Hold near 90% with the guidance vacuum priced; Goldman's FCI arithmetic caps the hawkish tail while Jefferies calls a 9–3 hawkish pause

0. Weekly Arc

The week ran from a data-versus-communication split into an outright positioning debate. Oct 19–20: the blackout opened with the hold near 84–90% and the Treasury buyer base, not the Fed, identified as the marginal price-setter. Oct 21–22: December hardened into a near-unanimous survey fact while a hawkish communication score hit a five-year high. Oct 23–24: the long end’s largest one-day rally since March 2023, attributed to term premium and AI-credit fear rather than disinflation, and twelve-month pricing shrank. Oct 25: the hold became unanimous and the correlation debate turned on an AI-trade unwind. Today: the last previews before the decision, and the argument is entirely about how a guidance-free Chair gets read.

1. Policy Narrative & Expectations

The net change over the past ~24h is that the meeting itself has stopped carrying information and the entire trade has migrated to how the statement and press conference are worded. October is priced at roughly a 90% hold on CME and LSEG ^[1][2][3][4], with one editorial citing the CME contract at less than a one-in-six chance of action ^[5] — a different statistic from a different vintage rather than a disagreement. December is where the dispersion sits: CME puts 23.8% on unchanged, 68.9% on a cumulative 25bp and 7.3% on a cumulative 50bp ^[3][4]; Soochow reads an 85% December hike probability with cumulative pricing of 2.1 hikes by next April and 2.7 by next September ^[6]; Jefferies reads roughly 90% for December ^[7]; the FT has only a 9% October probability alongside two to three quarter-point rises by end-2027 ^[8]; and Nomura notes December pricing has already fallen below 90% ^[9]. The genuinely new analytical contributions cut in opposite directions. Goldman quantifies the constraint on the hawks: its financial-conditions index has tightened about 30bp since September, only slightly more than a single 25bp hike is normally worth and far below the four hikes priced to a ~4.55% peak in autumn 2027, and because that tightening came mostly from rates, which themselves depend on hike expectations, the FOMC is unlikely to treat it as excessive or as a substitute for action ^[10][11]. Jefferies quantifies the constraint on the doves: with less than 10% priced for next week against core-goods inflation at a 21-month high, core services still firm and housing disinflation reversing, a 9–3 hawkish pause is the likely outcome and the terminal rate probably needs to be higher than priced ^[7]. Nomura’s frame is the bridge between them — the bar for a central bank put is far higher than in the Powell era, so none of this week’s three G3 meetings should deliver the dovish recalibration markets have started to price, with the BOJ most hawkish, the Fed next and the ECB most dovish ^[12].

1.1 FOMC Officials’ Remarks

• [ONGOING] Neutral/swing — Vice Chair Philip Jefferson and New York Fed President John Williams: both want to see additional data before the Fed does anything more ^[5], and Jefferson’s recent speech language — that any further adjustment to the target range to support a more timely return to 2% depends on data trends, the evolving outlook and the balance of risks — is the wording Goldman expects the statement to borrow ^[10][11].
• [ONGOING] Dovish — Governor Christopher Waller: the Committee does not feel urgency on further hikes, which need not come at consecutive meetings but should arrive within a reasonable timeframe ^[10][11].
• [NEW] Dovish — San Francisco Fed President Mary Daly: she sees no cause for urgency “given that soaring mortgage rates and other higher borrowing costs already are slowing parts of the economy”, and adds that “wage growth is cooling, which suggests things may be less overheated than the inflation rate suggests” (Chicago Tribune editorial, relaying her remarks) ^[5].
• [NEW] Chair Kevin Warsh: no fresh public remarks in this batch ahead of his Wednesday 2:30pm ET press conference ^[13]; the operative expectation is that he “may continue the practice of avoiding forward guidance” per Nomura ^[9], that he gave no hint of a slower pace at the September press conference per Goldman ^[11], and the editorial page’s read is that he has been “tight-lipped” and has yet to earn much credibility ^[5] — FT’s institutional point is that this guidance vacuum is itself a Treasury-volatility source ^[8].

1.2 Policy Signals & Institutional Communication

• [NEW] Goldman’s FCI arithmetic is the most consequential number in the batch: the index tightened ~30bp since September, roughly one hike’s worth and far short of the four hikes priced, was flat for most of the year as higher equity prices and narrower spreads offset rates, and the FOMC is unlikely to view the move as excessive or as a reason to skip ^[10][11]. It also attributes the entire 1.1pp core PCE overshoot to the mis-measured portions of tariffs, the energy shock and AI price pressure, forecasts core PCE at 3.0% this year falling to 2.2% in 2027, sees the target range at 4–4.25% in 2026 and 3.5–3.75% at end-2027, and therefore expects no March 2027 hike ^[10][11].
• [ESCALATED] Deutsche Bank expects the message to stay hawkish and pre-empts the substitution argument: its economists see recent inflation data supporting further tightening over time even if the Fed pauses, and it expects Warsh to clarify explicitly that tight financial conditions are not a substitute for higher policy rates, with the press conference focused on how the Fed balances still-elevated inflation and solid growth against the recent tightening ^[1].
• [NEW] Jefferies publishes the hawkish-pause call and the asset map: a <10% market price against 21-month-high core-goods inflation, still-firm core services and reversing housing disinflation makes a 9–3 hold the base case, with the terminal rate likely needing to be higher than priced; its own forecasts are a 4.13% funds rate at end-2026 and 4.38% at end-2027, a 10-year at 5.5% then 5.3%, an S&P 500 target of 8,000 then 9,000, and gold at $4,595 then $5,000 ^[7]. Its positioning conclusion is duration and high-grade credit, plus contrarian overweights in US consumer discretionary, small caps, Brazil and Greece, on the logic that AI capex is the cycle’s engine and growth scares in it are what push long-end yields down ^[7].
• [NEW] Nomura’s Naka Matsuzawa removes the central-bank put: this week’s G3 meetings will not deliver the dovish recalibration markets are starting to price, the threshold for a Fed put is far higher than in the Powell era, and unless credit tightening is very evident no easing signal will be released; the hawkish ordering is BOJ, then Fed, then ECB, implying JPY > USD > EUR, with a live risk that hawkish holds aggravate rather than relieve credit concerns ^[12].
• [NEW] Nomura’s FX team (Yujiro Goto) frames the asymmetric outcome: all three central banks hold, communication is unlikely to materially lift hike expectations, so USD/JPY and EUR/JPY upside should stay limited ^[9]; a tougher anti-inflation stance from Warsh could push the December hike probability back above the recently breached 90% level and support USD/JPY through higher Treasury yields, but US growth worries, equity and credit pullback risk and the midterms make any such strength temporary ^[9].
• [NEW] Bank J. Safra Sarasin keeps a much longer hiking path: neither the Fed nor the ECB needs to rush in October, but without further tightening inflation will not return to 2%, so it holds a December 25bp hike followed by three quarterly hikes next year for a 4.75%–5% terminal rate as demand gradually cools ^[14].
• [NEW] Monex Europe isolates the dollar-sensitive sentence: the Fed will likely preserve December optionality without formal guidance, and if Warsh hints that rising yields have already tightened financial conditions enough, December hike pricing could reverse and weaken the dollar ^[2].
• [ONGOING] The every-other-meeting consensus and the blackout hold: Goldman’s skip-October, hike-December call and its view that the bar for deviating before year-end is high remain unchanged ^[11].

2. Key Data & Market Read

• [NEW] Data release — US October flash PMIs: the composite eased from September but stayed within the top of its five-year range, manufacturing rose to its best since May 2022, services slipped, and selling-price increases were the fastest since August 2022 with demand firm, delivery times longer and backlogs rising ^[15][16]. Market read: GF Macro takes the cost pass-through detail as supporting a relatively tight policy bias ^[16].
• [NEW] Data release — housing: starts fell to a four-month low with single-family permits at their lowest since early 2023, pending home sales hit their lowest on record back to 2001, the NAHB index fell to its lowest since June 2012, and purchase applications dropped over the past four weeks, with residential investment expected to keep dragging growth into year-end ^[15][16]. Market read: the cleanest transmission evidence that the policy rate is biting, and the doves’ only hard data.
• [NEW] Data release — survey and nowcast upgrades: Bloomberg’s October survey marked Q3 growth up and Q4 CPI up with further upgrades through 2027 ^[6][17], Atlanta Fed and New York Fed nowcasts both point to strong Q3 growth ^[6], and the forecast cluster for Thursday’s Q3 GDP release implies a strong print ^[17]. Market read: a resilient-growth, firmer-price backdrop that leaves the hawkish pause intact and gives the doves nothing.
• [NEW] Data release — labour and consumer at the margin: initial claims declined while continuing claims rose, and retail gasoline prices fell on the week ^[15]. Market read: no labour-market crack, a small relief on the household purchasing-power squeeze.
• [ONGOING] September core CPI and the weak September payroll print: the core print was the second consecutive above-expectation month-on-month reading, and the payroll miss had earlier pushed October hike odds down before the CPI reversed them ^[6][17].
• [NEW] Calendar into the decision: core PCE and Q3 GDP land Thursday, with durable goods, consumer confidence and new home sales Tuesday, jobless claims Thursday and the employment cost index Friday ^[1][18].
• [NEW] Narrative impact: the data are doing exactly what they did all month — firm enough on activity and prices to keep the hawkish pause honest, soft enough in housing and the household sector to give the doves a defensible wait — which is why the statement wording, not the data, is Wednesday’s event.

3. Financial-Conditions Signals

• [NEW] Dollar & rates — levels and the driver: the 10-year sits at 4.98%, the 30-year at 5.302% and the 2-year at 4.6752% ^[1], after a week in which the 2-year fell 5bp, the 10-year 9bp and the 30-year 10bp with the long end outperforming ^[16]. The 10-year real yield rose to 2.65% while the 10-year breakeven rose to 2.35% ^[12], but GF Macro’s decomposition is the important part: the week’s fall in breakevens reflected a liquidity premium — risk-off selling of less-liquid TIPS into nominals — rather than a decline in inflation expectations, with the negative correlation between 10-year breakevens and both the VIX and the OAT-Bund spread the strongest since the Middle East conflict began ^[16].
• [ESCALATED] Credit — the AI complex is now the pricing point for the whole risk complex: CoreWeave’s five-year CDS touched 1066bp while Meta, Nvidia and Oracle CDS all hit record highs, and the high-grade hyperscaler bond spread widened on the week ^[16]; index yields sit near 6% for investment grade, about 6.3% for long investment grade and above 7% for BB, with recent issuance running roughly $22bn of investment grade against $3bn of high yield ^[7]. The corrective note is that US tech CDS spreads and the French-German spread narrowed by Oct 23 ^[12].
• [NEW] Liquidity and the FCI: the Bloomberg US financial-conditions index fell to 0.80 from 0.88 a week earlier and 1.21 two weeks earlier, with the euro-area index at 0.91 from 1.19 and 1.12, and offshore dollar liquidity deteriorated with the yen-dollar and euro-dollar swap bases both widening ^[15]; Goldman’s own index is 30bp tighter since September ^[10][11].
• [NEW] Flows: global money funds swung from a net outflow to a net inflow, global bond funds notched a 78th consecutive week of inflows at a low percentile, US bond funds drew a low-percentile inflow, and US equity inflows fell sharply week-on-week while Japan was the strongest incremental equity destination and China equity inflows doubled on domestic passive money with foreign funds still leaving ^[19].
• [NEW] Dollar and positioning: the dollar index traded a 101–102 range and closed at 101.75, roughly flat on the week, with USD/JPY at 156.85 ^[16]; Barclays’ month-end rebalancing model signals mild dollar buying against most majors, its sentiment indicator remains bullish, but crowded long positioning has already capped the dollar in recent weeks ^[20].
• [NEW] Correlation evidence, measured rather than asserted: the correlation between 5-year Treasury yields and the Nasdaq 100 rose notably over the past month, and the sensitivity of Treasury yields to oil turned negative once Brent broke $110 — the market reading higher oil as a growth risk rather than an inflation risk ^[16]; since mid-September TIPS real yields have tracked the Philadelphia Semiconductor Index closely, evidence that US growth expectations are now bound to the AI narrative ^[6][17].
• [NEW] Commodities and volatility: Brent’s most-active contract rose 6.24% on the week to $116.72 with implied oil volatility up to 60.02% from 50.74% ^[16]; spot gold fell 4.20% to around $4,013/oz and closed below $4,000 on Friday even as real yields declined ^[16][6]; MOVE fell to 103.3 while VIX rose to 19.5 ^[12], with a separate read putting VIX at 19.45 from 19.99 ^[16].

4. Global Central-Bank Linkages

• [NEW] ECB: a hold at 2.5% on Thursday is treated as near-certain, with swaps pricing of another quarter-point move this year cut to about a 66% chance from fully priced earlier in the month ^[8]; French 10-year yields have risen above 5%, some 1.7 percentage points over Germany, and Berenberg’s Holger Schmieding argues the ECB “has no mandate and no appetite to fund French fiscal follies”, reinforcing the case for standing pat ^[8]. LBBW expects the decision and press conference to cool medium-term rate expectations and looks for two more hikes in total to a 3% peak by spring 2027 ^[21], while Nomura ranks the ECB the most dovish of this week’s three ^[12].
• [ESCALATED] BOJ: the Bank is expected to hold at the meeting ending Friday, with attention on how clearly the board supports a December move and on whether its most hawkish member proposes a rise next week ^[8]; Nomura expects a hold with an upgraded inflation outlook and assessment to keep December alive, on a baseline of one hike every three months, with faster tightening if yen weakness materialises ^[9]. Governor Ueda has argued that proceeding at an appropriate pace helps long-term yields form stably, and Deputy Governor Himino warned that delaying the adjustment while upside price risks rise could lead to economic deterioration ^[9]. Market-implied December hike probability fell to 60–70% at one point before recovering to 75% ^[12], JGBs followed Treasuries lower with the 10-year at 2.8% and the 30-year at 3.82%, and the Nikkei fell 4.03% on the week ^[16].
• [NEW] BOE: HSBC expects a 25bp hike to 4.00% on Nov 5 by a 7–2 vote with Dhingra and Taylor dissenting, driven by resilient activity and energy-driven inflation, while signalling through the statement and projections that markets price too many further hikes and cutting its two- and three-year inflation forecasts below target ^[22]; Morgan Stanley expects both the BoE and the BoC to hike as “insurance tightening” rather than look through the energy shock, and warns that if demand stays resilient these separate decisions could merge into a synchronised global tightening response ^[23].
• [NEW] PBoC: Governor Pan Gongsheng has set out the plan to switch the policy rate from the 7-day reverse repo to the overnight reverse repo; Goldman expects the switch next year, which would take the policy rate from 1.4% to 1.25% for an effective 15bp cut, as part of a broader framework modernisation rather than a stimulus signal ^[24].

5. Asset Implications

Quadrant	Current probability tilt	Key asset implication	Anchoring narrative
Growth↑ + Inflation↑	Rising modestly	Commodities↑ TIPS↑ nominal long bonds↓ — selling prices at their fastest since August 2022, survey and nowcast upgrades to Q3 growth, oil at $116.72 with implied vol at 60%, and two houses (Jefferies, Bank J. Safra Sarasin) publishing terminal rates above what the market prices	§1.2 / §2
Growth↑ + Inflation↓	Falling	Stocks↑ long bonds↑ gold↓ — supported only by Goldman’s supply-driven overshoot decomposition (core PCE to 2.2% in 2027, no March hike) and by the argument that the AI growth scare, not inflation, is what is repricing the long end	§1.2 / §3
Growth↓ + Inflation↑	Rising	Commodities↑ gold↑ stocks↓ — CoreWeave, Meta, Nvidia and Oracle CDS at records, hyperscaler spreads wider, housing at multi-year lows, residential investment still a drag, and the market already trading oil as a growth negative rather than an inflation input	§3 / §2
Growth↓ + Inflation↓	Unchanged at a low probability	Long bonds↑↑ stocks↓ credit spreads↑ — the long end rallied on the week and term premia fell, but the breakeven decline was a liquidity premium rather than cooling expectations, and money funds are the assets absorbing the flow	§3

Stock-bond correlation call: the sign is currently negative, but for the third distinct reason in three weeks, and each reason matters for how long it lasts. This week’s negative print comes from two measured sources: AI-credit stress at record CDS levels, which hits the equity sleeve’s earnings complex and simultaneously bids duration, and a mechanical liquidity premium in TIPS, where risk-off money sold the less-liquid inflation hedge and bought nominals ^[16]. That second channel is the more dangerous one for a risk-parity book, because it means the measured decline in breakevens is not a disinflation signal — the inflation hedge failed in a risk-off week precisely when its convexity should have paid. The structural call remains a positive, inflation- and supply-driven regime, and three items today strengthen rather than weaken it: the correlation between 5-year yields and the Nasdaq 100 rose over the past month, the yield sensitivity to oil has turned negative only because the market is reading oil as a growth risk, and that reading reverses the moment oil is read as inflation again — which is exactly what a $116 Brent with 60% implied volatility invites. Add the guidance vacuum FT identifies: a Chair who refuses to frame the path leaves term premium to be set by flow and by data, which is the regime in which stocks and bonds fall together on supply news.

Risk-budget implication: do not add long-end duration beta into Wednesday. The bond sleeve should be sized at or below neutral and expressed in the belly and the curve, because the week’s gains came from a liquidity premium and short covering rather than from cooler prices, and because Jefferies’ 10-year forecast of 5.5% at end-2026 sits a long way above spot ^[7] while Goldman’s end-2027 funds path sits well below market pricing ^[11] — those two cannot both be expressed with the same duration position. Treat TIPS with more caution than in prior weeks: this week proved the breakeven can fall for liquidity reasons in a stress episode, so size the inflation hedge off realised stress behaviour rather than off the level of breakevens. Keep commodities at or above benchmark and book energy explicitly as the inflation leg — it is simultaneously the CPI driver, the reason December stays priced, and the only leg whose correlation to the other two is reliably negative in this regime — but size it to event risk given the reported planning for strikes on Iran after the midterms ^[15]. In credit, take Jefferies’ quality bias rather than index beta: hold senior investment grade and the higher-rated part of the structure, and avoid the AI-adjacent cohort whose CDS has just printed records. Equities get the smallest tactical budget relative to volatility, with the consumer-discretionary, small-cap, Brazil and Greece contrarian overweights noted but sized small, since the same AI-credit complex that sets index earnings also sets the marginal financing cost the long end is pricing. Gold should be bought on weakness as reserve diversification rather than as a rates hedge: it fell 4.20% on the week and closed below $4,000 even as real yields fell, which is precisely the sign that it is trading on funding and positioning rather than on the rate channel. Finally, keep a genuine cash allocation — money funds swung back to inflows this week, and a guidance-free Fed is the environment in which cash optionality is worth its carry ^[19][8].

6. Contrarian & Tail Risks

• Consensus fragility: the baseline is now “a hold near 90%, a hawkish pause, December is the meeting, and no central-bank put this week” — and it is unanimous across six houses, which is the fragility. Four assumptions are falsifiable. First, the October price is three statistics: roughly 90% hold on CME and LSEG, an identical 90% read from a second tracker, and a less-than-one-in-six chance quoted from the same CME contract by an editorial page — the last of which is a rounding and vintage artefact, not a rival view, but it is being used to argue a market mispricing. Second, the December path is genuinely unresolved: 68.9% for a cumulative 25bp against 85% and roughly 90% on other reads, with the 2027 endpoint ranging from Goldman’s 3.5–3.75% with cuts to Bank J. Safra Sarasin’s 4.75–5% terminal and Jefferies’ 4.38% — at most one of those is right, and the shape of the curve implied by each is incompatible. Third, the “hawkish pause” base case is only hawkish if the statement says something; Goldman expects the committee to borrow Jefferson’s data-dependence language and warns that if Warsh deems it too close to forward guidance the statement may omit the policy outlook entirely ^[11], which is a genuinely two-sided communication risk rather than a hawkish one. Fourth, both of today’s most quantitative houses are arguing about a reaction function neither can observe: Goldman says tighter conditions cannot substitute for hikes ^[10][11], Deutsche Bank expects Warsh to say exactly that ^[1], and Monex says the opposite sentence — yields have tightened enough — would reverse December pricing and weaken the dollar ^[2]; one sentence on Wednesday resolves it.
• Second-order transmission: four chains are not in the price. The AI-credit chain is now the primary one and has a term sheet: record CDS on a compute provider and on Meta, Nvidia and Oracle, a wider hyperscaler spread, and an equity complex that supplies index earnings while its funding cost rises ^[16] — a single variable that hits both sleeves at once and therefore can flip the correlation sign faster than any macro data. The liquidity-premium chain is the second and is the one risk-parity books systematically underprice: this week the inflation hedge lost money for liquidity reasons in a risk-off week, which means the sleeve that is supposed to pay in a growth scare failed to pay, and the same mechanism would repeat in any disorderly episode. The sovereign-substitution chain is the third: French 10-year yields above 5% and 1.7 percentage points over Bunds, with the ECB explicitly unwilling to fund French fiscal policy, makes US duration the substitute haven for euro-area risk so US term premium can move for reasons with no Fed content ^[8]. The energy-logistics chain is the fourth and is already measurable: Middle East-to-China crude freight rates are far above their historical $1–3 per barrel, refinery maintenance delays and higher shipping costs reinforce each other, and the diesel market is expected to stay unusually tight into mid-2027 with an El Niño food-price shock a live second-round risk — an inflation channel that operates entirely outside the policy rate ^[23].
• Source quality control: volume is not confirmation. Several items are scheduling-only or social posts and are carried as colour, not as reporting: the Investing.com and Jesse Cohen items list the decision, statement and press conference times but carry no policy content ^{[25][13][26][18]}, and the WSJ interactive profiling the 19 people who influence the decision explicitly carries no substantive content ^[27]. Three blocks are out of scope for the Fed narrative and are excluded: a CICC note on Hong Kong international bank valuations whose only Fed-relevant content is the discount-rate channel ^[28], and a Guosheng A-share timing model with no US macro content ^[29]. There is significant franchise concentration: three Huachuang products supply the data and flow detail ^[15][19], two Nomura products supply the central-bank and FX frames ^[9][12], two Soochow products supply the pricing and correlation work ^[6][17], and the Jefferies and Goldman blocks are each single-franchise views — read them as houses, not as corroboration. Genuine conflicts deserve side-by-side reading rather than averaging: October hike probability at roughly 10% against a cited less-than-one-in-six; December at 68.9%, 85% and roughly 90%, which are a cumulative-move probability, a hike probability and a rounded reading; 2027 policy paths spanning 3.5–3.75%, 4.13%, 4.38% and 4.75–5%; VIX at both 19.5 and 19.45 and MOVE at 103.3 across vintages; and a substantive interpretation conflict in which the week’s breakeven decline is read as a liquidity premium by one house ^[16] while the same move is treated as inflation-expectation softening by another ^[12]. Finally, the Chicago Tribune item is an opinion editorial arguing for a hike, not reporting — its Daly and Jefferson/Warsh material is relayed quotation and should be treated as second-hand ^[5].

---

Appendix: Additional Sources

• ^[30] Gelonghui — Goldman relay: hold this week, 25bp in December
• ^[31] WSJ — Asian currencies consolidate against the dollar ahead of the Fed decision
• ^[32] WSJ — officials’ calls for patience plus the approaching midterms make a surprise hike hard to pull off
• ^[33] Bloomberg — TD Securities’ Gennadiy Goldberg advised betting against a hike hours before the inflation report, illustrating how positioning still follows the Fed
• ^[34] WSJ — Treasury yields edged higher in Europe; French bonds outperformed after Moody’s left the rating unchanged
• ^[20] Barclays — month-end rebalancing model: mild dollar buying, EUR and JPY neutral
• ^[35] Minsheng Securities — “super central bank week”: Fed, ECB and BOJ all expected to hold
• ^[4] Everbright Futures — base case is a hold, but the trading logic is shifting from cut expectations to re-hike risk
• ^[36] Financial Juice — Fed and ECB meetings scheduled this week (single source / unverified)
• ^[14] Bank J. Safra Sarasin (Karsten Junius) — neither the Fed nor the ECB needs to rush in October

---

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