〔Overnight Brief〕September Hike Locked at 85-90% on 3.4% August CPI, 10Y Nears 5% on +57bp YTD Move; Warsh Caught Between Preventive Tightening and Trump's No-Win

0. Weekly Arc

Markets have all but locked in a 25bp September hike: probability has climbed to 85-90% on CME futures after the August CPI/PPI/NFP cluster ^[1][2][3][4]. The 10-year is closing on 5%, and the curve has bear-steepened all year ^[5][2][6]. Yet the political axis runs the other way — Trump on Sunday renewed his push for cuts, and Hassett said both he and the President see "no reason" to hike ^[7][4][8]. Warsh steps into a "no-win" setup: hike and defy the principal, or hold and let expectations drift ^[9][4]. A Shanghai Securities News-cited brokerage outlook argues the base case is preventive tightening, with A-share risk appetite expected to repair after the meeting lands ^[10][11]. Net: a hike is priced, but the framing of it — one-and-done vs. first of a series — is the actual fight ^[10][11][2].

1. Policy Narrative

• **[ESCALATED] Preventive hike framing (single-source / brokerage relay):** a Shanghai Securities News roundup calls the 25bp base case "preventive tightening" and says the post-meeting path is risk-on for A-shares once overseas uncertainty clears ^[10][11].
• **[NEW] Hassett, on the inflation data:** "cautious" about hiking; said both he and Trump see no reason to hike and that holding steady into the election is important; pledged Trump will defend Warsh's independence "if Warsh raises rates" ^[12][7][8]. Three separate items, same axis.
• **[NEW] "No-win" framing (single-source relay):** a Jin10 preview argues Warsh is trapped — hike and defy Trump, or hold and let expectations un-anchor — and calls this his "real credibility test" ^[9][4].
• **[NEW] Serial-hike history (brokerage):** the 2022-2024 cycle delivered 11 consecutive hikes; the base case has rarely been a one-and-done, and a separate Moomoo piece raises the late-1980s parallel ^[2][13]. Contradiction with the "preventive" framing above is unresolved in the packet.
• **[NEW] Background drumbeat:** Trump publicly pushing for cuts in IBD and AP coverage; Barron's and TradingKey both flag the FOMC and dot plot as the week's main event ^{[14][15][16][17]}.

2. Key Data

• **[NEW] August CPI:** headline +0.4% m/m and 3.4% YoY; core +0.3% m/m, the largest single-month gain since April ^[2][3]. The proximate trigger.
• **[NEW] August PPI:** +5.4% YoY; final-demand goods +1.1% m/m; processed energy +7.3% m/m ^[3]. Energy is doing the work.
• **[NEW] August NFP:** +162k; unemployment 4.1% ^[1][3]. Labor leg holds.
• **[NEW] Market read:** CME futures priced September at ~90% intraday after CPI; a Jin10 relay puts the current read "above 85%" ^[1][4]. Sina brokerage puts the 25bp probability at 90% ^[2]. Treat 85-90% as the band, not a point.
• **[NEW] One brokerage's read of the macro state:** a Gelonghui piece calls the US a "quantum superposition — can hike or not" ^[1]. Decorative, but captures the policy uncertainty.

3. Bond Market and Flows

• **[ESCALATED] 10Y near 5%:** a Sina brokerage note says 10-year yields are "approaching the 5% threshold" ^[2]; Bloomberg flags a 5% Treasury as a "new risk" for markets and the economy ^[6].
• **[NEW] YTD curve shift (as of Aug 31):** 2Y +87bp, 10Y +57bp, 30Y +41bp ^[5]. Bear-steepening — front end re-prices the hike, long end re-prices inflation persistence.
• **[NEW] Foreign holdings — against the narrative:** TIC data shows foreign investors were net buyers of $154.4B in H1 2026; total foreign holdings $9.30T as of June 30 vs. $9.49T at end-Q1, with the decline fully explained by a $124.9B negative valuation effect ^[5]. The "foreign dump" story is not supported by this data.
• **[NEW] Equity outflows:** US equity ETFs saw $4.5B in outflows as the hike was priced in ^[18].
• **[NEW] Bond-market push (MarketWatch):** argues the bond market is pushing for hikes even though they "won't bring down gas prices" — the channel is expectations, not supply ^[19].

4. Geopolitical and Energy Backdrop

• **[ESCALATED] Dual-strait risk:** Iran tightened controls around the Strait of Hormuz on Sept 6; a Sept 8 US missile strike on an Iranian tanker near Kharg Island prompted retaliation threats against Kuwaiti and Bahraini port traffic; Houthi control of Bab el-Mandeb continues to thin transits to single digits in some windows ^[3]. Both Hormuz and Bab el-Mandeb under simultaneous stress means crude, LNG and Asia-Europe shipping costs rise together ^[3].
• **[ESCALATED] Inflation persistence channel:** the Sina brokerage note flags energy as the main driver of the August CPI surprise, with September oil strength expected to keep headline sticky ^[2]. A Center Square op-ed makes the parallel point: the Fed cannot pump more oil ^[20].
• **[NEW] Fed-PPI linkage:** August PPI processed-energy +7.3% m/m is the upstream signal for the next CPI energy line ^[3].

5. What Decides Next

• The FOMC decision and dot plot this week are the obvious pivot; the open question is whether the move is a one-and-done or opens a serial cycle ^[2][16][13]. The packet disagrees with itself: a brokerage says 11-of-11 historical precedent, Moomoo raises the late-1980s parallel, the Shanghai outlook calls the move preventive and A-share-positive ^{[10][11][2][13]}.
• Falsifiable test: dot plot and Warsh press-conference language on "further moves" vs. "data-dependent." Geopolitical falsifier: any de-escalation at Hormuz or Bab el-Mandeb would compress the energy tail and loosen the hike case ^[3].
• Source quality control: most policy framing comes through single-source social relays (Jin10, Gelonghui, Cailian, brokerage roundups) ^{[10][11][1][7][9][4][8]}; the TIC foreign-holding figure and the official August CPI/PPI/NFP prints are the most solid anchors in the packet ^[5][3].