Hike odds leap to ~70% on an above-forecast PPI and $100+ oil; today's CPI is the final word

0. Weekly Arc

The arc from Jackson Hole to today has been a one-way hawkish grind punctuated by two dovish interruptions that did not hold. Chair Warsh’s Aug 28 keynote set a “hike unless data excuse it” default; Governor Waller’s conditional-hold remarks halved September odds to a coin flip; the Sept 4 payrolls beat re-leaned them toward ~60%; and this week an above-forecast PPI, oil above $100 and a hawkish ECB hike pushed pricing to ~70%, collapsing the September decision entirely onto Friday’s CPI.

1. Policy Narrative & Expectations

The past ~24h delivered the sharpest hawkish repricing since Jackson Hole. September hike odds have moved from roughly 60% to a 70–73% band — CME FedWatch at 71.8% ^[1] and a 28.7%/71.3% hold-hike split ^[2], Bianco Research marking a 73% close ^[3], and multiple CME- and futures-based reads at ~70% ^{[4][5][6][7][8]} — after August PPI accelerated above expectations and Brent broke back above $100 ^[9][5][10]. The move is no longer about September alone: traders have fully priced a first hike by October at the latest ^[11][12][10], attach close to 60% to a second increase in December ^[5], and market prices now imply the Fed will almost certainly have delivered one, if not two, hikes by year-end ^[13]. The marginal driver is the data-and-energy channel rather than a new Fed voice: a hot PPI, Brent’s surge and a hawkish ECB hike compounded into the repricing ^{[9][14][4][15]}. Set against that is an unusually explicit divergence between markets and forecasters — a Bloomberg survey found a majority of economists expect the Fed to hold at this month’s meeting and through end-2027, citing tempering inflation and the proximity of the midterms, with only 13 of 48 economists surveyed Sept 4–9 expecting a hike this month ^[16][17]. The September decision has therefore narrowed to a single question: whether today’s core CPI validates the hawkish repricing or pushes back against it ^{[18][14][19][20]}.

1.1 FOMC Officials’ Remarks

No sitting FOMC official spoke in the past 24h; the only fresh voice on the committee’s reaction function came from a former governor.

• [NEW] Dovish (former official — single source / unverified): Stephen Miran, former member of the Federal Reserve Board of Governors and presented by Larry Kudlow — “The Fed has backed themselves into a place where policy is ultimately going to be a function of the next inflation print…. As the inflation data have been getting better, the Fed has bizarrely been getting more hawkish…. It’s very confusing.” He argued a hike “tells America that the Fed really doesn’t respond to data” ^[21].
• [ONGOING] Hawkish: Chair Kevin Warsh — no fresh remarks; the Jackson Hole frame still anchors the debate, with the Fed to “have work to do” unless inflation falls toward the 2% target at a “sufficient speed,” and the PCE deflator reaffirmed as the official yardstick ^{[22][23][10][5]}.
• [ONGOING] Neutral/swing: Governor Christopher Waller — the operative recap remains his earlier-month remarks that he would consider a hike “if inflation comes in hot” but would be willing to hold if August CPI shows progress toward 2% ^[22][23].

1.2 Policy Signals & Institutional Communication

• [NEW] Deutsche Bank’s AI scoreboard shows a unified hawkish committee: its proprietary tool scores the average sentiment of all Fed officials up 0.6 points to 7.2 since July, a record high since the sample began in October 2021 and above the pandemic-era peak; 18 of 19 officials sit above the neutral 5.0 line. Hammack scores 9.8 (the highest in the dataset), Collins rose more than a point to 8.6, and Warsh scored 8.2 overall and 8.8 on the Jackson Hole speech; Waller is the only voting member whose score fell (to 6.5, on a “hawkish pause” stance), while Powell (5.5) and Bowman (3.8) are stale because neither has made monetary-policy comments since May. The implied trigger: a core PCE print more than 20bp above expectations would tilt the committee to hike ^[24].
• [NEW] Market vs forecasters — an explicit split: the Bloomberg survey’s hold-through-2027 consensus directly contradicts ~70% market odds for a hike this month ^[16][17], while Fed watcher outlooks are described as inertial — only three banks (UBS, Barclays, SocGen) formally shifted toward a September hike after Jackson Hole, and half now see no cuts this year ^[25].
• [NEW] The communications-framework debate is now formal: Goldman Sachs chief economist Jan Hatzius argues the 40-year transparency revolution has been positive because a clearly communicated reaction function accelerates transmission, and that weakening transparency risks more speculation, unhelpful volatility and delayed pass-through; former governor Stephen Miran supports reducing transparency, argues forward guidance made markets data-insensitive (citing post-pandemic housing over-stimulation and SVB’s duration exposure), and advocates abolishing the policy dot plot while keeping the economic projections; former Vice Chair Donald Kohn takes the middle, criticizing over-reliance on the SEP median and warning that a complete absence of “narrative” would cause market confusion and a lack of accountability ^[26].
• [NEW] PCE methodology change: the BEA’s Sept 30 release will adjust price measurement for legal services, computer software and investment advisory; multiple economists expect the changes to produce lower PCE readings, and Citi estimates the core PCE methodology revision could cut the y/y rate by about 35bp ^[10][27].
• [ESCALATED] The policy threshold has hardened on supply-side inflation: officials have explicitly signalled readiness to raise rates again if inflation does not improve in the near term, but the dominant price drivers this year — tariffs and energy — are historically insensitive to rate changes, leaving the Fed’s main tool with limited traction ^[7][28]. Deutsche Bank warns markets to guard against more aggressive tightening than expected ^[24].
• [ONGOING] Decision mechanics: today’s CPI is the last major inflation input before the Sept 15–16 meeting, and the PPI components feed directly into the PCE calculation the Fed targets ^{[18][4][13][5]}.
• [ONGOING] Political pressure: President Trump has revived his pressure campaign against the central bank, pressing his handpicked chairman Kevin Warsh to act ^[29].

2. Key Data & Market Read

• [NEW] August PPI (released 9/10): came in above expectations and accelerated sharply from July — the largest gain in three months, driven by energy — with transportation and warehousing, hospital services and airfares all jumping, undoing some of the recent inflation deceleration ^{[9][5][30][10][13]}. Market read: 10Y yields broke above 4.9% and September hike odds jumped from roughly 62% to ~70% within the session ^[31][5]; traders extrapolated CPI from the published PPI components ^[11]. Offsetting detail: core PPI excluding food and energy rose only modestly and slightly below forecast, the mildest aspect of the release ^[32][10].
• [ESCALATED] August CPI (due 9/11 08:30 ET): consensus is centred on a modest core rise, but the policy payload is asymmetric — a core print at or above 0.3% m/m would essentially confirm a September hike, while ~0.2% is neither an inflation shock nor enough to convince the Fed that inflation is returning to 2% fast enough, and a print at or below 0.1% would make a hold the base case ^[33][34][14]. El-Erian frames the same line: above 0.2% would “rattle yet again market rate expectations,” below 0.2% would moderate them ^[35]. Forecaster dispersion is narrow but consequential (BofA and Barclays firmer, Citi and JPMorgan softer), and JPMorgan’s five-scenario S&P 500 map is keyed symmetrically on the core m/m threshold, with cooler prints read as more favourable for risk assets ^{[33][36][37][27]}.
• [NEW] The CPI/PCE conversion is the real battleground: with the PPI print in hand, BofA’s Stephen Juneau tracks core PCE at a monthly rate that rounds up rather than down, which he says should “greenlight a hike” at next week’s meeting ^[5][23]; Julian/Citi’s milder core CPI and core PCE estimates would instead keep the Fed on hold ^[27].
• [NEW] Weekly claims and PPI details: weekly unemployment insurance claims indicated the job market remains stable, and jobless claims tracked low enough that BMO and TradeStation read the data combination as making it hard for the Fed not to hike ^[13][5].
• [ONGOING] August payrolls (released 9/4): the far-above-consensus print and its upward revisions remain the reason the labour market is no longer treated as a constraint, with markets pricing roughly 15bp of a hike for September after it ^[6][34][38].

3. Financial-Conditions Signals

• [ESCALATED] Rates: the 10Y pushed to ~4.94–4.97% intraday, closing in on the 5% level — its highest since late 2023 ^{[39][40][6][14][41]}; the 30Y touched 5.37% intraday, the highest since 2007 ^[9][42][43], and today’s $22bn 30-year auction cleared at 5.308%, the highest winning yield since 2001 ^[9]. The 2Y rose sharply, its largest single-day move since the April 2025 tariff shock ^[44][8]. Global long ends moved in tandem: UK gilt yields at their highest in nearly 30 years, French long yields at more than 20-year highs, Australian yields at 15-year highs, New Zealand swaps up 22bp and the Japanese benchmark up 9bp ^[8][14].
• [NEW] The long-end driver is real rates and term premium, not policy expectations: BofA attributes most of the adjustment to real rates rather than inflation expectations, with the US deficit projected at about 6.5% of GDP next year and interest expense at roughly 3.3%; TS Lombard’s Dario Perkins argues the term premium holds all the upside in an era of supply shocks where bonds’ hedging properties structurally erode; Deutsche Bank’s Jim Reid instead frames the move as a long normalization, noting the average 10Y yield since 1800 was 4.5% ^{[45][46][47][48]}. One-year US inflation swaps are nudging 2.70%, up almost 100bp since the start of August ^[8].
• [NEW] Treasury buyback underdelivers: the first expanded 10–20Y operation fell short of its announced cap (roughly $5.2bn against a $6bn ceiling), and the shortfall triggered fresh selling rather than calming the market; BofA frames the buyback as “more signal than liquidity impact” relative to a market segment above $1 trillion, argues cutting long-end issuance (for example scrapping the 20-year) would be more effective, and says the programme’s price-influencing intent is inherently in tension with Chair Warsh’s desire to reduce forward guidance and learn from market signals. Secretary Bessent called the criticism “nonsense” ^{[44][49][45][4][50][51][27][52]}.
• [REVERSED] Dollar: the recent weakening reversed — the Bloomberg Dollar Spot Index rose as much as 0.4%, its largest intraday gain since Aug 28, on the PPI print and the oil surge ^[30]; BMO’s Mark McCormick says the market underestimates inflation and its effect on rates, the dollar and risk appetite, and warns financial conditions will likely tighten further in the autumn ^[30]. Offsetting structural narrative: HSBC reads the year’s dollar weakness as a credibility problem — widening fiscal deficits, uncertainty over the Fed’s reaction function and political interference such as joint yen intervention — while cautioning there is still no conclusive evidence of a large-scale strategic shift out of dollar assets ^[53].
• [NEW] Credit — the AI complex is the stress point: Gundlach reports AI-related investment-grade spreads widening from about 50bp to about 125bp and AI-related high-yield from about 180bp to about 325bp, while non-AI high-yield spreads sit near this year’s historic lows ^[54]. El-Erian notes remarkably limited spillover from rate risk to credit risk so far ^[55]. On the household side, the average 30-year fixed mortgage rate breached 7% for the first time since May 2025 ^[50][56].
• [ESCALATED] Commodities as the inflation transmission belt: Brent settled above $107 and WTI above $102 — oil up more than 18% in September and more than 50% from July lows — with the Strait of Hormuz still effectively shut and Bab al-Mandab at risk; Goldman’s Daan Struyven sees rising risk of a spike above $120, and BofA flags diesel rather than crude as the real-economy pain point, with the diesel-crude spread at a record ^{[44][14][4][57][43][58]}.
• [ONGOING] Flows: US equity funds saw a third consecutive weekly outflow (the largest three-week outflow since January 2026) while global bonds, cash and money market funds kept attracting inflows — defensive positioning rather than panic ^{[42][36][43][59]}.

4. Global Central-Bank Linkages

• [NEW] ECB: delivered the quarter-point hike to 2.5% on 9/10, which President Lagarde called a “no-brainer,” adding that inflation will be “longer lasting” than she and her colleagues had anticipated; the ECB raised some inflation forecasts and warned inflation is set to remain well above target for an extended period, a shift from its more benign July energy assessment ^{[4][8][60][5][15]}. Markets fully price three more quarter-point hikes by mid-2027 and see the deposit rate ending the year at 2.75%, while ECB officials expect further tightening with another hike possible as soon as next month ^[8][15][61]. BofA judges the ECB more hawkish than it expected but less aggressive than the market assumes, and warns that without clearer communication euro front-end rates and 10Y yields remain hostage to energy prices ^[49]. Cross-asset read: the ECB hike added fuel to the global bond selloff ^[15][60].
• [ESCALATED] BOJ: Citi expects a 25bp hike to 1.25% at this week’s meeting, followed by three more (January, June and December next year) to a 2% terminal, and explicitly flags that the path depends on the Fed — a Fed hike this year would pull the next BOJ move forward to December. Citi calls market pricing of ~2.2 hikes by January overly aggressive ^[62]. HSBC’s OIS pricing shows a 100% probability of a 25bp hike on Sept 18 with 75bp cumulative through April 2027 ^[53]. Separately, sources say the BOJ may signal readiness to accelerate rate hikes and holds no fixed view on the terminal rate — single-source / unverified ^[63]. On flows, GPIF’s roughly ¥320tn asset base means a 1pp increase in the domestic bond allocation cap implies about $20bn of flows ^[53].
• [ONGOING] PBoC / reserve diversification: no material new PBoC signal in this batch beyond the yen-unwind and dollar-credibility framing ^[64][53].
• [NEW] Others: Norway’s sovereign wealth fund (NBIM) proposes cutting the government-bond share of its fixed-income benchmark from 70% to 50% under a market-cap-based country framework, a change BofA estimates could lift the JGB allocation to about 7.4% — roughly ¥2.8tn of incremental demand — flattening the JGB curve ^[49]. BofA also flags residual South African Reserve Bank hike risk on elevated inflation expectations and possible Fed tightening, and the Reserve Bank of India’s challenge absorbing a liquidity surge after stabilizing the rupee ^[64]. Barclays holds EM resilient for now but sees Q4 stress tests from US rate volatility, higher fuel prices, rising European bond supply, Brazil election uncertainty and Fed repricing ^[65].

5. Asset Implications

Quadrant	Current probability tilt	Key asset implication	Anchoring narrative
Growth↑ + Inflation↑	Rising	Energy/commodities and TIPS remain the cleanest expressions — Brent above $107 with a record diesel-crude spread; nominal long bonds stay blocked as real rates and term premium, not inflation compensation, drive the selloff, and the buyback backstop has proven weaker than advertised	§2 / §3
Growth↑ + Inflation↓	Falling	The disinflation path is losing its pricing: resilient activity plus renewed inflation risk have removed near-term easing expectations entirely, so front-end/belly carry is now a bet against the CPI print rather than with the policy path; equities’ AI-earnings insulation is the only offset	§1.1 / §2 / §3
Growth↓ + Inflation↑	Rising (tail)	The supply-shock stagflation leg is the live tail — oil, tariffs, chip shortage and AI capex demand are all rate-insensitive, so hikes cannot restore bonds’ hedging properties; within equities the high-free-cash-flow, low-leverage factor is the historical archetype, and gold remains the structural hedge rather than long TIPS	§2 / §3 / §4
Growth↓ + Inflation↓	Falling	The hold-camp path — a soft core CPI plus the Sept 30 BEA methodology revision mechanically lowering measured PCE — would unwind the ~70% pricing; the shunned duration complex and the 5% zone on the 10Y are where the risk budget would be rebuilt	§1.2 / §2

Stock-bond correlation call: the regime is decisively the inflation/policy-driven positive-correlation configuration — the hardest for risk parity. Thursday was a joint selloff: the 10Y jumped toward 5%, the 30Y hit its highest since 2007, the 2Y posted its largest single-day move since April 2025, US equities fell, and small caps and materials led the decline, i.e. equities and bonds are being repriced by the same policy-and-energy variable rather than by differentiated growth news. The term-premium and supply channels reinforce this: most of the long-end move is real rates and term premium rather than inflation compensation, which is precisely the configuration in which duration fails to hedge equity drawdown. The escape hatch is narrow and data-contingent — a soft core CPI that lets the Fed hold, or a hike explicitly framed as a one-off credibility repair, would compress the policy-uncertainty premium and partially restore medium-duration hedging value. Gold’s role has shifted structurally: it is a hedge against dollar-credit and fiscal risk, not equity drawdown, and CFTC positioning (net futures at 55% of open interest) shows it is already crowded as a macro hedge.

Risk-budget implication: Under this correlation structure, underweight long-end nominal duration into the CPI print and express the rate view through curve and relative-value structures rather than outright short-duration risk — BofA’s recommended expressions are long CAD 5s30s steepeners against short US equivalents to capture the Treasury’s active long-end management, long 30-year ASW spread, long 10y20y inflation swaps and a long 20-year UST fly targeting the rich long belly. Overweight commodities and energy as the cleanest expression of an oil-led supply shock that policy cannot offset, and keep the commodity/TIPS sleeve sized for a Brent path toward $120 rather than a reversion. Hold gold as the structural fiscal/dollar-credit hedge but respect its crowding and the real-rate headwind. In credit, keep risk short-duration and favour non-AI over AI-linked issuers — the AI complex is where spreads have already blown out and where supply keeps arriving “like an avalanche.” In equities, retain moderate beta with the AI-earnings-insulated quality tilt while trimming rate-sensitive small caps and materials, and treat the high-free-cash-flow, low-leverage factor as the stagflation-consistent core. Finally, keep a convex duration option — not a core position — on a soft core CPI print, since the intermediate and belly segments are the least crowded and would see the largest forced repositioning.

6. Contrarian & Tail Risks

• Consensus fragility: the ~70–73% hike pricing is a market construct, not a Fed commitment — Timiraos’ point is that Warsh’s Jackson Hole speech convinced investors a hike was more likely without naming a trigger, so the market filled the information gap with a magnitude of hikes Warsh never promised; a weak CPI would throw the assumption into chaos. Deutsche Bank’s own scoring tool implies the committee only leans to hike if core PCE exceeds expectations by more than 20bp, and BofA’s threshold analysis puts the hold-versus-hike line near a core PCE monthly rate of about 0.25%. Externally, the Bloomberg survey’s economists disagree outright, expecting a hold through end-2027, and only 13 of 48 expect a hike this month. Falsifiable pillars: (1) core CPI lands at or below the consensus core path rather than at 0.3%+; (2) PPI details do not continue to feed through into core PCE; (3) oil does not hold above $100 into the FOMC; (4) the dollar’s response is asymmetric — a soft CPI print would trigger a larger selloff than the rally a hot print delivers, and a Fed that fails to follow through after hot data would damage credibility and hit both the dollar and long-end Treasuries simultaneously; (5) the buyback programme’s failure to reach its cap is read as a signal about the limits of official intervention rather than a one-off execution shortfall.
• Second-order transmission: the fiscal-dominance chain is the live structural tail. Long-end yields are being set by real rates, term premium and supply, with the deficit projected near 6.5% of GDP and interest expense still growing — BofA argues intervention impairs price discovery, weakens the information policymakers are trying to extract, and risks moral hazard, and concludes fiscal dominance “has not fully arrived but is no longer far off.” The Treasury’s attempt to influence long-end prices is internally inconsistent with a Fed chair trying to reduce forward guidance while learning from markets. The AI-financing channel is the equity transmission vector: AI-related IG and HY spreads have already blown out while non-AI credit sits near year lows, AI cloud giants are spending over 90% of operating cash flow on infrastructure, and a 50–75bp rise in data-centre financing costs is judged insufficient to make them scale back — so the marginal growth project’s hurdle rate keeps rising. Beyond that: the 30-year mortgage above 7% is the household transmission belt, and the political-fiscal coupling is newly active — a pledge of $5,000 per adult if Republicans hold Congress, estimated at $1.2–1.3tn of extra spending, which markets are brushing aside as improbable against a 6% deficit, solid growth and elevated sovereign yields. Finally, the midterm elections are flagged as the channel through which US gasoline prices and equity performance transmit to EM.
• Source quality control: hike odds are a band across instruments and snapshot times — 71.8% (CME FedWatch, Thursday afternoon), 71.3% cumulative 25bp per a CME read, and a 73% close per Bianco Research, alongside ~70% in CNBC, LSEG and Reuters reads and a ~68% implied probability per HSBC’s OIS framing — whether pricing is quoted pre- or post-PPI drives the difference, so sequencing matters more than any single number. Evidence of genuine conflicting signals: BofA expects the ECB to be more hawkish than it forecast but less than markets price, while the market fully prices three further hikes; the same batch has a Bloomberg survey of economists forecasting a hold through 2027 against ~70% market odds for a hike this month. Single-source or unverified items include former governor Miran’s remarks (social relay), the BOJ’s supposed readiness to accelerate hikes (“sources say”), the deerpointmacro OIS repricing posts, and Financial Juice headlines. Several fact attributions in this batch compress time (Waller’s and Warsh’s quotes date to earlier-month speeches and Jackson Hole, not to today) — recaps should not be mistaken for fresh guidance. Do not treat Gundlach’s Shiller-PE or model-derived fair-value figures as market prices.

---

Appendix: Additional Sources

• ^[66] WSJ — investors await CPI; yields retreat a little after threatening 5%
• ^[58] WSJ — oil pulls back after closing above $107; Treasurys around multiyear highs
• ^[67] Jin10 — historical angle: a single Fed hike has not killed equity bull markets (headline only)
• ^[68] Bloomberg — global bond selloff pushes Treasury yields toward a key level
• ^[69] WSJ — gold rises in early Asian trade
• ^[70] Reuters — US stocks end lower as PPI and oil stoke hike worries
• ^[71] Jin10 — “rates near 5%, but the bond market hasn’t told the whole truth” (headline only)
• ^[57] Bloomberg — Asian stocks and bonds set to decline after the US selloff
• ^[21] Nick Timiraos / Stephen Miran — former governor’s dovish case (social relay, unverified)
• ^[20] @deerpointmacro — hawkish OIS repricing to late-July peaks; limited further scope

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