〔Overnight Brief〕Last Hike Was the Only Hike, Natixis and Citi Tee Up September CPI: Bond Selloff Marker, 64% Sweep Odds, $166.4B Cash Hoard Refuse to Cooperate

0. Weekly Arc

The dovish repricing that began with August's soft PCE (headline 0.3% m/m vs 0.4% expected, 3.4% y/y vs 3.7%; core 0.2% m/m vs 0.3%, 3.0% y/y vs 3.3%) and September's 29k nonfarm print (vs 90k consensus, two prior months revised down by 60k) is now being institutionalized on the sell-side ^[1]. Natixis called the September CPI in line and labelled last month's hike the cycle's only one ^[2], while Citi framed a 'dovish surprise' around core inflation approaching 2% ^[3]. Yet the cross-asset picture refuses to confirm: a long-dormant fixed-income selloff marker reasserted itself overnight ^[4], and Hartnett's cash-gathering trade hit its fastest pace since April 2020 ^[5]. Net: a rates-dovish, positioning-bearish split, with tomorrow's CPI as the arbiter.

1. Fed and CPI: One-Hike Consensus Takes Shape

• **[NEW] Natixis:** September CPI should print in line; last month's Fed rate hike 'may be the only one this cycle' ^[2].
• **[NEW] Citi:** positioning for a 'dovish surprise' from the Fed, with core inflation near 2% cited as the key ^[3].
• **[ESCALATED] Dovish data inputs:** August PCE printed below consensus on every line; September nonfarm payrolls at 29k vs 90k expected, with two prior months revised down by a combined 60k ^[1].
• **Source-quality flag:** the 'one-and-done' framing is a single-broker view from Natixis ^[2]; Citi's 2% core trigger is a call, not yet data ^[3].

2. Midterm Positioning: The Cash Trade Accelerates

• **[NEW] BofA / Hartnett:** November 3 midterm framed as the largest potential catalyst for ±10% global equity moves; probability of a Democratic sweep of both chambers rose to 64% from 49% in September, with US equities flagged for >10% downside if realized ^[5][6].
• **[NEW] EPFR flows via BofA:** $166.4B into money market funds in a single week, the largest since April 2020; US Treasuries saw $16.6B of net inflows, the largest since April 2025; financials posted their biggest weekly outflow ($3B) since March 2026 ^[5].
• **[NEW] BofA Bull-Bear Indicator fell to 8.1 from 8.8, still in 'sell' signal territory ^[5].
• **[ESCALATED] Global MMF AUM climbed to $8T from $5T in 2023 ^[5].
• **[NEW] Baird Strategas / Verrone:** markets remain 'highly rotational'; bond yields still not 'truly competitive for your capital every day' ^[7].
• **Source-quality flag:** the 64% Democratic-sweep probability is a single-sourced market-implied number; the $166.4B inflow figure is BofA-cited EPFR data, also single-source ^[5][6].

3. Cross-Asset and Bond Signals

• **[NEW] Bond market:** a long-dormant fixed-income selloff marker reasserted itself overnight, prompting fresh Wall Street chatter about what it means for duration ^[4].
• **[NEW] Deutsche Bank:** crude/refined-product divergence is widening; oil price forecast raised ^[8].
• **[NEW] Partners Group / Amoroso (Bloomberg):** described it as a 'phenomenal time to be an investor, not a consumer,' while flagging near-term volatility risk and rising global yields ^[9].
• **[ONGOING] Q3 cross-asset review (per broker synthesis):** 10-year US Treasury yields broke above the 5% threshold; China 10-year CGB yields fell below 1.7% to 1.68% at quarter-end; international crude approached $110/barrel; gold pushed higher before retracing ^[10].

4. Single-Stock Calls: SpaceX, Cloudflare, OpenAI Methodology

• **[NEW] Morgan Stanley on SpaceX:** 'Tera'Merica' thesis — reindustrialization beyond chips into rockets, satellites, ground stations, turbines, autonomous vehicles, industrial/humanoid robots, ships and drones — extends SpaceX's TAM; Overweight retained, $300 target ^[11].
• **[NEW] JPMorgan on SpaceX:** plan to acquire national low-band wireless spectrum from Grain Management complements the mid-band EchoStar deal; JPMorgan reads the move as extending SpaceX's wireless ambitions beyond phones into IoT, autonomous driving and robotics ^[12].
• **[NEW] Oppenheimer on Cloudflare:** target raised to $430 from $380, Outperform maintained, on accelerating security-product adoption and AI-agent tailwinds; expects Q3 results above company guidance ^[13].
• **[NEW] OpenAI revenue disclosure:** ~$50B annualized run-rate as of end-September — below the widely cited $70B figure, the gap explained by OpenAI excluding partner-channel sales (Anthropic includes AWS/Google Cloud distribution); Q3 y/y +77%, enterprise +107% ^[14].
• **[NEW] Micron Q4 (fiscal):** record revenue of $54.2B, +31% q/q and +379% y/y; gross margin 87%; 26 long-term agreements covering ~$150B in locked-in orders; 75% of 2027 capacity already sold ^[14].

5. China A-Shares: Post-Holiday Selloff, Domestic Resilience Watch

• **[NEW] Post-holiday tape (October 8 close):** Shanghai Composite -0.79% at 3811.90, Shenzhen Component -2.07% at 12620.90, ChiNext -3.15% at 3036.66; turnover rebounded toward ~RMB1.7T; petrochemicals, utilities and transport led; electronics, telecoms and machinery lagged ^[15].
• **[NEW] September PMIs all back above 50:** manufacturing 50.1%, non-manufacturing 50.2%, composite 50.7%, with production-led character and demand durability still to be confirmed ^[1].
• **[NEW] National Day demand:** cumulative cross-regional trips October 1–7 of 2.144B, +0.7% y/y, with rail leading; key shopping-mall sales flat vs May 1 holiday but better than Dragon Boat and Mid-Autumn; box office extended three consecutive years of decline ^[1].
• **[ONGOING] Strategy frameworks:** Shanxi Securities' four-part AI research series — position, diagnose, monitor, decide — frames the A-share tape as a K-shape divergence where policy-driven rebounds should not be chased while structural bottoms reward upper-arm exposure ^{[16][17][18][19]}.
• **[ESCALATED] Dividend strategy:** stable-yield dividend底仓 (banks, utilities) favored for October; cyclical dividends face earnings-cycle headwinds; CSI dividend yield minus 10-year CGB yield sits at the lower end of its normal range ^[20].

6. What Would Falsify It

• **Falsifiable tests:** tomorrow's September CPI (consensus: in line per Natixis ^[2]); any fresh Fed speaker pushback on the 'one-and-done' framing; a reset in the 64% Democratic-sweep probability; a turn in the dormant bond-selloff marker ^[4]. Truist's data argues bull markets past year four rarely end abruptly — so the relief rally has historical tailwind even as Hartnett's cash signal screams caution ^[5][21]. Contradiction is the trade: dovish rates, defensive flows.