〔Day Digest〕Fed Lifts to 3.75-4.00% on a 16-of-18 Hawkish Dot; Goldman Adds October, Deutsche Bank Flags the Shallowest Cycle in Modern History — Stocks-and-Yields-Rising Is Back on the Table

0. The Arc

The September FOMC pulled forward a hawkish repricing: rates lifted 25bp to 3.75-4.00%, the dot plot showed 16 of 18 members expecting at least one more 2026 hike, the neutral rate median jumped from 3.06% to 3.25%, and Chair Walsh used 'removing a dose of accommodation' three times while calling financial conditions 'far from restrictive' ^[1]. The 2-year Treasury yield jumped intraday, and the market priced October at 51% (up 8pp) and December at 78% ^[1]. Two strands now coexist: banks rapidly embracing the steeper path ^[1][2], and the same move being read as a near-one-off relative to past cycles ^[3] or a cycle shallow enough to be a modern record ^[4].

1. Policy Narrative — Banks Quickly Stack Revisions

• **[ESCALATED] Goldman (David Mericle):** added an October 25bp hike to the base case; scenario probabilities are 50% for two hikes to 3.25-3.50%, 35% for three hikes with a higher terminal, 15% recession; the firm offsets the path change with a new 25bp cut in March 2028 ^[1].
• **[ESCALATED] Industrial Securities Chief Economist Liu Yu:** 'may hike at least another 25bp'; terminal between 4-5%, with the >5% path constrained by debt dynamics unless productivity gains are confirmed ^[2].
• **[NEW] Deutsche Bank (historical lens):** if market pricing of three more hikes by mid-2027 is correct, this would be the shallowest cycle in modern record, below the 137bp of 1986-87; historical averages are 22 months and 478bp (medians 15 months and 313bp) ^[4].
• **[NEW] Morningstar (contrarian):** the Fed raised rates but still expects cuts next year ^[5].
• **[NEW] JPMorgan Global Macro Conference:** 8 of 9 DM central banks tracked expected to hike before year-end; Taylor-rule gap ~100bp, risk skewed 'faster bigger'; stocks and bond yields can rise together, S&P 500 year-end target 8,000, positioning described as underweight-not-extreme ^[6].
• **[NEW] UBS (Arend Kapteyn):** 'calm before the storm' — Sept-Oct are the year's most volatile months, Senate control is a coin flip, and since 1950 the ruling party has lost an average 25 House and 3 Senate seats in midterms ^[7].

2. Market Read — A Shares, Gold, Asia FX

• **[NEW] Gold:** closed at a four-week low on the Walsh hawkish combination; the watch-item is whether the long end can retreat from 5% ^[8].
• **[NEW] A-shares (CICC):** this Fed hike is not a normal cycle opener, A-share impact 'to be observed'; near-term reaction already largely priced, with effects limited if it is a one-off ^[3][9].
• **[NEW] Asia EM equities (Morgan Stanley, Jonathan Garner):** hawkish Fed + high oil raise the odds of a third 2026 pullback in coming weeks; team expects two more 25bp hikes (December and next March) to a 4.25-4.5% terminal in Q1 2027; underweight Australia, prefer Japan, Korea, Taiwan; USD trade-weighted back above 100 for the first time since late July ^[10].
• **[NEW] New Zealand Superannuation Fund (CEO Jo Townsend):** 14.2% annual return, top-ranked globally; warns US equity returns sit near 2x their 20-year annualized pace, sees mean reversion ahead ^[11].

3. Sell-Side Single-Stock and Sector Highlights

• **[NEW] Morgan Stanley — Vivace Biotech (HK):** upgrade to Overweight, target raised to HK$93 from HK$85, on positive six-month PFS ratio for LBL-024 (PD-L1/4-1BB) in 1L NSCLC at WCLC; next catalysts are ESMO October data and 2027 H1 NMPA approval ^[12].
• **[NEW] Goldman — GF Securities (A/H):** A-share target raised to 37.81 yuan (from 35.83), H-share to HK$23.55 (from HK$22.32); 2026-28 revenue and net profit forecasts each raised 5% on stronger investment income and overseas contribution ^[13].
• **[NEW] Goldman — Hong Kong property:** new policy was 'fairly limited' as expected; maintains 3-4 year upcycle call, prefers SHK Properties, Henderson Land, Sino Land ^[14].
• **[NEW] Morgan Stanley — China advanced packaging:** segment set to reach 100bn yuan by 2029; prefers equipment makers over OSATs, OSAT picks are share-gainers ^[15].
• **[NEW] CLSA — Zhipu:** year-end ARR guidance raised to 3bn (from 2.4bn) post 5bn strategic round; Outperform ^[16].
• **[NEW] Guosen — Haiguang Information:** H1 net profit 1.798bn yuan (+49.7% YoY), Q2 1.111bn yuan (+59.8% YoY, +61.7% QoQ); 2026-28 net profit forecasts 4.253/6.745/8.721bn yuan, current PE 129/81/63x ^[17].
• **[NEW] China Galaxy — fiber:** computing network build-out and AI demand jointly tighten supply, sustain high prosperity ^[18].
• **[NEW] Hua Long Securities — 2026 H1 review:** Q2 all-A cumulative net profit +15.98% (prior +6.89%), single-quarter +24.96%; Q2 style breakdown: growth +57.02%, cyclical +25.80%, financial +17.92% ^[19].
• **[NEW] CITIC CLSA — Philippines property:** capital-recycling gains booked straight to equity inflate reported book value, so reported ROE is mechanically depressed; adjusted 'real' ROE significantly higher — e.g. Robinsons Land 2025 reported ROE fell after each successful divestment ^[20].
• **[NEW] CITIC CLSA — Singapore property:** Royal Holdings/RB Capital acquire Scotts Square for S$310m (S$2,369/sqft); Centurion buys a JB worker dormitory for 53m, capacity up 22% to 43,980 beds; S-Reits -2% on the week, YTD institutional outflow S$1.3bn vs property-stock inflow S$235.3m ^[21].
• **[NEW] Industrial Securities strategy — AI crowding:** upstream hardware (optical modules, fiber, PCB, memory, semi materials/equipment) sits at multi-year lows, while some mid/downstream tracks are at multi-year highs ^[22].
• **[NEW] Soochow Securities — 6G:** part of the 'ten future industries' series; frames 6G around communication-sensing-computing-AI fusion and lists terahertz, RIS, and computing-network as key techs ^[23].
• **[NEW] Wanlian Securities — banks:** allocation case still intact on dividend yield and valuation; prefers stability names, with select regional and undervalued state-owned banks for expectation-gap trades ^[24].

4. Falsifiable Tests and Source Notes

• Three datapoints would settle the path debate: the next inflation print against the Walsh 'not restrictive' frame ^[1][8]; whether the long end holds at 5% or breaks lower ^[8]; the Q4 2026 / Q1 2027 macro mix against Morgan Stanley's 4.25-4.5% terminal ^[10].
• Contradiction is the story: Industrial Securities sees at least 25bp on top of Wednesday's move with a 4-5% terminal ^[2] while Deutsche Bank calls the whole cycle the shallowest in modern record if pricing is right ^[4]; Morningstar still looks for cuts next year ^[5] versus Goldman's October add-on ^[1].
• Source-quality flags: the JPMorgan macro confab is a single conference note ^[6]; the Liu Yu quote routes through a People's Finance channel rather than the underlying research ^[2]; the UBS midterm history is sourced to UBS research but presented in summary form ^[7]; Haiguang forecast revisions are an explicit downward reset from 26-27 net profit 5.494/6.385bn yuan to 4.253/6.745bn yuan, with 2028 added at 8.721bn yuan ^[17].