August core PCE lands soft and cuts October hike odds below 40% — yet the 10-year pushes past its 2007 peak to a 24-year high

0. Weekly Arc

The week’s arc ran from hawkish accumulation into a dovish data shock the bond market refused to honour. It began with a fresh speaker chorus and a boom-time September flash PMI that put October hike odds near 70% and the 30-year above 5.5%; it turned on New York Fed President John Williams’s “no need for urgency,” which cut October odds toward a coin flip; and it ended with a soft core PCE print plus a retroactive methodology revision that took October odds below 40% — while long yields instead set fresh multi-decade highs.

1. Policy Narrative & Expectations

No new policy action in the past ~24h — the target range stands at 3.75%–4.00% after September’s first hike in three years, with 16 of the 18 forecasters who submit projections still expecting at least one more 2026 move and the year-end median raised to 4.1% ^[1][2] — so the marginal information sits in four places. First, the data finally turned helpful to the doves: the Fed’s preferred gauge rose less than expected in August on core and in line at the headline, and the BEA’s annual methodology change lowered recent core readings retroactively to 2021 ^[3][4][5][2]. Second, the price of the cycle fell further: CME FedWatch now shows an October hold at 61.8% against a 25bp hike at 38.2% ^[6], with a separate reading at 37.1% after a 13.8 percentage-point one-day drop ^[7], swaps pricing about a 36% chance ^[8], and Business Insider recording the move from roughly 70% at the start of the week ^[9]. Third, the long end ignored all of it — the 10-year traded above 5.3% intraday, past its 2007 peak to the highest since 2002, and the 30-year surpassed 5.6% ^[10][11][12]. Fourth, the sell-side redistributed itself again without converging: JPMorgan keeps an FOMC hawkish bias with the next hike in December ^[5][13]; Barclays maintains a December 25bp hike followed by no change through end-2027, arguing the stronger revised growth profile contradicts the assumption embedded in the Fed’s September projections ^[14][15]; Goldman pushed its second-hike forecast to December and said no further hikes may ultimately be needed ^[4]; Citi’s baseline is that the Fed holds for the rest of this year and through 2027 ^[3]; and Jefferies says it is more confident central banks will not deliver the magnitude of hikes markets price ^[16]. The net change is a genuine dovish repricing of timing with unchanged December conviction, and an unusually explicit split between the front end (pricing a pause) and the long end (pricing a structurally higher cost of capital).

1.1 FOMC Officials’ Remarks

• [NEW] Hawkish — Minneapolis Fed President Neel Kashkari: the session’s only substantive new speaker and the day’s clearest hawkish signal, delivered against a market that had just priced out October ^[17]. Per Gelonghui he said inflation “remains too high, at roughly 3%,” that the new data had not changed his outlook, that the neutral rate may be higher than expected and he is uncertain where it is now, and that “the longer the economy stays strong, the more I question how restrictive monetary policy is” ^[18]. Financial Juice relays add that he warned rolling supply shocks may boost inflation expectations ^[19], that he expects one more hike this year and another in 2027 ^[20], and that consumers continue spending and job seekers have positions ^[21] — single source / unverified social relays. The asymmetry is worth flagging: the same relay stream has him hoping the central bank “can reduce inflation with restrained measures” ^[22] and keen to see the outcomes of the Fed’s task forces ^[23], so his hawkishness is about the destination rather than a call for near-term action.
• [ONGOING] Hawkish — Governor Michael Barr: reiterating that the Fed most likely needs to keep raising rates to bring inflation back to 2% ^[24].
• [ONGOING] Hawkish — Alberto Musalem: maintaining the same hawkish line in the same relay ^[24].
• [EASED] Neutral — Chicago Fed President Austan Goolsbee: moved to a wait-and-see stance, saying he wants clearer evidence that inflation is coming down, and separately noting a record gap between consumer sentiment and actual spending ^[24][25] — a softening from the “playing with fire” framing he used earlier in the week.
• [ONGOING] Mixed/dovish — New York Fed President John Williams: said there was “no need for urgency” after September’s adjustment and that “we have time to gather more information” ^[2], while a Citi relay has him also saying the Fed “still has work to do” and hinting another hike later this year may be appropriate ^[26].
• [ONGOING] Chair Kevin Warsh: still avoiding any hints on timing and describing September’s increase as the Fed removing only a “dose of accommodation,” leaving open both the extent of further tightening he deems necessary and whether it means moves at consecutive meetings ^[2].

1.2 Policy Signals & Institutional Communication

• [EASED] October hike pricing fell further across venues: CME FedWatch holds at 61.8% for a hold versus 38.2% for a hike, with December split 13.3% hold / 56.7% cumulative 25bp / 30% cumulative 50bp ^[6]; another CME reading fell to a 37.1% hike probability from 50.9% the prior day, with at-least-one-more-by-December dropping to 86.8% from 91.6% ^[7]; swaps price about 36% ^[8]; and Business Insider records the round trip from roughly 70% to 37% within a week ^[9].
• [NEW] JPMorgan keeps the hawkish bias: growth resilience plus the Fed’s attention to above-target inflation keep the stance hawkish, with the FOMC expected to preserve that bias and the next hike in December, though mild near-term momentum may let it avoid hiking at every meeting ^[5][13].
• [ESCALATED] Citi’s baseline turns explicitly into a hold: it expects the Fed to leave rates unchanged at the October meeting and for the remainder of this year and through 2027, resting on downward core PCE revisions and sub-0.20% monthly core CPI, plus only moderate job growth of about 85k with unemployment at 4.2% ^[3][26].
• [ESCALATED] Goldman pushes the next hike out: it reads the below-expectation core PCE and larger-than-expected methodological revisions as lowering October hike odds, moving its second-hike forecast to December and holding open that no further hikes are needed; it also marks resilient Q2 GDP and a wider-than-expected goods deficit, trimming its Q3 GDP tracking slightly ^[4].
• [NEW] Barclays turns the GDP revision into a hawkish argument: the annual revision shows the economy grew more strongly in H1 2026 than previously thought, with private domestic final purchases accelerating, which raises doubts about how restrictive policy really is; under Okun’s law, above-trend growth should translate into labour-market tightening, contradicting the assumption embedded in the Fed’s September projections — so Barclays keeps its December 25bp hike and an unchanged stance through end-2027 ^[15][14].
• [NEW] Nomura blames the curve steepening on communication, not data: the UST steepening reflects insufficient Fed communication on the policy path, leaving markets without a clear read on the reaction function and pushing up long-end yields and UST volatility, with real yields at cycle highs on worsening supply-demand and rising term premium ^[27].
• [NEW] Jefferies doubts the hiking path: it is more confident central banks will not deliver the magnitude of hikes markets price, calls current Treasury yield levels attractive for long-term investors, but says confidence on the next 10bp in the 10-year is low, and notes money markets price 93bp of cumulative Fed hikes over the next 12 months ^[16].
• [NEW] The constructive long-end case gets an airing: a Financial Times column by Absolute Strategy Research argues the yield rise reflects stronger real growth rather than inflation risk or policy uncertainty, imposes capital discipline on governments and corporates, reduces financial engineering that had benefited non-bank lenders, and improves long-run optionality for allocators given stretched US equity valuations ^[28].
• [NEW] Fed approves stress-test changes: the Board approved revisions to its annual stress tests by a 6-1 vote, with Vice Chair for Supervision Michelle Bowman saying the reform keeps the framework robust by making tests more transparent, granular and risk-sensitive, and the Fed also sought comment on better reflecting banks’ fee-income business models ^[29][30].
• [NEW] Fed oversight report: Chair Warsh said the Board of Governors will work with Foster to set fixed metrics, budget and timeline for the remaining projects, with Foster reporting progress to the Board and the Chair ^[31].
• [NEW] The next meeting is political-adjacent: the Fed gathers in late October, just days before the midterm elections, with most officials expecting inflation to reach 2% only by 2029 and most expecting rate reductions to be delayed until 2028 ^[2].
• [ONGOING] The delivered decision and projections stand: a 25bp hike to 3.75%–4.00% on Sept 16, a unanimous vote, and a dot plot whose median implies one more move ^[32][1].

2. Key Data & Market Read

• [NEW] Core inflation came in soft — the session’s decisive print: the Fed’s preferred gauge rose less than expected on core and in line at the headline, and prior months were revised down, leaving the three-month annualized core rate at roughly a two-year low and below both the September SEP median and the Chair’s own estimate ^{[3][4][33][34][5]}. The market read was immediate: traders pared October hike bets, short-term rate futures rose and short-dated yields slipped while longer yields held steady ^{[8][35][36][34]}. Two offsets deserve naming: core is still running at or above 3% year-on-year for a seventh consecutive month on JPMorgan’s count ^[5], and a material part of the improvement comes from the BEA’s methodology revision to software, portfolio management and legal services, applied retroactively to 2021 — institutions caution that US inflation remains elevated under either yardstick ^[2][24][13].
• [NEW] Growth was revised up, not down: the annual revision lifted second-quarter growth and first-quarter growth, driven by stronger services consumption, with private domestic final purchases — a gauge the Fed itself watches — at their fastest since early 2023 and household income and the saving rate revised sharply higher ^{[15][14][4][3][33]}. Market read: “the growth story is rocking and rolling,” which is precisely why the softer inflation print did not trigger bond-market celebrations ^[33][37].
• [NEW] September ADP beat: private payrolls rose more than consensus, led by education, health and leisure, with wage growth described as still solid by ADP’s chief economist, and job-stayer pay growth steady ^[4][24].
• [NEW] Consumer: strong in August, fading into September: nominal personal spending jumped and real spending posted its largest gain in more than a year, but JPMorgan’s card data showed controlled retail sales up only marginally in the first half of September, attributed partly to promotion timing ^[24][38][5].
• [NEW] Trade was the offset: the August goods deficit widened well beyond expectations on a surge in industrial-supply and capital-goods imports, which both JPMorgan and Goldman read as a mechanical drag that trims third-quarter GDP upside even as it signals strong investment ^[5][4][3].
• Narrative impact: this is the first print in three weeks that genuinely supports the pause case rather than merely failing to refute the hike case — and the market took it that way, cutting October odds roughly in half. But the data flow itself remains two-sided in a specific way: the inflation surprise is partly a measurement event that leaves core sticky at 3%, while the growth revision cut against the Fed’s own assumption that above-potential growth would not tighten the labour market. That combination is why the front end rallied and the long end did not.

3. Financial-Conditions Signals

• [ESCALATED] Rates: the long end stretched to fresh multi-decade highs — the 10-year above 5.3% intraday, past its 2007 peak to the highest since 2002, and the 30-year above 5.6% — capping a quarter of relentless selling, with the move led by the long end as the curve steepened rather than flattened ^{[10][11][39][40][13][41]}.
• [NEW] The driver decomposition is now contested between real growth and term premium: Nomura shows both the 10-year real yield and the breakeven rising, alongside a terminal-rate proxy rebounding ^[27]; Absolute Strategy Research argues stronger real growth — not inflation risk or policy uncertainty — is doing the work, with real yields simply reflecting the true cost of capital and accelerating AI capex pushing them higher ^[28]; and Goldman’s trading desk frames high real yields as the “clearing price” a fiscal-dominant economy must pay to attract private capital, which in turn lifts the hurdle rate for equities ^[41].
• [NEW] Market functioning, not fundamentals, is where the stress is: the main sellers of long Treasuries this week were hedge funds and REITs holding large MBS positions, as rising borrowing costs cut prepayments, extended MBS durations and forced sales of other long bonds; leveraged funds in the futures market were forced to rebalance; and Citi’s rates desk says marginal buyers have yet to appear ^[32]. Privorotsky adds yen appreciation compressing global balance-sheet capacity, carry-trade unwinding, and extreme rate volatility leaving market makers unwilling or unable to supply liquidity ^[41].
• [ESCALATED] Volatility: MOVE rose to 110.5 and VIX rebounded to 16.3 ^[27], with UBS arguing rate volatility has finally activated and should stay above equity volatility into the October FOMC and the midterms ^[42].
• [ESCALATED] Dollar: the dollar rose to a three-month high against a basket of currencies, with September its best month since March, all G10 currencies down against it except the yen, and the index near its year-to-date high ^[43][44][27].
• [ONGOING] Gold: bullion traded around $4,155 an ounce, steady as softer inflation lowered hike bets, with high bond yields keeping the door open to renewed selling and price action described as unconvincing ^[45][46].
• [NEW] Credit is calm on the surface and complacent underneath: Wellington reports credit markets remained fairly stable with spread widening driven by issuance rather than deteriorating fundamentals, though index valuations are tight with wide internal dispersion ^[47]; UBS warns high-yield spreads are unusually complacent and face a “double hit” of rising yields plus widening spreads, with the correlation between yields and high-yield spreads at a historic high ^[42].
• [NEW] The yield rise is global and, in Europe, sharper: the French-German 10-year spread hit its widest in 14 years and the WSJ frames US anxiety as mild compared with what is happening across the Atlantic ^[11][48][49], while the UK 30-year gilt climbed to a level not seen since 1998 ^[50].
• [ONGOING] Fiscal: US public debt has exceeded $40tn ^[32].

4. Global Central-Bank Linkages

• [ESCALATED] BOJ: Nomura puts the probability of an October hike at 21%, linked to US expectations with limited further upside, and notes the JGB curve is flattening in contrast to Treasuries, the two-year auction was absorbed smoothly, and Japanese market volatility has fallen sharply as the Takaichi government gradually exits its reflation stance — with Japan’s bonds and equities decoupling from their US counterparts ^[27].
• [ONGOING] PBoC: the Sept 29 structural package stands — PSL rate cut, broader PSL scope and larger re-lending quotas — with no move on headline policy rates ^[51][52].
• [NEW] Others: the quarter’s tightening came as a bloc — RBA, ECB, BOJ and Fed all hiked in Q3 after only one G20 central bank moved in Q1 — reinforcing the global tightening impulse ^[32]; and UBS Wealth Management says major central banks have made clear rates will need to remain high near term amid sticky inflation and resilient growth ^[33].

5. Asset Implications

Quadrant	Current probability tilt	Key asset implication	Anchoring narrative
Growth↑ + Inflation↑	Unchanged at elevated levels	Energy and commodities remain the cleanest expression with Brent near $100 and yields highly sensitive to oil, while nominal long bonds stay capped by a 10-year above 5.3%; carry and curve structures still dominate outright duration	§2 / §3
Growth↑ + Inflation↓	Rising, and it is the day’s real gain	The soft core PCE alongside an upward GDP revision is the cleanest version of this quadrant yet, favouring quality cash-flow equity beta and belly carry over long duration; the constraint is that the long end did not confirm it	§2 / §1.2
Growth↓ + Inflation↑	Unchanged (tail)	The forced-selling loop and the MBS duration-extension channel are the mechanism; commodities and gold, not duration, remain the hedge while the household spend channel is already thinning	§3 / §6
Growth↓ + Inflation↓	Rising modestly	October odds below 40%, Citi’s hold-through-2027 baseline and a soft core trend support the tail, but a December hike is still fully priced and ~90bp of tightening over 12 months remains in the curve	§1.2 / §3

Stock-bond correlation call: the regime remains positive correlation — the configuration hardest for risk parity — and today isolated the mechanism more cleanly than any session so far. The key new fact is the asymmetry: an inflation print that missed on the dovish side and dovish-leaning commentary moved the front end and hike odds materially, yet the 10-year still made a new 24-year high and the 30-year cleared 5.6%. That means duration is no longer repricing the variable equities reprice (the expected policy path); it is repricing something that does not mean-revert on a soft print — real rates, term premium, supply, and a mechanically forced seller base with no marginal buyer. The counter-print is equally specific and should be named: if the rise is genuinely driven by stronger real growth rather than inflation risk, it is the one configuration in which equities can absorb higher yields, and Fidelity’s Jurrien Timmer puts numbers on it — a move to a 6% 10-year implies a P/E near 16x versus 19–20x today, which 30% earnings growth could offset without a 2022-style drawdown. Directional call: positive correlation persists, but for a new reason — not inflation compensation, which is not the swing factor, but a term-premium-and-market-structure channel that a dovish Fed cannot fix. The tell remains the same as last week and is now sharper: if a soft core print and dovish commentary cannot rally the long end, bonds are not a hedge.

Risk-budget implication: hold commodity and energy risk at or above benchmark. Energy is still the sleeve whose driver is orthogonal to the Fed’s reaction function, UBS explicitly notes yields are highly sensitive to oil prices, and the same desks that favour the hawkish tilt also name energy as the stagflation hedge — but recognise that the energy-to-rates transmission is itself under strain, with one Goldman desk arguing falling oil no longer relieves bond-market pressure. In rates, do not rebuild long-end duration. The combination is unusually hostile: the long end is deeply oversold yet shows no trend exhaustion, the 10-year sits at the 100th percentile of one major desk’s forecast range, the marginal buyer base has become price-sensitive, and a mechanical forced-selling loop is active — the robust expressions remain belly and curve structures, with any duration funded there. Note the constructive counterweight for patience: Jefferies calls current yield levels attractive for long-term investors and the FT’s constructive case rests on real-growth-driven yields, both of which argue for staged rather than absent duration. In credit, keep duration short and prefer spread over pure duration, but size for the complacency UBS flags: high-yield spreads do not price the historically tight link between rate volatility and spreads, the yield-to-HY-spread correlation is at a historic high, and the recommended structure is explicit convexity rather than cash beta. In equities, hold moderate beta expressed through cash-flow durability, and respect the internal composition: the Nasdaq led while the median stock fell 5% over the past month with the index barely moving, and index-level resilience is concentrated in exactly the sector most exposed to the discount rate — so size equity risk against a rates scenario, not an earnings scenario. Hold gold as the structural fiscal and dollar-credit hedge, sized for a real-rate headwind: $4,155 spot, unconvincing price action, and high yields keeping renewed selling alive, set against a still-intact official-sector and debasement bid.

6. Contrarian & Tail Risks

• Consensus fragility: the baseline is now “the Fed pauses in October, probably still hikes in December, and the long end has repriced for something the Fed does not control.” Five falsifiable pillars hold it up. First, the pricing is a band, not a number — 36% on swaps, 37.1% and 38.2% on two CME vintages, and one desk reading near 50% after hawkish official remarks — so venue and timestamp dominate any quoted level, and the probabilities themselves are derived from futures prices rather than observed intentions. Second, the market and the sell-side cannot both be right: ~90bp of cumulative tightening is priced over 12 months and a December hike is fully priced, while Citi’s baseline is no change through 2027 and Goldman says no further hikes may be needed. Third, the disinflation that triggered the repricing is partly a measurement event — the BEA revision lowered core readings retroactively to 2021, core is still at or above 3% year-on-year for a seventh straight month, and the Fed itself is unlikely to change its view that more work is needed. Fourth, the growth revision cuts against the doves: if above-trend growth translates into labour-market tightening, the Fed’s September assumption is wrong in the hawkish direction. Fifth and most fragile, the entire equity valuation structure rests on the assumption that weak data buys lower discount rates — and today weak inflation data bought nothing at the long end.
• Second-order transmission: four chains are not in market pricing. The forced-selling chain is the most active: rising long yields cut MBS prepayments, extend MBS durations and force REITs and hedge funds to sell other long bonds; leveraged futures accounts rebalance mechanically; and the marginal buyer has not appeared — a loop in which higher yields create their own supply, with one portfolio manager describing it flatly as a vicious cycle. The funding-and-carry chain is the second: yen appreciation compresses global balance-sheet capacity, unwinds carry trades, and leaves market makers unwilling or unable to supply liquidity precisely when rate volatility is elevated. The real-rate business-model chain is slow but structural: a decade of cheap funding built an entire non-bank lending complex, and a normalisation of real rates challenges those models directly, which argues against treating credit calm as credit safety. The household chain is the fourth and is already thinning: real labour income growth is slowing and gasoline prices are rising, while the August strength in nominal spending was partly one-off and card data has faded in September — so aggregate consumer resilience is doing more work than the marginal consumer’s experience. Layered on top, a cross-border fiscal chain: French spreads at a 14-year wide, UK 30-year gilts at levels last seen in 1998, and debt above $40tn in the US, all of which price through term premium rather than through the Fed’s reaction function.
• Source quality control: a large share of the most quotable items today are single-source social relays and should be treated as indicative only — the entire Kashkari stream, which is real but arrives as one-line Financial Juice and Gelonghui posts reproduced across several indices; El-Erian’s yield-level and regime posts; Lacalle’s headline-only market framing; Zaccardi’s core-PCE line; and the Investing.com headline on falling hike bets. Three data-quality flags deserve precision. First, one UK outlet’s newsletter reports the 10-year yield rising “four basis points, to 4.29 per cent,” which conflicts with every other level in this batch by roughly a full percentage point and is best treated as a typographical error rather than a fifth venue for the same print. Second, the 10-year level is reported at 5.29% (JPMorgan’s close), 5.30% (Bloomberg, and separately El-Erian) and 5.306% intraday (Gelonghui), and the “highest since 2002” and “24-year high” framings coexist with “past its 2007 peak” — level-based triggers must specify venue and timestamp. Third, the August inflation numbers arrive in two incompatible shapes across sources — 3.0% core year-on-year with a 2.05% three-month annualized rate in the sell-side notes, versus 2.6% headline year-on-year in one BEA relay — so any single figure should be checked against the methodology change before use. Finally, several claims are model- or framework-derived rather than observed — the GDP-revision-to-Okun’s-law inference, the curve-steepening-as-communication-failure thesis, the Shiller CAPE return comparison, and the hedge-fund-through-tightening-cycle history — and are useful for framing a distribution rather than as price levels.

---

Appendix: Additional Sources

• ^[14] Barclays — Okun’s law reasserting itself; December hike then unchanged through end-2027
• ^[53] Bloomberg — bond yields climbing almost daily; the trend’s signal on inflation and borrowing costs
• ^[54] Bloomberg — resilient growth, commodities, hikes, deficits cited among selloff drivers
• ^[55] Financial Juice — Japan Cabinet Office expects the BOJ to conduct policy appropriately on prices
• ^[56] Financial Juice — Kashkari: neutral rate probably higher for the time being
• ^[57] Financial Juice — Kashkari: uncertain about the current neutral rate
• ^[58] Financial Juice — Kashkari: neutral rate could be higher than expected
• ^[59] Financial Juice — Kashkari: the longer the economy stays strong, the more I question policy tightness
• ^[21] Financial Juice — Kashkari: consumers continue spending and job seekers have positions
• ^[60] UBS Global Wealth Management — hedge funds generated positive returns in every Fed tightening cycle since 1994

---

This report is a macro-mechanism analysis, not investment advice.