Economists flip to a hike into the FOMC; the dot plot and the "after" carry the risk

0. Weekly Arc

The past week completed a hawkish consolidation that began with Chair Warsh’s Aug 28 Jackson Hole keynote. The Sept 4 payrolls beat re-leaned September pricing toward ~60%; the Sept 10 PPI and Sept 11 core CPI broke the market through, lifting hike odds from ~70% to an 85–90% band and pulling the 10-year toward 5% and the 30-year to ~5.35%. The ECB hiked on Sept 10 and the BOJ is expected to follow on Sept 18. The live question has migrated from whether the Fed hikes to what the dot plot and Warsh’s press conference imply about a second move.

1. Policy Narrative & Expectations

The past ~24h brought no fresh Fed voice but a decisive change in the composition of the hike consensus. The Reuters poll conducted after Friday’s inflation report shows 85% of economists — 86 of 101 — now expect a quarter-point increase to 3.75%–4.00% at the Sept 15–16 meeting, the first since July 2023, with a near-53% majority (37 of 70) expecting at least one further hike by end-March versus 56% expecting steady rates in the prior week’s poll ^{[1][2][3][4][5]}. That closes the market-versus-forecaster divergence that ran through the previous week’s briefings, and it leaves pricing as the marginal driver rather than the data: futures price close to a 90% chance of a move this week and roughly four increases by end-July 2027 ^[1], while CME FedWatch shows a cumulative 25bp hike for September at 86.2% and, by October, a 50.1% chance of 25bp against 42.9% for 50bp ^[6][7]. The consequence is that the meeting’s information content now sits almost entirely in the dot plot and the press conference: Goldman Sachs’ Sept 13 note explicitly frames the hike as a move to satisfy near-90% market-implied odds rather than a fundamental judgment ^[8][9][10], Morgan Stanley argues the market has over-extrapolated a single move into a meeting-by-meeting cycle ^[11], and Jefferies and ING both expect the Fed to deliver less tightening than forwards price ^[12][13].

1.1 FOMC Officials’ Remarks

No public FOMC remarks in the past 24h. The batch’s official content consists of a synthesis of remarks made between the July FOMC and today (see §1.2).

1.2 Policy Signals & Institutional Communication

• [REVERSED] The economist consensus flipped to a hike: a Reuters poll now shows 86 of 101 economists expecting a quarter-point move to 3.75%–4.00% on Sept 16, against 65 of 93 expecting a hold in the Sept 9 poll ^[2][4][5]. BofA’s Stephen Juneau says Warsh has “boxed himself into a position where the Fed will forgo a hike only if data is very weak,” and Reuters notes the mainstream expectation of 2027 rate cuts no longer exists ^[5][1].
• [ESCALATED] The sell-side is clustered on a hike but split on the path: Goldman Sachs’ David Mericle (“A Hike Without a Signal?”) expects minimal statement changes, no guidance on the future path, the terminal fed funds forecast raised to 3.25%–3.5% and a median dot showing only one 2026 hike by a 10–8 majority, with Governor Waller dissenting because recent core PCE fell below his 2.8% threshold ^[8]. BofA expects the SEP 2026 median at 4.125% (from June’s 3.75%) and 2027 at 3.875% ^[14]; HSBC likewise expects a 4.125% 2026 median with a further hike in December ^[15]; Deutsche Bank expects a 75bp total across September, December and March ^[16]; Huatai Research expects the 2026 median to be raised to two hikes ^[17]; TD Securities expects a 75bp cumulative cycle from September to Q1 2027 ^[18]; BNP Paribas expects three hikes running to Q1 2027 ^[19].
• [NEW] Goldman’s stated rationale is market pricing, not fundamentals: the revision “was driven less by its economic outlook and more by financial market pricing, with investors largely expecting a rate increase” ^[8][9][10], and its probability-weighted rate path is more dovish than market pricing ^[8].
• [NEW] The bank tally has flipped within days: of the 20 major investment banks tracked by Nick Timiraos, 17 now expect a hike this month, and almost every Fed watcher who had expected a hold changed their forecast after the CPI ^[20]; as recently as last Thursday about half of that table still expected a hold through year-end ^[20].
• [NEW] Dot-plot transmission may be smaller than assumed: Chicago Fed research relayed by Christophe Barraud finds that when the median SEP fed-funds projection comes in 25bp above what Fed watchers expected, market-implied rates move only about 5bp on the day, consistent with dot plots being read as conditional assessments rather than commitments — single source / unverified ^[21][22].
• [ESCALATED] The communication critique is now mainstream: Bloomberg argues Warsh has rightly questioned the Fed’s communications but that his arrival as chairman only worsened the problem, and that he must rethink how he explains policy after this week’s meeting ^[23]; Reuters describes “a new regime under Warsh of no rate guidance, along with heightened uncertainty” that left few economists ready to commit ^[1].
• [NEW] A political-cover read on the hike: Goldman Sachs suggests Warsh may use majority opinion as cover to raise rates before the midterm elections rather than bear Trump’s anger alone — secondary attribution ^[24].
• [ESCALATED] The 2027 threshold is being repriced via the voter rotation: hawkish voters Logan, Hammack, Kashkari and Paulson all rotate out this year, and the 2027 cohort (Goolsbee, Daly, Barkin and the incoming Atlanta Fed president) reads as more neutral, implying a higher bar for another hike next year and, if growth weakens, room for cut expectations to open ^[25][26].
• [ESCALATED] Political pressure intensified over the weekend: Trump again urged the Fed to keep rates low, saying “I don’t know, but we should have the lowest interest rates in the world,” and has floated cutting trade ties with economies running deficits against the US if the Fed does not cut ^{[27][28][29][30]}.
• [ONGOING] Treasury buyback shortfall: the first upsized long-dated operation set a $6bn cap but bought only $5.19bn, with Chengtong noting the shortfall ^[11][31].
• [ONGOING] Market-implied hike odds sit in an 85–90% band across vendors and snapshots ^{[1][14][32][33][34][35][36][6][37][38]}.

2. Key Data & Market Read

• [ONGOING] August CPI (released 9/11): headline in line with expectations, core above on the month ^{[33][39][40][41]}.
• [NEW] The composition and the PCE conversion are where the hawkish calls rest: BofA reads the core upside as concentrated in one-off core services — a concentrated wireless phone service repricing plus volatile airfares and lodging — and raised its August core PCE tracking estimate accordingly; it argues this should not be over-read as a deterioration in longer-term inflation trends ^[33]. Deutsche Bank reads the CPI as implying an above-threshold monthly core PCE ^[16]; Morgan Stanley raised its August core PCE forecast to 0.25% m/m, implying still-elevated 3/6/12-month annualized readings ^[11]. Working the other way, HSBC flags that the forthcoming methodology change could produce a lower revised core PCE year-on-year, while cautioning policymakers focus on trend and upside risks from energy ^[15].
• [NEW] Activity and labour cross-checks support the hike case: the Atlanta Fed GDPNow tracker shows strong Q3 real GDP growth with private domestic demand close behind and the wage growth tracker accelerating slightly ^[16]; August payrolls came in above expectations ^[15].
• [NEW] Consumer survey worsens: the University of Michigan’s September sentiment index slumped to a multi-year low while one-year inflation expectations jumped ^[42].
• [NEW] The calendar ahead is consumer- and inflation-heavy rather than Fed-heavy: US retail sales and import prices, jobless claims, Canada CPI, UK CPI, Japan national CPI and New Zealand GDP ^[43][44][41]; Barclays forecasts US August retail sales below consensus and UK CPI in line ^[44].
• [NEW] Narrative impact: the data justify the hawkish read but do not settle the path. CICC’s counter-reading is that the August increase was driven by oil, airfares and hotels — factors unlikely to persist — and that year-end CPI will probably fall back from current levels unless the oil centre holds above roughly $95 ^[45]. The market’s own behaviour (risk assets rallying on a hot print, with only the front end repricing) says it is trading “uncertainty removed” rather than an inflation shock ^[46][47][48].

3. Financial-Conditions Signals

• [NEW] Credit & banking: BofA lays out the fork explicitly — a hike supports an orthodox “twist flattening,” while a surprise hold would be read as extremely dovish and could push the 30-year yield toward 5.75% with buybacks unable to stop it ^[14]. The same desk reports positioning broadly bearish with CTAs near extreme shorts across the curve and asset managers cutting duration in favour of spread products, setting up a possible technical short-covering rally if Warsh sounds dovish ^[32]. Deutsche Bank’s asset allocation team reads bond positioning at roughly the 5th percentile as historically coincident with a near-term yield peak, while noting systematic strategies still hold high equity longs, which supports the market but limits further buying ^[49]. JPMorgan notes US and European private-credit growth remains strong, evidence that current yield levels have not strangled activity and financial conditions remain effectively loose ^[50]. Barclays argues higher rates affect the US economy less than in past cycles because AI investment offsets housing-market weakness, warning bond bulls expecting a yield-driven slowdown ^[51].
• [ONGOING] Liquidity: Morgan Stanley reads the buyback’s price-sensitivity as evidence of a non-QE, price-conscious Treasury and suggests lower future long-end issuance rather than intervention ^[11]; the buyback undershoot itself remains the operative fact ^[11][31].
• [NEW] Demand for duration is thinning: BofA reports custodial holdings fell another $11bn and foreign reverse repo balances fell $8bn last week, a combined year-to-date decline of roughly $110bn, with foreign participation in Treasury auctions back below 20%; short-term government funds took the largest inflows while intermediate funds and intermediate corporate bond funds saw outflows ^[32]. Deutsche Bank’s flow data show equity funds +$9.8bn, bond funds +$17.5bn and money market funds +$12.9bn, with US equity funds down $3.9bn for a third consecutive week ^[49].
• [NEW] Dollar & rates — the long-end driver is now a stated disagreement: Barclays says the term premium has risen back to pre-financial-crisis levels and the 10-year is approaching 5% with bonds lacking appeal ^[51]; JPMorgan counters that the 10-year term premium is already at a 10-year high, most repricing is done, and the yield rise is driven mainly by improving activity and earnings rather than runaway inflation expectations ^[50]; CICC decomposes the 10-year into a term-premium component and an implied policy expectation well above the current policy rate, implying more than two hikes are already discounted ^[45]; El-Erian (social, unverified) attributes the move less to the Fed and more to an ex-ante imbalance between demand for bond financing and its supply ^[52]; Chris Whalen (single source, opinion) argues bond vigilantes have seized pricing power and the Fed has lost control of long-term rates ^[53][54]. On the relationship between assets, Soochow’s market pricing shows a wide divergence in the number of hikes discounted over the next year across instruments, with rates pricing several and equities far fewer ^[45].
• [NEW] Rates and volatility tape: Monday’s session saw the 2-year down 3.3bp to 4.610% and the 10-year down 1bp to 4.964% while the 30-year rose 0.8bp to 5.362%, across a week in which the 2-year rose 26bp to 4.63%, the 10-year 18bp to 4.96% and the 30-year 11bp to 5.35% ^[18][41]. The 10-year TIPS real yield rose to 2.60%, breakevens sit near 2.37%–2.40%, the MOVE index jumped to 82.2 and VIX eased to 15.8 ^[55][41][38]. The dollar index closed the week at 99.09, down 0.07% ^[41]; BofA attributes recent dollar softness partly to global central-bank curve repricing eroding its rate-differential advantage ^[14], while Barclays notes the dollar premium remains high with sentiment negative, so only a clearly dovish FOMC outcome unlocks further weakness ^[44].
• [NEW] Commodities, gold and equities: Brent rebounded above $107 after a Saudi East-West pipeline attack, with roughly 4% of global oil supply reported at risk, and US futures fell with Nasdaq 100 futures down 1.2% amid renewed concerns over the pace of AI development ^{[56][36][57][58]}. Gold slipped toward $4,340/oz for a third weekly decline ^[59][39]; the week’s London range was narrow and the weekly loss modest ^[60]. Copper fell on the LME, its first weekly decline since June ^[61]. US indices closed the week lower with the rate-sensitive small-cap index materially weaker while semiconductors outperformed ^{[42][41][60][36]}.

4. Global Central-Bank Linkages

• [ONGOING] ECB: the Governing Council unanimously raised all three key rates by 25bp to a 2.5% deposit rate on Sept 10, with a statement describing inflation as “well above target for an extended period” — language last used in the 2022–23 hiking cycle ^[37][41].
• [ESCALATED] BOJ: the market widely expects the policy rate to rise from 1% to 1.25% at the Sept 17–18 meeting ^[62][63][44]. Japanese yields have moved with it, the 10-year JGB near 3% and the 30-year around 4% ^[62][41], and yen carry-trade unwinding is described as accelerating, with the yen the week’s strongest currency ^[41]. Barclays forecasts hikes in September 2026, January 2027 and July 2027 to a 1.75% terminal ^[51], while Nomura is explicitly sceptical that the BOJ will accelerate, seeing limited further yen upside given crowded speculative long positioning ^[55][64].
• [NEW] PBoC: the central bank signalled an expanded macroprudential and financial-stability role and a review of macroeconomic and financial work, and its deputy governor set out plans to improve the reserve system and guide short-term money-market rates to run more smoothly around the policy rate — social/unverified relays for the mandate items ^{[65][66][67][37]}.
• [NEW] Others: the BOE is expected to hold ^[63]; Barclays forecasts a 12.5bp hike from the Bank of Canada on Sept 17 while consensus expects no change, and favours long AUD and SEK against a bearish NZD ^[44]; the next global “super central bank week” lands in late October with the Fed on Oct 27–28, the ECB on Oct 29, the BOJ on Oct 29–30 and the Bank of Korea on Oct 22 ^[41].

5. Asset Implications

Quadrant	Current probability tilt	Key asset implication	Anchoring narrative
Growth↑ + Inflation↑	Rising	Energy/commodities and TIPS stay the cleanest expressions — oil back above $107 on a fresh supply event and UMich inflation expectations rising — while nominal long bonds are blocked by a term premium restored to pre-GFC levels and a 30-year near 5.35%	§3 / §2 / §4
Growth↑ + Inflation↓	Falling	The “one-off calibration” path: equity strategists are turning bullish on the view that stronger growth absorbs the hike, and Goldman’s probability-weighted path is more dovish than market pricing — the expression is quality and cyclicals rather than duration	§1.2 / §2 / §3
Growth↓ + Inflation↑	Rising (tail)	The supply-shock leg is live and policy-insensitive — a pipeline attack putting a meaningful share of global supply at risk, with the inflation pass-through landing after the August survey period, so nominal bonds hedge neither leg; gold’s rate-insensitive official and allocation bid is the hedge, not duration	§2 / §3 / §4
Growth↓ + Inflation↓	Falling	The hold tail is expensive: BofA estimates a surprise hold pushes the 30-year toward 5.75% with buybacks unable to stem it, which is why the short-covering setup is a trade, not a thesis	§1.2 / §3

Stock-bond correlation call: the regime remains the inflation/policy-driven positive-correlation configuration that is hardest for risk parity, but this week’s tape carries an important distinction. Monday’s move was front-end-led with the 30-year rising — a term-premium and supply signature, not a growth shock — while the week’s cross-asset pattern (equities down, oil up sharply, gold down on the week) is the classic inflation-and-policy signature rather than a growth-driven one ^{[18][41][56][59]}. The joint selloff risk is structural: the long end is being set by a term premium that three different houses attribute to three different causes (pre-GFC term premium restoration, an ex-ante supply/demand imbalance in bond financing, and a still-elevated neutral rate), and none of those are resolved by a single 25bp move ^[51][52][45]. The escape hatch is narrow and event-specific: if Warsh frames the hike as a one-off and the median dot shows only one 2026 hike, the policy-uncertainty component of the term premium can compress and long duration can partially recover its hedge value ^[8][11]. If the dot median shifts to two hikes and the tone is hawkish, both legs of a risk-parity portfolio are repriced by the same variable again — and the wide cross-asset divergence in how many hikes are discounted means equities would have the most to catch up on.

Risk-budget implication: underwrite the FOMC as a two-sided density rather than a directional bet, and do not treat 86–90% pricing as a signal. Keep commodity and energy risk at or above benchmark: energy is the only sleeve whose driver is genuinely orthogonal to the Fed’s reaction function, and the marginal supply event this weekend reinforced that. In rates, stay underweight long-end nominal duration into the decision but pre-commit to rebuilding it on a calibration read, and prefer curve and relative-value expressions to outright shorts given the two-sided term-premium risk and the documented extreme short positioning that makes a dovish surprise a squeeze risk. Recognize that the demand base for duration is thinning — custodial holdings and foreign official participation are both falling — so any duration re-build should be funded in the belly rather than the 30-year. In credit, keep duration short and favour spread products over pure duration; private-credit growth confirms conditions are still loose, so carry is not yet the problem, but the refinancing and AI-financing pipelines are. In equities, retain moderate beta tilted to quality and cash-flow durability, and treat the wide gap between the number of hikes equities and rates are discounting as the most likely source of a catch-up repricing. Hold gold as the structural fiscal and dollar-credit hedge, sized for continued real-rate whipsaw, given that its marginal buyer has shifted away from rate-sensitive speculative money — it is not equity insurance.

6. Contrarian & Tail Risks

• Consensus fragility: the 86–90% pricing is a market construct whose meaning is unpriced, and Goldman’s own note concedes the hike is being justified by the market’s expectation of the hike rather than by fundamentals — a circular validation that would be exposed by a weak dot plot or a soft opening statement. The falsifiable pillars are narrow: (1) the entire core exceedance rests on a concentrated wireless-services repricing plus volatile airfares and lodging, and the core PCE conversion is only modestly above the threshold Waller has used to justify dissent; (2) the SEP is the swing variable, with Goldman and Citi expecting one 2026 hike, BofA, HSBC and Huatai expecting the median to imply two, and one of those being wrong; (3) the one-and-done precedent is thin — since the 1990s the Fed has delivered only one single-and-stop hike, in 1997, after which it held for 18 months before cutting; (4) the forecasters who flipped did so within days of the print, and half the bank table expected a hold as recently as last Thursday, which is a measure of how little independent information the flip contains; (5) Barclays itself flags an expectation gap between its hike call and a market consensus of no change in its FX publication, and Financial Street Securities still leans toward a hold being marginally favoured, while Morgan Stanley argues the market has over-extrapolated; (6) the Chicago Fed work suggesting dot-plot surprises move market rates only marginally cuts both ways — the dot plot may matter less than the market believes, but a hawkish dot with no market response would leave the term premium to do the tightening.
• Second-order transmission: three chains are not in market pricing. The energy pass-through chain has a fresh input — a pipeline attack putting a meaningful share of global supply at risk — and its inflation impact will land in September and October data rather than the August survey period, which means the Fed would be hiking into an impulse rather than ahead of it; CICC’s counter-argument is that the August drivers were transient and that year-end inflation falls unless oil holds above roughly $95, but that threshold is close to current levels. The fiscal channel is the structural tail: US Treasury debt above $40tn with annual interest expense above $1tn, roughly a third of personal income tax receipts already going to debt service, a Treasury actively shortening duration while buying back the long end (in effect distorting the curve rather than clearing it), and a Treasury Secretary publicly positioned as an active participant in currency and bond markets — none of which a 25bp policy move addresses. The AI-financing channel is the equity transmission vector: the five largest cloud providers’ long-term debt has roughly doubled since early 2025, AI investment is already a large share of sequential GDP growth, and the crowding-out of private long-duration financing is itself one of the named drivers of long-end yields; if long yields approach 5.5%, the marginal growth project’s hurdle rate rises. Layered on top, the political channel is active — open presidential pressure for the world’s lowest rates, threats to cut trade ties if the Fed does not cut, the pre-midterm optics of hiking, and a 2027 voter cohort expected to be more neutral — which raises the bar for follow-through rather than for the September move itself.
• Source quality control: hike odds remain a band across instruments, vendors and snapshot times — 86.2% (CME FedWatch), “close to 90%” (futures, Reuters), 88% (fed funds futures versus 62% the prior Monday), 87.3% (CME, two separate citations), ~85% (BofA’s read versus 35% two weeks earlier) — so sequencing matters more than any single reading. Single-source or unverified items include the Chicago Fed dot-plot paper relayed through a social post, El-Erian’s characterisation of the yield surge, Chris Whalen’s “bond vigilantes” thesis and his claim that the Fed has lost control of long-term rates, the PBoC mandate items relayed by Financial Juice, the Goldman-sourced suggestion that Warsh may seek political cover, and the Trump quotes relayed by Gelonghui and Financial Juice. Two items in the batch carry no extractable content and should not be treated as evidence: the Christophe Barraud post, which contains only a title and link to a Chicago Fed Letter, and the Investing.com item, which is a question rather than a report. Direct conflicts should be read side by side: BofA describes CPI-day curve behaviour as a real-yield-driven twist flattener while Monday’s tape was front-end-led with the long end selling; the long-end driver is variously attributed to term-premium restoration, supply-demand imbalance in bond financing, and fiscal/credibility concerns; and BofA’s forecast that a hold would push the 30-year to 5.75% sits against JPMorgan’s view that most term-premium repricing is already done. Finally, the “17 of 20 banks” figure and the poll tallies are snapshot counts that have changed within days and should be treated as sentiment, not as information.

---

Appendix: Additional Sources

• ^[3] Financial Juice (poll relay, social) — Fed to hike at least twice by end-March 2027: 37 of 70 economists
• ^[4] Financial Juice (poll relay, social) — 86 of 101 economists expect a hike to 3.75%–4.00% on Sept 16
• ^[68] Morgan Stanley via Gelonghui — quality and cyclicals outperform after a first hike
• ^[12] Jefferies (Mohit Kumar) — Fed will hike less than the 3.5 hikes priced by forwards
• ^[18] Gelonghui — Treasury curve close, 2Y/10Y/30Y session levels
• ^[67] Financial Juice (social) — PBoC to intensify review of macroeconomic and financial work
• ^[69] Goldman Sachs, “Fed Chatterbox: September Edition” — divided committee; Hammack, Waller, Williams, Barr, Schmid, Kashkari and Warsh positions
• ^[34] Commerzbank Research — inflation data reinforces hike expectations; fed funds futures 88% vs 62%
• ^[70] Financial Street Securities — leans toward September hold being slightly advantaged
• ^[71] Investing.com (social) — question post only, no reported facts

---

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