the question was asked: the floor under the long end was never the war. It is the bid, and this week the bid got thinner while the deficit got heavier.

The shorthand: a credible ceasefire framework arrived and the 30-year held 4.97%, higher on the week, not lower. The Bank of Japan lifted its policy rate to 1.00%, a 31-year high, and cut its bond buying. Credit compressed (HY OAS 2.71) rather than widened, so this is not a fear trade. USD/JPY barely moved at 160, so the carry unwind is a slow burn, not yet a shock. The war premium leaving could not lift the long end off the floor, because the floor is the vanishing buyer, not the conflict. Warsh meets that floor tomorrow.

On Wednesday, Kevin Warsh chairs his first FOMC with headline inflation at its fastest since 2023 and a hot producer pipeline already on the board. On Friday, the US and Iran are set to sign in Geneva. The market is asking what the new Chair will do. The desk's answer: he inherits a Fed that cannot cut and will not hike at its debut, and the thing actually moving the long end is not in Washington. It is in Switzerland.

On the same day, the US pipeline ran hot beyond energy and the ECB hiked for the first time since 2023, justifying it on a broadening of price rises. The euro was sold on the hike. The 30-year held its third test in a week. Inflation is spreading past the oil story on both sides of the Atlantic, and the dollar is the relative winner.

The shorthand: core m/m 0.2% (soft, vs 0.3% consensus), headline y/y 4.2% (the base-effect re-acceleration, dead on), the 2-year unchanged at 4.14%, the 30-year holding 5.01% above the floor, gold off 2.4% on higher real yields, the dollar firm, the VIX higher, and an AI complex that was already selling into the print and that a dovish core could not rescue. A soft core is necessary for a dovish turn. It is nowhere near sufficient.

Tomorrow's May CPI is the second test of the rate floor in a week. The monthly number is set to cool while the annual rate re-accelerates, and that split is the trap. Here are the three paths the print can take, the levels that decide which one lands, and why the 30-year may hold above 5.00% either way.

The shorthand: Nasdaq -4.2% Friday with a record point drop, S&P 500 -2.65% for a ~$1.8T market-cap wipeout, SOX -8.8%, VIX +39%, HY OAS 2.76 close (from 2.74 prior; credit refused to widen), Alphabet announcing a ~$80bn equity raise to help fund $180-190bn of 2026 capex, Anthropic confidentially filing for IPO at a ~$965bn last private mark, and the mainstream reporting cycle picking up the mechanism: model routing. The era of one model at premium price is being tested. The IPO valuations resting on that era are the cleanest tell on which way it breaks.

May NFP printed 172,000 against an 80,000 consensus. The 2-year repriced 21 basis points intraday and the 30-year is now above 5.00%. The Fed reaction function the April minutes ratified just took the strongest test the data could give it. The floor held.