sembr.live
中

Fed Watch

Hawkish consolidation after Jackson Hole: September odds hold in the 56–60% band, Barclays flips to two hikes, dollar firms while gold's debasement trade pauses

Warsh's Jackson Hole debut left September hike odds in a ~56–60% band and split the Street — Barclays now expects two 2026 hikes while Morgan Stanley and Nomura read the hawkish turn as optionality and tactics — with August payrolls and CPI set to arbitrate the follow-through .

30 sources ~39 min

0. Weekly Arc

The week’s arc resolved around Warsh’s Jackson Hole debut: after months of deliberate ambiguity, the Chair’s 8/28 keynote flipped the market’s default from “hold unless data force action” to “hike unless data excuse it,” lifting September odds from ~35% into a 56–60% band. Treasuries bear-flattened as tightening was priced at the short end, the dollar firmed toward 100, and gold’s three-week rally cracked — while the long end held on Treasury buyback support. Barclays flipped to a two-hike call; Morgan Stanley and Nomura pushed back. The data calendar now arbitrates: payrolls Sept 4, CPI Sept 11.

1. Policy Narrative & Expectations

The past ~24h brought consolidation rather than fresh repricing: September hike odds held in a ~56–60% band (CME FedWatch 56.9% hike / 43.1% hold [1]; Nomura 58% [2]; GF and Everbright Futures 57% [3][4]; Industrial Securities nearly 60% [5]), with December cumulative pricing now above one hike and ~50% odds of a second by year-end [4][2]. The marginal development is institutional: Barclays reversed its no-hike forecast and now expects 25bp hikes in September and December [6][7], while Morgan Stanley holds its no-change call with only 50–75bp of hikes if wrong [8] and Nomura reads the hawkish turn as a tactical communication move rather than a genuine prelude to September action [2]. The calendar ahead is dense — Beige Book Sept 3, Waller’s inflation-outlook speech Sept 3, payrolls Sept 4, then CPI before the mid-September FOMC [4][9][10].

1.1 FOMC Officials’ Remarks

  • [ONGOING] Chair Kevin Warsh (listed separately): the 8/28 keynote remains the anchor — 2% PCE target “firm and fixed” [9][11], “hard pressed to describe broad financial conditions as restrictive” [12], and “we have work to do” unless inflation returns to target clearly and fast enough [4][11][7], but no explicit September commitment: “committed to a discipline, not to a decision” [9][3].
  • [NEW] Chair Kevin Warsh — marginal speech details: placed responsibility for 65 months of above-target inflation “squarely with the central bank” [11][12]; blamed the ~$7tn balance sheet as the main reason inflation is above target and advocated shrinking it to drain liquidity, which could mean smaller future rate adjustments [13]; stated “inflation is a result of policy choices” [13]; introduced a new PCE-breadth diffusion gauge — 54% of the basket rising >3% over 12 months vs a 32% pre-pandemic average [9][12][14]; and explicitly cut AI productivity gains and wage growth out of the near-term reaction function [9][12].
  • [NEW] Calendar: Governor Waller is scheduled to speak on the inflation outlook on Sept 3 [4]. No other FOMC officials made fresh remarks in the past 24h.

1.2 Policy Signals & Institutional Communication

  • [NEW] Barclays reversed: now expects 25bp hikes in September and December, taking the target range to 4.00–4.25%, and frames the Fed’s H2 pivot from easing to tightening as a response to inflation-rebound/overheating risk [6][7][10].
  • [NEW] Morgan Stanley: maintains a no-change call, interpreting Warsh’s hawkish tone as preserving optionality; the error bar is small — only 50–75bp of hikes if wrong [8]; separately it flags that under Warsh the Fed has turned hawkish on the 2% target, rates-as-main-tool and balance-sheet shrinkage, projecting the balance sheet could shrink $1.5tn or more next year [13].
  • [NEW] Nomura: argues the hawkish pivot is likely a tactical communication move rather than genuine urgency to hike in September, noting similar June remarks were not followed by July action and that only ~58% September pricing itself reflects market skepticism [2].
  • [NEW] Timiraos: several former central bankers believe Warsh “flipped the script” — the data will now have to give the Fed a reason not to move next month; a hold would revive July’s bond-market queasiness, while a hike would cut against the appointing president’s wishes weeks before the midterms [15].
  • [NEW] CICC: a short-lived “preventive” hike would be “bad news out” (1997 analogy: post-hike yields peaked and equities bottomed); with one hike priced, the 10Y’s fair-value center is ~4.5–4.7%, and if no hike materializes the current ~4.7% yield implies a rich 90–100bp term premium, leaving trading room [9].
  • [ONGOING] July FOMC record: 9–3 hold at 3.50–3.75%, with Hammack, Kashkari and Logan dissenting in favor of a 25bp hike [12].
  • [NEW] Fed–Treasury boundary: Warsh did not comment on the Treasury’s expanded long-end buyback plan [11]; Dong Tao argues his hike-tilt effectively cooperates with Bessent’s curve-flattening operations — “from the bond market’s reaction, Bessent won” [16].

2. Key Data & Market Read

  • [ONGOING] July PCE: headline 3.7% y/y — above the 3.6% consensus — with core steady at 3.3% y/y; the print kept September hike expectations elevated in the wake of Warsh’s speech [11][16][5].
  • [NEW] July PCE internals: core ran slightly hot m/m (+0.25% vs +0.2% expected) with ~0.11pp from one-off portfolio-management fees, while the headline eased to +0.16% m/m and the Dallas Fed trimmed-mean was milder at +2.29% y/y — the fee/methodology distortion caps the hawkish read [3].
  • [NEW] Q2 GDP second estimate: held at +1.5% annualized with consumption revised up to +3.4% and private domestic final purchases at +4.2% — a “strong domestic demand” read [3][7].
  • [NEW] July durable goods mix: headline orders beat (+1.1% vs +0.5% expected, transport-led), but core capex missed (+0.2% vs +0.7% expected) — slowing business-investment momentum [5].
  • [NEW] Labor-market revisions: employment growth in the year through March was more moderate than first reported, raising the stakes for the monthly employment report [17]; July payrolls fell 23k with unemployment at 4.1% [18]. August NFP previews cluster around +60k (Dong Tao) to +65k (Morgan Stanley) [16][13].
  • [NEW] August CPI nowcast: the Cleveland Fed nowcast sees August core CPI cooling to roughly +0.20% m/m / +2.38% y/y — if realized, the no-hike base case strengthens [12].
  • [NEW] Narrative impact: Dong Tao argues labor data now matter less to Fed policy than the upcoming CPI [16]; GF lays out the three conditions that would materially raise September odds — August core CPI at or above +0.3% m/m with core services re-accelerating, the 6-month PCE diffusion measure back above 50%, or oil above $90 restoring positive 3-month energy inflation / 5y5y breakevens breaking 2.5% [12].

3. Financial-Conditions Signals

  • [NEW] Rates — the hawkish bear-flattening: on speech day the 2Y rose 12bp to 4.34% (one-month high), the 10Y +4bp to 4.71%, the 30Y just +3bp to 5.22%, and the 10Y-2Y spread compressed from 47bp to 39bp [12]; the move was concentrated at maturities below 2y [14]; the 2Y slipped 2bp to 4.32% in Asian trade Monday [17]. The 10Y real yield sits at 2.40–2.42% [2][3][12], breakevens ~2.31–2.32%, and MOVE rose to 71.0 [2]. Futures now price a peak rate of 4.00–4.50% maintained long-term, with 2027 cuts no longer traded [4].
  • [NEW] Credit: spreads are at historical lows (IG 79bp, HY 263bp), C&I loan growth swung from -3% to +8.7% y/y, M2 growth is 5.4%, and the July SLOOS showed looser commercial lending standards — the evidence behind Warsh’s “not restrictive” read [12][9].
  • [NEW] Dollar: the index closed the week at 99.68, +0.85%, briefly breaking 99.70 on the speech [3][19][11]; Barclays’ framing: the dollar premium is ~5%, near the upper end of the post-”Liberation Day” range — breached only once in the past decade — so the bar for further USD weakness is high even with sentiment negative [20][21]. CICC’s model sees a 96–98 range for H2 [9].
  • [NEW] Gold — debasement trade pauses: spot gold fell 2.95% Friday to $4,453.67 (one-week low $4,444.80), COMEX gold -3.43% to $4,504.10, breaking a three-week winning streak; silver lost the $70 level and the key $67.69 support [11]. Positioning is the fragility: CFTC speculative net longs hit an 11-month high in gold and a 15-week high in silver as of Aug 25 [11]. House views span “buy the dip” (UBS, CICC, China Post Securities, Donghai Securities) and “watch the 4,458 / 4,400 supports” (trader technicals, Guotai Junan Futures) [11][9][22][23]; UBS keeps a $4,600 year-end / $5,400 end-2027 target and Goldman $4,900 year-end [11]; CICC’s static support is 4,400–4,600 with $5,500 the key watershed [9].
  • [NEW] Liquidity & money: broad-money growth is accelerating — Divisia M4 ~6.8% y/y in June with quarterly annualized above 8% — making inflation expectations harder to anchor [24]; CICC sees USD liquidity overall not tight, consistent with Warsh’s read [9].
  • [NEW] Treasury buyback follow-through: the 30Y is back near or above 5.2% [24]; by Aug 21 it had returned to 5.27%, erasing the Aug 19 announcement relief [25]. Analysts characterize the ~$4bn/operation program as marginal, signal-only liquidity support that may raise interventionist concerns [24][21].

4. Global Central-Bank Linkages

  • [NEW] ECB: a 25bp September hike is fully priced [20][21]; Schnabel hints further hikes may be needed to prevent second-round effects [7]; French 10Y (4.139%) now yields more than Italian (4.108%) — a reversal in the market’s assessment of the two countries’ fiscal positions [16], and Barclays prefers German over French assets with a tactical BTP-Bund short by year-end [26]. The ECB is reassessing the stability of US financial policy after the Treasury’s FX and bond-market interventions [27].
  • [NEW] BOJ: markets price >80% odds of a September hike, but policy space is limited; if the Fed hikes, the BOJ will find it hard to accelerate and may have to tolerate a weaker yen and higher JGB yields [2][21]. Deputy Governor Himino’s tone is hawkish without committing to September [7]; Tokyo core CPI rose to 1.8% y/y [5]; 10Y JGBs are near 2.89% and the 30Y at 4.07% [3]; the yen broke back below 160 (160.10) [3].
  • [NEW] BOE: Bailey says second-round inflation effects are mild, allowing the Bank to “wait and see”; Barclays expects a hold, with the next move not until after 2027 [7][10].
  • [NEW] PBoC: Barclays expects the 7-day reverse repo rate to stay at 1.40% through 2026 [10]; the RMB’s August appreciation below 6.72 was its strongest since February 2023 [28].
  • [NEW] Others: Barclays sees the BOK hiking in August 2026 and February 2027 and the RBI in February 2027 [10]; RBA policy is more restrictive than expected with risks skewed to hikes, while the market’s hawkish repricing of the RBNZ may be excessive; it recommends receiving CZK 2y IRS [21]. Central-bank gold demand remains a structural bid: Q2 net purchases ~289t, +62% y/y, with ~45% of surveyed central banks planning further accumulation [29][25].

5. Asset Implications

This section is inference — anchored to the facts above.

QuadrantCurrent probability tiltKey asset implicationAnchoring narrative
Growth↑ + Inflation↑Steady-to-risingWarsh’s resilience read (capex ~9%, profits +20%+, credit at historical lows) plus 54% PCE breadth keep commodities/TIPS as the cleaner expressions; long nominal bonds stay blocked by the term-premium wall, cushioned only by the buyback put§1.1 / §3
Growth↑ + Inflation↓FallingCleveland Fed nowcast (~+0.2% m/m August core CPI) and Dallas trimmed-mean 2.29% support the no-hike camp; equities held up (S&P +0.49% on the week) on Nvidia strength; front-end carry has been repriced hawkish§2 / §3
Growth↓ + Inflation↑Rising (tail)Hormuz transits still below normal, fresh Iran sanctions and a $120–150/bbl oil tail keep stagflation live; gold is the hedge but mid-correction; long bonds do not hedge in this quadrant§2 / §6
Growth↓ + Inflation↓FallingThe hold-through-2026 camp (Morgan Stanley, SPDB International, Guosen Futures) — disinflation resumes, Fed holds; expression is front-end/belly carry plus a tactical long-end entry if no hike (CICC’s 90–100bp term premium)§1.2 / §2

Stock-bond correlation call: the regime is still the inflation-driven, positive-correlation configuration hardest for risk parity — but the marginal shape has improved for long duration. The bear-flattening (2Y +12bp vs 30Y +3bp) shows the policy shock being absorbed at the short end while the long end barely moved, and GF frames the month’s shift as credibility restoration: the market moved from demanding long-end compensation (post-July) to pricing tightening at the short end. Short-dated bonds now carry policy-path risk rather than hedging it; long duration has regained some hedging value precisely because the 30Y held. The caveat is the fiscal/term-premium anchor — with the 30Y back above 5.2% on supply, deficits and $40tn debt, any flaring of the credibility issue re-couples stocks and long bonds. The cross-asset shock absorbers remain the dollar (5% premium resistance per Barclays) and gold, whose crack is rate-driven, not debasement-narrative-driven.

Risk-budget implication: Respect the repriced front end — Barclays’ tactical pay-2Y-SOFR at 4.08% is the clean expression of the rate-risk-premium tail even if the Fed holds [26]. On the long end, prefer CICC’s tactical framing: with one hike priced, the 10Y’s 4.5–4.7% fair-value band leaves trading room, and a no-hike outcome implies a rich 90–100bp term premium [9]. Hold gold as the strategic two-sided hedge but size for whipsaw — CFTC positioning at 11-month highs means the crowded trade can overshoot below CICC’s 4,400–4,600 support zone before the structural bid (central-bank buying, debasement) reasserts [11][9][29]. In credit, carry is intact at historical-low spreads but spread-rate correlation risk is live if hikes land [12][26]. Avoid chasing USD shorts at Barclays’ 5% premium resistance; express the dollar view through gold and EM FX instead [20][21]. In equities, keep beta moderate — small caps (Russell 2000 -1.51% on the week) are the fragile node versus a hawkish Fed [3].

6. Contrarian & Tail Risks

  • Consensus fragility: the ~56–60% September-hike band rests on one speech, not one data point, and the named arbiters land within two weeks. The contrarian stack is thick: Nomura reads the pivot as tactical communication with a skeptical market [2]; Morgan Stanley holds its no-change call (50–75bp if wrong) [8]; DWS warns the market is “doing the Fed’s job” by overpricing hikes the Fed may not deliver [17]; ABN AMRO says credibility concerns resurface if Warsh again fails to support a hike while inflation stays sticky [17]; Goldman flags a considerable risk of the curve replaying the July FOMC scenario if the Fed holds without a clear explanation [17]; Guosen Securities expects a “harsh words, soft heart” style for the year [14]; and Guotai Junan Futures warns a September no-show would restart the “dollar debasement” trade [23]. CICC warns of the broader tail — if credibility keeps eroding, the US could face a simultaneous stock-bond-dollar selloff that eventually forces a hike [9]. Falsifiable pillars: (1) August core CPI near +0.2% m/m (Cleveland nowcast) [12]; (2) payrolls do not badly miss (China Post: the September base case holds unless NFP significantly undershoots) [22]; (3) GF’s three hike-trigger conditions stay absent [12]; (4) Warsh follows rhetoric with action [17][15].
  • Second-order transmission: fiscal–monetary fusion is the live structural tension — Dong Tao reads Warsh’s hike-tilt as explicit cooperation with Bessent’s curve-flattening (“Bessent won”), while Barclays warns the “Treasury twist” may be read as a precursor to inflationary financial repression that undermines the Fed’s price-stability commitment [16][20][21]. Ray Dalio warns of a US debt crisis in ~3 years if the path is unchanged [11]; BofA’s Hartnett makes long-end yields the core variable of global asset pricing [11]; broad-money growth (Divisia M4, quarterly annualized >8%) is accelerating against the intervention [24]. Demand-side red flags: foreign Treasury holdings fell $72.1bn in June (Japan -$26.4bn, China at an 18-year low) and H1 net foreign buying collapsed ~94% y/y [25]; AI-related corporate issuance ($194bn YTD at the four biggest tech issuers vs $108bn in all of 2025; Goldman forecasts $250bn in 2026 and $400bn in 2027) competes directly with government duration [25]. Oil remains the inflation switch: Hormuz transits are still well below normal and fresh US sanctions keep the $120–150/bbl tail live [5][18][28]. Europe adds divergence — French fiscal stress and the Rhine drought (water levels below the 2018/2022 lows) are supply-side inflation inputs [26][16].
  • Source quality control: September odds are a band, not a point — 56.9% (CME FedWatch [1]), 58% (Nomura [2]), 57% (GF [3], Everbright Futures [4]), ~60% (Industrial Securities [5]), 59% (Century Securities [29]). Gold weekly figures conflict: one source reports a Friday close of $4,453.67 (-2.95%) [11] while GF’s weekly read shows spot near $4,562.75 (-0.42%) [3] — likely differing fixing/close conventions; treat as a band. FOMC meeting dates are variously given as Sept 15–16 [11] and Sept 16–17 [9]. Single-source / unverified items: Timiraos’ “flipped the script” post (labeled social/unverified) [15], Bianco’s “liquidity never left / Treasury wallet” regime [30], Dalio’s three-year crisis timeline [11], and the “Liz Truss moment” bond-vigilante framing [30]. Barclays’ two-hike call directly conflicts with Morgan Stanley’s hold-through-2026 — volume of hawkish commentary is not confirmation.

Appendix: Additional Sources

  • [2] Nomura — Matsuzawa Morning Report: tactical-communication read, 58% September pricing, 2Y forward OIS 4.08%
  • [29] Century Securities — gold driver analysis: dollar-credit repricing, central-bank buying +62% y/y
  • [18] SPDB International — hold-through-year-end base case; Hormuz diplomacy and the oil tail
  • [20] Barclays — FX & EM weekly: ~5% dollar premium, financial-repression warning
  • [24] Yicai — twist-op as signal-only; Divisia M4 acceleration; AI issuance competition
  • [12] GF Securities — Warsh reaction-function analysis; three September-hike trigger conditions
  • [23] Guotai Junan Futures — 58% September odds; dollar-debasement restart risk if no hike
  • [14] Guosen Securities — “harsh words, soft heart”; intraday curve moves; hall-of-mirrors critique

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

30-day review of this series 8/6 – 9/5

  • September-hike odds round-tripped on the data-Fed seesaw. Pricing swung from a mid-August dovish low near 27% after soft payrolls and cooling CPI to a post-Jackson-Hole peak near 66–70% as Warsh’s hawkish keynote and Barr’s “act decisively” language took over, then settled back to a coin flip after Waller’s conditional-hold tilt—with the Sept-11 CPI installed as the arbiter.

  • The long end repeatedly defied both the Fed and the Treasury. The 30-year climbed to 2007-era highs above 5.3% despite Bessent’s Aug-19 buyback expansion, whose relief faded within days; the driver narrative shifted from rate-path repricing to a term-premium and real-rate wall.

  • The policy default flipped from “hold unless data force action” to “hike unless data excuse it,” then partially reversed. Warsh’s Aug-28 debut and Barr’s remarks set the hawkish marker, but Williams and Waller’s pushback restored a genuine two-variable fight over whether disinflation or hot core readings govern the September decision.

  • Fed communication itself became a market-moving storyline. Warsh’s “play the ball, not the referee” guidance pullback, the meeting-count reform trial balloon, and the renewed Cook-removal attempt each lifted the policy-uncertainty premium, making every data release behave like a mini-FOMC.

  • The dollar weakened while gold absorbed the fiscal-credibility risk. The DXY slid from near 100 to three-month lows below 99 as Treasury activism fueled de-dollarization chatter, while gold oscillated between rate-driven corrections near $4,400 and debasement-led peaks above $4,600.

  • Global central banks tightened around the Fed’s indecision. BOJ September-hike odds consolidated near 75–84% and the ECB’s path hardened, while coordinated yen intervention and the FIMA channel added a second-order Treasury-demand layer to the long-end story.

Sources30

  1. 美联储9月维持利率不变的概率为43.1% 格隆汇快讯 Score 66
  2. 松泽晨报：沃什的鹰派转向是真实意图还是战术策略？ 外资研报 Score 61
  3. 【广发宏观团队】托底、换档、转型：理解当前政策的三位一体 郭磊宏观茶座 Score 60
  4. [光大期货]宏观周报：美联储加息曲线全面上修，“更高更久”重新定价 内资宏观研究 Score 64
  5. 兴证宏观 | 财政干预叠加鹰派JH 美债黄金双向波动——20260830海外周报 段超宏观研究 Score 63
  6. 沃什杰克逊霍尔首秀后，巴克莱推翻年内不加息预测，预计美联储年内将加息两次。美债收益率曲线趋平，也意味着市场正消化美联储“长期维持高利率”的可能性。点击... 金十-快讯 Score 61
  7. 全球经济学周报：央行需采取行动 外资研报 Score 63
  8. 宏观担忧与科技亮点：数据疲软下政策微调可能性上升 外资研报 Score 61
  9. 中金：沃什转鹰意味着什么？ 中金点睛 Score 66
  10. 全球经济周报：央行利率预测与关键数据前瞻 外资研报 Score 66
  11. 沃什放鹰重挫黄金，机构激辩金价走势 第一财经-资讯 Score 64
  12. 【广发宏观陈嘉荔】沃什Jackson Hole演讲的新信号 郭磊宏观茶座 Score 62
  13. 周日启动：全球宏观展望：通胀是一种选择……但如何选择？ 外资研报 Score 64
  14. [国信证券]Jackson Hole点评：刀子嘴、豆腐心？ 内资宏观研究 Score 61
  15. The Fed’s default had been to hold unless the data argued for a hike. After Warsh’s Jackson Hole speech, several former central bankers think he fli... Twitter·财经快讯 Score 62
  16. 陶冬：沃什重新定价市场风险 第一财经-资讯 Score 61
  17. 沃什讲话推高加息预期，债券投资者仍不信美联储会行动 华尔街见闻 Score 66
  18. [浦银国际证券]月度美国宏观洞察：沃什再次放鹰，美联储9月会加息吗？ 内资宏观研究 Score 62
  19. [东海证券]资产配置周报：美联储鹰派预期持续，寻找国内自主创新、进口替代、内需潜力主线 内资策略报告 Score 60
  20. 外汇与新兴市场周度观点：美元溢价过山车 外资研报 Score 62
  21. 美元溢价过山车：在极端值中Fade美元溢价的价值 外资研报 Score 61
  22. [中邮证券]有色金属行业报告：美联储放鹰引发贵金属价格调整，建议逢低布局 内资行研 Score 61
  23. [国泰君安期货]海外宏观及大类资产周度报告 内资宏观研究 Score 64
  24. 美债收益率操控困局 第一财经-资讯 Score 61
  25. 美债收益率走高背后的财政隐忧 第一财经-资讯 Score 61
  26. 全球投资组合经理摘要：利率、风险与韧性 外资研报 Score 63
  27. 杰克逊霍尔年会期间，美联储官员主动安抚欧洲同行，但无法保证特朗普政府不会说变就变。美国财政部近期干预汇市和美债，让欧洲央行重新审视美国金融政策的稳定性... 金十-快讯 Score 61
  28. [国信期货]宏观月报：国内广义财政有望逐步发力 海外美元指数或大幅震荡 内资宏观研究 Score 61
  29. [世纪证券]大周期行业周报（8月第4周）：美联储“放鹰”或扰动金价 内资行研 Score 62
  30. https://www.youtube.com/watch?v=h96bNFN29tw Rational Dissent Episode 006. Can the Treasury Bully the Bond Market? In a regime where financial conditio... Twitter·宏观市场 Score 64