〔Day Digest〕Target Revisions Split: Microsoft $630-$650, Alphabet $345-$460, and a 10-2 Fed Hold Tests the AI Rebound

0. Daily Arc and Source Check ^{[1][2][3][4][5][6][7][8][9]}

• **[NEW] Split research tape:** Two Microsoft target increases and four Alphabet target reductions dominate the corporate research flow, while a 10-2 Fed hold and an options-based AI hedge create a two-sided setup ^{[1][2][3][4][6][7][8][9]}
• **[ONGOING] Source quality:** The Susquehanna and DA Davidson revisions each appear in duplicate newswire versions, while the Goldman, Nomura, Gundlach, TrendForce and CITIC arguments are relayed article summaries and should be treated as attributed rather than independently verified claims ^{[3][10][4][11][8][12][13][14][15]}
• **[NEW] Thin item:** The company-research teaser names no company and truncates both story lines, so it cannot support a stock-specific conclusion ^[5]
• **[ONGOING] Immediate test:** Alphabet and Microsoft had not yet reported when Nomura published its stabilization call, leaving the earnings response as the near-term check on the rebound ^[13]

1. AI Targets and the Options-Defined Rebound ^{[1][2][3][4][6][7][8][16][13]}

• **[NEW] Microsoft targets:** Stonex raised its target from $600 to $630, while DA Davidson lifted its target from $550 to $650 ^[1][4]
• **[NEW] Alphabet targets:** MoffettNathanson cut its target by $25 to $345, Susquehanna cut from $460 to $400, TD Cowen cut from $485 to $460, and Morgan Stanley cut from $400 to $375 ^[2][3][6][7]
• **[ONGOING] Options over spot:** Goldman Sachs recommended capturing AI upside through option structures rather than direct spot holdings, pairing limited-loss structures with downside puts as insurance ahead of technology earnings and the Fed decision ^[8]
• **[ONGOING] Event risk:** The same report said earnings support and light positioning could lift stocks by year-end, but AI concerns, Middle East conditions and U.S. midterms were expected to sustain near-term volatility ^[8]
• **[NEW] Volatility contradiction:** Goldman described Mag 7 implied volatility as exceptionally low and Neocloud pricing as signaling that a crisis had only just begun, while a broader market note put the Mag 7 relative volatility ratio at the 15th percentile of the past four years despite hedge-fund exposure near a historical peak ^[8][16]
• **[ONGOING] Stress gauge:** The one-month at-the-money implied-volatility spread between Nasdaq 100 and the S&P 500 remained at the 94th percentile even after its sharpest three-day compression in four years ^[16]
• **[ONGOING] Stabilization call:** Nomura cross-asset strategist Charlie McElligott said U.S. stocks had reached a critical stabilization after two weeks of left-tail and growth-scare repricing, citing Anthropic's S-1 and SK Hynix's earnings beat; the same tape included an approximately $1.6 billion Mag 7 call-option block, Nasdaq implied-volatility compression, and the S&P 500 1% below its high ^[13]
• **[ESCALATED] Fragility:** McElligott counted roughly 100 hedge-fund managers leaving during the turmoil and warned that another stop-out wave may only be a matter of time ^[13]

2. Fed Hold, Hawkish Minority and Inflation Risk ^[17][14][9]

• **[NEW] Fed decision:** The Federal Reserve held the target range at 3.75%-4.00% as expected, with a 10-2 vote versus 9-3 in July; Hammack and Cook sought a 25bp hike ^[9]
• **[NEW] Statement:** The Fed said it would gather more information before determining the appropriate policy stance, while reiterating that inflation remained elevated and reaffirming its commitment to price stability ^[9]
• **[ONGOING] 2026 path:** Kaiyuan Securities assigned a base case of one additional hike in December 2026, citing reduced pressure after September's hike, election politics and the risk of unmoored inflation expectations; it said the 2027 path depends heavily on U.S.-Iran developments and oil pass-through ^[17]
• **[ESCALATED] Inflation warning:** DoubleLine Capital founder and chief investment officer Jeffrey Gundlach said the Fed was aiming but not pulling the trigger, and his model expected CPI to begin with a 4% handle in the coming months because of energy prices, persisting for at least five months through the following April report published in May ^[14]
• **[ONGOING] Credit and allocation:** Gundlach said AI-related bond issuance was being received coldly and that investment-grade ratings diverged from market pricing, likening the setup to the ABX pattern before the 2006-2007 subprime crisis; he maintained 30% equities, 30% bonds, 20% commodities and 20% cash-like assets ^[14]
• **[ONGOING] Policy tension:** The Fed statement retained its inflation concern, Kaiyuan Securities retained a December hike as its base case, and Gundlach expected a 4% CPI handle, leaving energy and inflation transmission as the next decision variables ^[17][14][9]

3. Tight AI Hardware Against a Weaker Long-Duration Narrative ^[12][15]

• **[NEW] Inference pull:** TrendForce said the shift from model training toward inference and AI agents was lifting CPU demand; each CPU requires a power module and DDR5 memory, doubling high-capacity MLCC usage around CPU boards and memory modules ^[12]
• **[NEW] Supply path:** TrendForce expects Nvidia's Vera Rubin to ramp monthly through Q4 2026 and AMD's MI450 to begin pulling high-capacity specialty products, while ODM forecasts severe shortages in Q1-Q2 2027 and a supply gap lasting through the first half of 2027 ^[12]
• **[ONGOING] Pricing power:** Sellers retain control of pricing, Q1 2027 negotiations are due to begin in November 2026, and Japanese suppliers are likely to follow Samsung Electro-Mechanics' price increase ^[12]
• **[NEW] Narrative split:** CITIC Securities said SOXX had fallen 16% since the start of October, the largest monthly pullback in this AI cycle, even as supply remained tight across GPUs, CoWoS, HBM and ABF substrates ^[15]
• **[NEW] Bull case:** CITIC Securities identified Neocloud CDS-yield jumps and slowing annual recurring revenue at leading model companies as the main sources of pessimism, and expects easing inference compute plus new-model releases to broaden AI applications and reaccelerate revenue; it recommends focusing on low-valued compute, memory, optical-communication and CPU names ^[15]
• **[ONGOING] Falsifier:** The bull case depends on new models and easier inference compute restoring annual recurring revenue growth, since that is the stated transmission mechanism from AI deployment to model-company revenue ^[15]

4. Cross-Market Tape and Chinese Fundamentals ^[18][19][20]

• **[NEW] China and Hong Kong:** The China Merchants International daily wrap showed the Shanghai Composite up 0.9%, the Shenzhen Component up 2.0% and the ChiNext up 3.4%, with turnover rising to about RMB1.63 trillion; the Hang Seng gained 0.2% and Hang Seng Tech 1.4%, while southbound funds sold HK$804 million for an 11th consecutive session ^[20]
• **[ONGOING] Currency:** The dollar closed at 6.666 against the onshore renminbi in the night session and 6.662 against the offshore renminbi ^[20]
• **[NEW] Northeast Asia:** The Nikkei 225 rose 1.1% and the KOSPI 6.6%; SK Hynix reported record earnings and announced about 100 trillion won of shareholder returns, lifting its shares 10.9% and Samsung Electronics 9.0% ^[20]
• **[ONGOING] Europe and United States:** The Stoxx 600 edged up 0.1%, with Stellantis results supporting autos and copper weakness pressuring miners; the Dow rose 0.2% and the S&P 500 and Nasdaq each 0.3%, with technology and energy leading while utilities and real estate lagged ^[20]
• **[NEW] Industrial profits:** Kaiyuan Securities' review showed January-September industrial profits up 15.5% year on year versus 15.7% previously, revenue up 6.6%, and the revenue profit margin at 5.68% versus 5.66%; September monthly profit growth accelerated to 14.4% from 4.2% ^[19]
• **[NEW] Cost signal:** Cost per RMB100 of industrial revenue fell to RMB85.02, down RMB0.54 year on year and RMB0.05 month on month, while PPI rebounded to its year-to-date high ^[19]
• **[ONGOING] Value view:** Multiple institutions said the A-share adjustment reflected external disruption and sentiment release rather than a fundamental reversal, while Anxin Fund Value Investment Department deputy general manager Zhang Ming argued that low-valuation blue chips had entered an allocation window ^[18]

5. Quantum Research Radar ^[21][22][23]

• **[NEW] Gate benchmark:** Scientists at the Jülich Center and other institutions performed a superconducting-qubit quantum operation in 6.8 nanoseconds at more than 99.98% accuracy, versus the tens of nanoseconds typically required, setting a new quantum-gate benchmark in Nature Communications ^[22]
• **[NEW] Detector progress:** Kaiyuan Securities' quantum-technology report said superconducting nanowire single-photon detectors had reached 99% detection efficiency, 32-photon number resolution and thousand-pixel array formats, with products moving toward rack-based deployment ^[23]
• **[NEW] Limited detail:** Guangxi released a blue book on its marine economy, but the packet provides only a headline and no supporting figures ^[21]