Post-hike digestion: Goldman says the market over-prices the cycle as the 10-year holds at 5%

0. Weekly Arc

The arc ran from the Sept 11 core CPI beat into the Sept 16 FOMC, where the first hike in three years landed 12–0 with a dot plot pointing to one more 2026 move and no forward guidance. The post-meeting phase has been digestion rather than extension: the long end pushed to 5% and then partly reversed, the BOJ hiked to 1.25% and the BOE held hawkishly with a QT tweak, and the sell-side split into a “cycle” camp and a “calibration” camp. Today’s incremental development is that the dovish analytical case broadened from an outlier to a house view, while the inflation and dot-plot facts that justified the hike are unchanged.

1. Policy Narrative & Expectations

No new policy action in the past ~24h, but the composition of the hike debate shifted on two margins. First, pricing cooled slightly from its peak: after the hike landed, expectations of further increases eased and the tape behaved as “bad news exhausted” rather than as a fresh tightening impulse, with the probability of an unchanged target rate rising and the hike probability falling to 57.6% ^[1], while Nomura’s read of market pricing puts October hike odds at 53% against 47% a week earlier ^[2]. Second, and more consequential, the case against the market’s cycle pricing broadened from a minority view into house calls: Goldman Sachs’ Dominic Wilson and Josh Schiffrin argue the market has strongly priced a complete tightening cycle of about four hikes while their base case is one more in October followed by a longer pause, and that “if you ask me whether it’s two or six, I’d say two is far more likely” ^[3], while Nomura’s Naka Matsuzawa argues market-implied hike expectations are too high and will be revised down because Chair Kevin Warsh’s hawkish posture is political rather than his true policy intent ^[2]. Against that, Oxford Economics expects a December hike followed by a long pause and explicitly rejects a “one-and-done” read ^[4], Morgan Stanley still projects two more hikes to a 4.25%–4.50% terminal ^[4], and Northeast Securities warns December hike risk should not be underestimated under current guidance even though it leans toward a 2016-style small-hike-then-pause pattern ^[5]. The net effect is a wider two-sided distribution on the path with the decision itself fully priced.

1.1 FOMC Officials’ Remarks

• [ONGOING] Hawkish — Chair Kevin Warsh (listed separately): the operative line from the meeting remains that “The plain fact is that inflation is too high” ^[6], with the move framed as recouping only “a small portion of the easing” and an economy that looks able to withstand a modest rise in rates ^[4].
• [NEW] Neutral — New York Fed President John Williams: made data-dependent, wait-and-see remarks, per Nomura’s strategy note ^[2].

1.2 Policy Signals & Institutional Communication

• [ONGOING] The decision and the projections: the FOMC raised the target range 25bp to 3.75%–4.00% by a unanimous 12–0 vote, the first tightening since July 2023, with the median end-2026 policy-rate projection lifted to 4.1% from June’s 3.8% and the dot plot signalling one more hike this year ^[4][6].
• [NEW] Goldman Sachs’ framework (Dominic Wilson, Josh Schiffrin): the market has priced a full ~four-hike cycle, their base case is one more hike in October then a longer pause, Schiffrin personally leans to two (two to three) hikes, and Wilson describes this as the clearest long-side window recently and flags 30-year real yields above 3% as attractive on a medium/long-term view ^[3].
• [NEW] Nomura: market-implied Fed hike expectations are too high and should be revised down; Warsh’s hawkish posture stems from “behind-the-curve” concerns rather than his true intent, and he may actually want to avoid an October hike given the political difficulty before the midterms; the house favours equities and bonds and is bearish the dollar ^[2].
• [NEW] Northeast Securities: December hike risk should not be underestimated under current guidance, but weak endogenous inflation potential in wages and housing argues against a rapid consecutive cycle; it leans toward a 2016-style small hike followed by a sustained pause ^[5].
• [NEW] Oxford Economics: senior economist Bob Schwartz told Yicai the hike is unlikely to be “one-and-done” — the 1997 single hike aimed only at cooling an overheated stock market, whereas today’s is rooted in fundamentals with persistent external shocks — and expects a December hike followed by a long pause, with US inflation returning to a downward path next year if external shocks fade ^[4].
• [NEW] ING’s James Knightley wrote that only two FOMC members see no further hikes this year, indicating the committee has on the whole turned hawkish ^[7].
• [ONGOING] Morgan Stanley projects two more hikes (December and March) to a 4.25%–4.50% terminal rate held through end-2027 ^[4].
• [NEW] The bond-market question is now framed explicitly two ways: WSJ asks whether “bond yields rose because the economy is strong or because America’s credibility is on the line?” ^[8].
• [NEW] NYT’s DealBook assembles competing explanations for a 10-year at 5% and quotes Apollo’s Torsten Slok: “We have never seen such a dramatic re-pricing of the Fed” ^[6].
• [ONGOING] Political channel: President Trump wrote on Truth Social after the vote that rates should be 1 percent or less, and public criticism of the hike persists with a preference for rates at 1% or lower ^[6][9].
• [ONGOING] The Treasury’s expanded long-bond buyback is judged far too small to move yields — this month’s $5.19bn is dwarfed by the market ^[6].
• [NEW] A plumbing read that cuts against dysfunction fears: Stanford’s Darrell Duffie said “none are flashing red” among the dysfunction measures he watches, with auctions going smoothly by bid-to-cover, though at higher rates ^[6].
• [NEW] Week ahead: central-bank rate decisions and global flash PMIs top the calendar ^[10][11], with US September preliminary manufacturing and services PMIs on Wednesday Sept 23, plus jobless claims, August new home sales, preliminary durable goods orders and the final University of Michigan sentiment print ^[7].

2. Key Data & Market Read

• [ONGOING] August retail sales came in above expectations with consumer demand still firm, one of the data points that underpinned the hike ^[12][4].
• [ONGOING] August inflation ran hot and employment stayed resilient, the backdrop the hike and the hawkish post-meeting messaging were built on ^[1].
• [NEW] The expectations check points the other way: according to a University of Michigan survey, households expect inflation over the next five years to average less than they did a year ago, and TIPS-implied professional expectations put ten-year inflation around the target rather than above it — i.e. neither consumers nor professional investors have repriced a second inflation wave ^[6].
• [NEW] Narrative impact: the data still justify the hike without settling the path. The hawkish leg rests on sticky measured inflation, firm retail demand and a labour market the committee does not treat as a constraint; the dovish leg rests on anchored expectations and an economy that is not overheating ^[1][6][4]. That asymmetry is why the dot plot, the term-premium debate and oil, not the data, are the swing variables.

3. Financial-Conditions Signals

• [ONGOING] Rates: the benchmark 10-year yield climbed 5.3bp to 5%, again approaching its pre-2007-financial-crisis high, and the 2-year rose 6.2bp to 4.75% ^[4]; a separate close puts the 2-year at 4.76% and the 10-year at 5.01% ^[1]; the 10-year had hit 5% before the Fed vote, the highest since 2007, and has stayed near that level since ^[6].
• [NEW] Curve and breakevens: the 10y–2y spread has narrowed to 0.25% and the 5-year breakeven inflation rate has fallen to 2.31% ^[1]; across markets over the week the US 10-year fell 4bp with a twist flattening, the UK fell 12bp, Europe fell 3bp and Japan rose 1bp, with the US two-year forward OIS terminal-rate proxy flat at 4.40% ^[2].
• [ESCALATED] Equity flows: US equity funds recorded a fourth consecutive week of net outflows in the week to September 16, roughly flat versus the prior week and the second-highest level of the year ^[4].
• [NEW] The long-end driver is now a stated disagreement: NYT notes the term premium “can’t be directly observed, only inferred”, sets Goldman’s “high and sticky” call against Apollo’s Torsten Slok, who cites New York Fed data to say the term premium is well below its early-2010s levels and has moved sideways over the past year ^[6]; Goldman’s Dominic Wilson instead attributes the rise in real yields mainly to AI and enterprise financing demand combined with persistent fiscal deficits — the capital-competition narrative ^[3]. Supporting inputs: the federal deficit is near 6% of GDP, unusually high for healthy growth, and a surge of corporate borrowing for AI data centres and power plants has turned tech giants from buyers of Treasuries into competitors for investors’ funds ^[6].
• [NEW] Dollar views diverge: Nomura expects the dollar to soften next week as Fed hike expectations fall while BOJ hike expectations hold ^[2], whereas Goldman’s Josh Schiffrin names dollar strength among his most confident themes on the Fed hike, a clearly hawkish press conference and the AI narrative ^[3]; the dollar index fell rather than rose on September 17 and USD/JPY was around 157 after the BOJ meeting, suggesting the market has not bought a one-way dollar leg ^[9][13].
• [NEW] Hong Kong credit channel: the HKMA raised its base rate 25bp to 4.25%, but HSBC, BOCHK and Standard Chartered left HKD prime rates unchanged at 5%, 5% and 5.25% respectively, and one-month HIBOR sat near 2.91% before the hike — below the US policy rate — so a widening US–HK gap invites HKD-funded carry into USD assets and a later HIBOR adjustment if the peg’s weak side is touched ^[9].
• [ONGOING] Liquidity and supply: next week’s Treasury calendar carries auctions of 2-, 5- and 7-year notes ^[7], while the expanded long-bond buyback is judged too small to lower yields ^[6].

4. Global Central-Bank Linkages

• [ESCALATED] BOJ: the Bank raised its policy rate to 1.25% on a 7–2 vote, with board members Asada and Sato dissenting on the grounds that the economy is not strong enough for an immediate hike and inflation acceleration is not clear ^[14][13]; Governor Ueda said the policy “situation has already changed” with underlying inflation gradually moving toward 2%, did not rule out consecutive hikes or a single move larger than 25bp, but gave no timetable ^[13]; the BOJ also modified its climate-change financing operation, setting the loan rate to the average policy rate over the loan period with a ¥50tn total cap and a ¥10tn single-counterparty cap ^[14]. A BOJ working paper puts the long-run inflation-expectations anchor at roughly 1.6%–2.4%, and Nomura expects a terminal rate of at least 2.0% with further hikes possibly in December 2026 and March, July and December 2027 ^[14]; board member Masai noted 1–2 year real rates remain negative ^[14].
• [NEW] BOJ market read: the market took the outcome as “hiked as expected, pace not necessarily accelerating” — the yen weakened, with USD/JPY briefly above 157, JGB yields fell back and Japanese stocks strengthened ^[13]; yen-funded allocation into overseas assets is hard to reverse while US yields stay high, and the risk of a large carry unwind is judged smaller than earlier after yen net shorts fell sharply since the late-July intervention ^[13]; Joseph Wang’s one-line summary was “BOJ hikes and disappoints” ^[15] (single source / unverified).
• [ESCALATED] BOE: the Bank held Bank Rate at 3.75% with Governor Bailey acknowledging upside inflation risks and hinting at possible hikes in coming months, while market pricing fully discounts four hikes by June 2027 ^[7]; the BOE also tweaked its QT programme ^[15] (single source / unverified).
• [EASED] ECB: the ECB hiked last week, driven not only by high oil prices but also by the euro area’s recent strong performance ^[7]; however, market pricing of a European hike has fallen to 53% from 78%, part of a broader downward revision of global central-bank hike expectations ^[2].
• [NEW] PBoC: September’s outright reverse repo operations were rolled over in equal volume across both tenors on Sept 7 and Sept 15, ending the two consecutive months of increased-volume operations seen in July and August ^[12]; the Q2 monetary policy report formalises the 7-day OMO rate as the policy rate with DR001 as the target rate and a ±25bp corridor, and signals gradually more frequent overnight reverse repo operations ^[16]; policy is expected to concentrate on delivering existing structural and quantity measures with a low near-term probability of a rate cut, with incremental easing possible around Q4 if Q3 growth fails to stabilise on existing measures and base effects ^[16]. Domestic funding prices rose over the week, with DR001 up about 7bp and DR007 about 1.6bp, against a bank net interest margin at 1.4% ^[12].
• [NEW] Others: the HKMA raised its base rate 25bp to 4.25% by formula while local prime rates stayed put ^[9]; Switzerland, Sweden, Norway, Mexico, South Africa and Indonesia all hold rate decisions next week ^[7]; intervention risk during Japan’s long holiday period is rising ^[17].

5. Asset Implications

Quadrant	Current probability tilt	Key asset implication	Anchoring narrative
Growth↑ + Inflation↑	Rising	Energy/commodities and TIPS remain the cleanest expressions, with Brent near $104 and the oil-supply repair timeline still measured in weeks, while nominal long bonds are capped by a 10-year at 5% and a term premium several desks call high and sticky; the more robust expression is curve and relative value rather than outright duration	§3 / §2
Growth↑ + Inflation↓	Rising modestly	The analytical broadening of the dovish case — Goldman’s two-versus-four hikes, Nomura’s downward revision call, breakevens at 2.31% and household expectations below a year ago — supports belly carry as well as quality, cash-flow-durable equity beta, with Goldman’s year-end “all risk assets up” window explicitly conditional on oil falling	§1.2 / §2
Growth↓ + Inflation↑	Rising (tail)	The supply leg is policy-insensitive — pipeline repair measured in weeks, Red Sea flows under pressure, and a Fed that cannot offset an energy impulse — so nominal bonds hedge neither leg and gold, not duration, is the hedge; note the sequencing risk that oil strength lifts inflation expectations and term premium at the same time	§3 / §2
Growth↓ + Inflation↓	Unchanged	The credit-repair tail is cheaper to own than a week ago — dealer dysfunction gauges are not “flashing red”, auctions are clearing, and both household and market inflation expectations are anchored — but the Fed’s own non-guidance and the December-hike distribution keep a fast dovish pivot out of reach	§1.2 / §3

Stock-bond correlation call: the regime is still the inflation/policy-driven positive-correlation configuration that hurts risk parity, but today’s evidence sharpens why — and that matters more than the sign. The inflation-compensation channel is not doing the work: the 5-year breakeven has fallen to 2.31%, TIPS-implied ten-year expectations sit near target, and household five-year expectations are below a year ago, so duration is not failing to hedge because inflation expectations are unanchored. It is failing because the long end is being set by variables monetary policy cannot touch — fiscal supply with the deficit near 6% of GDP, corporate AI borrowing competing for the same pool of funds, and a term premium that even the analysts cannot agree on (Goldman’s “high and sticky” against Apollo’s “sideways and below early-2010s levels”). That is the configuration in which both legs of a risk-parity portfolio reprice off the same variable during a growth shock. The counter-consideration is worth stating plainly: if the “Fed is not the driver” thesis is right, the correlation structure itself shifts rather than merely its level, and no amount of dot-plot clarity restores duration’s hedge value.

Risk-budget implication: treat the delivered hike as fully priced and the path as a genuinely wide two-sided distribution, and treat the dovish broadening — not the hawkish dot plot — as today’s marginal information. Keep commodity and energy risk at or above benchmark: energy remains the sleeve whose driver is orthogonal to the Fed’s reaction function, and the physical supply story (weeks-long pipeline repair, pressured Red Sea flows) has not resolved. In rates, prefer curve and relative-value expressions to outright duration shorts, and note that the same evidence that argues against long-end duration (term-premium disagreement, a 6% deficit, AI issuance) argues for funding any duration rebuild in the belly; the fall in breakevens and the 10y–2y compression to 0.25% are the specific prints that make a flattener less attractive here than it looked a week ago. Accept that the demand ledger for duration is one-sidedly contested: equity funds have bled for four straight weeks while bonds have been the funding source of that reallocation. In credit, keep duration short and favour spread over pure duration, but recognise that the stress point has migrated from corporate balance sheets to the household and Hong Kong funding channels — mortgage and card transmission is the drag, and the HKD carry dynamic is a second-order amplifier if the peg’s weak side is tested. In equities, hold moderate beta expressed through cash-flow durability and AI-linked earnings; the split tape (Dow down sharply for the week, Nasdaq up, technology and communication services among the only advancing sectors) is the signature of a market absorbing higher rates by rotating rather than de-risking. Hold gold as the structural fiscal and dollar-credit hedge, sized for real-rate whipsaw rather than as equity insurance, and note that the sell-side consensus remains structurally constructive even under tightening.

6. Contrarian & Tail Risks

• Consensus fragility: the baseline is “hike delivered, one or two more to come, equities can look through it”, and it rests on assumptions that are today demonstrably unsettled. First, the market’s cycle pricing is not one number: Goldman reads roughly four hikes priced against a two-hike base case, Morgan Stanley models a 4.25%–4.50% terminal, Northeast Securities says December risk is understated, Nomura says pricing is excessive, and LSEG-based pricing implies one more hike before end-2026 and two in total by June 2027 — these cannot all be right, and none is a Fed commitment because the Fed has deliberately withdrawn guidance. Second, the sell-side’s own confidence is thin at the top: Goldman’s Dominic Wilson says it is “quite unusual” and he has “a faint unease” that everyone is so comfortable with the growth outlook, and that he increasingly feels rate pricing is approaching an extreme in one direction. Third, the justification is circular — the hike was delivered to satisfy market pricing that the Fed itself refuses to guide. Fourth, the long-end debate is genuinely “overdetermined”: economists’ term for too many independently sufficient causes, which means fixing the wrong one brings no relief.
• Second-order transmission: three chains are not in market pricing. The energy chain has a dated physical constraint — Saudi Arabia aims to restore partial pipeline operations within days, but repair of damaged pumping stations and full capacity recovery could take six to eight weeks, while Red Sea flows face continued pressure; because that impulse lands after the August survey period, the Fed is tightening into cost pressure rather than ahead of it. The AI-financing chain has flipped sign: corporate borrowing for data centres and power has turned the largest technology issuers from suppliers of funds to competitors for them, and that crowding-out is now named as a driver of the long end by Goldman itself — if long yields grind higher, the marginal project’s hurdle rate rises and the equity transmission becomes self-reinforcing. The fiscal-funding chain remains structural and self-defeating: a deficit near 6% of GDP in a healthy economy forces higher rates to clear auctions, and the Treasury’s own expanded buyback has been read by critics as confirming rather than dispelling the fear that there is no fix. Layered on top, a fourth, less-discussed chain: the Hong Kong funding channel, where an unchanged prime rate beside a higher base rate invites HKD-funded carry into USD assets, with the eventual adjustment arriving through HIBOR and the aggregate balance rather than through a policy decision.
• Source quality control: several items in this batch are social relays or single-source posts and should be treated as indicative only — Christophe Barraud’s week-ahead lists of central-bank decisions and flash PMIs, the Bloomberg “Fed officials raise interest rates / signals another 2026 hike” relay, Bob Elliott’s note on developed-market central-bank caution, Joseph Wang’s one-line summaries of the BOJ and BOE, and the Jin10 headline digest on gold positioning, the $7 trillion reallocation and Japan holiday intervention risk. Conflicts deserve side-by-side reading: the US 10-year is reported at 4.93% in Nomura’s note against 5.00%–5.01% in Yicai and Soochow, a timing and instrument difference that matters for any level-based trigger; weekly equity performance is reported inconsistently, with one account describing the S&P 500’s second consecutive weekly loss and the Dow’s largest weekly decline since March while another reports the Dow down 1.69%, the Nasdaq up 0.72% and the S&P down only fractionally; the direction of long-dated US yields after the hike is described as falling in some Chinese institutional notes and as holding at a post-2007 high in WSJ and NYT; and the term premium is simultaneously “high and sticky” and “below early-2010s levels and sideways”. One item flags its own internal inconsistency and should not be relied on: the Zhongtai Futures report’s headline says the Fed cut 25bp as expected while its body says the Fed announced a 25bp hike. Finally, two items in the batch are not usable as evidence for US policy — a China A-share timing report whose indicator set contains no US Fed content, and a promotional digest for a weekly macro roundup with no extractable facts — and one PBoC review piece is dated August 14 and should be read as background rather than a current signal.

---

Appendix: Additional Sources

• ^[10] Christophe Barraud — Week 39 preview: central-bank rate decisions, global flash PMIs, UN General Assembly
• ^[11] WSJ week-ahead relay — US PMI data in focus for FX and bonds
• ^[8] WSJ — did bond yields rise because the economy is strong or because US credibility is on the line
• ^[9] Macro commentary (WeChat) — HKMA base rate, unchanged HKD prime rates, RMB above 6.70 and the US–HK rate gap
• ^[14] Nomura — BOJ September MPM details: dissents, climate financing operation, terminal rate and expectations anchor
• ^[12] Zhongtai Futures — China bond market range view; Fed hike noted in body text (internal inconsistency flagged)
• ^[18] Huajin Securities — historical post-hike sector patterns (A-share focus, limited US read-through)
• ^[19] Guosheng Securities — A-share timing radar, no US Fed content
• ^[13] Industrial Securities — BOJ decision details, yen and JGB reaction, carry-unwind risk
• ^[16] Sinolink Securities — PBoC operations and framework review (report dated August 14)

---

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