AI Stocks & Markets

Today's Stories

1. 1

   Citi Declares 'Magnificent Seven' Label Obsolete for AI Stock Winners

   Citigroup analysts have concluded that the "Magnificent Seven" label—a term traditionally applied to seven mega-cap tech stocks—is no longer useful for tracking the companies winning in artificial intelligence, suggesting the investment framework needs to be updated. The original Magnificent Seven grouping has been the dominant lens through which Wall Street and investors have viewed mega-cap tech leadership and AI beneficiaries. If Citi's view gains traction, it may shift how analysts categorize and recommend stocks in the AI era, potentially affecting portfolio construction and valuation frameworks for large institutional investors.

   The readiness of other major investment banks and analysts to adopt or reject Citi's framing will shape whether the Magnificent Seven label persists in market discourse or whether a new taxonomy emerges for AI-driven stock leadership.

2. 2

   P&G Stock Gains as AI Selloff Deepens, Attracts Defensive Investors

   As AI stocks and semiconductors tumbled last week amid valuation concerns, Procter & Gamble advanced 4.2% in 2026 and gained 0.65% over the past five trading sessions—sharply outpacing the S&P 500 (down 0.40%) and Nasdaq Composite (down 0.66%). The 188-year-old consumer goods company, with a market value of $349.2 billion（約56兆円）, has emerged as a pocket of stability for investors seeking predictable earnings. Investors are becoming more selective as doubts grow about whether AI demand can sustain the productivity gains many expected. P&G's 136-year track record of uninterrupted dividends, 70-year streak of annual dividend increases, and blue-chip presence in the Dow Jones Industrial Average since 1932 (longer than any current component) make it an appealing alternative for portfolios seeking defensive positioning and lower volatility during AI market uncertainty.

   P&G is scheduled to report fiscal 2026 fourth-quarter earnings before market open on Wednesday, July 29. The company's fiscal 2026 guidance projects 1% to 5% reported sales growth and flat to 4% organic sales growth, with core EPS expected to reach $6.83 to $7.09 per share. Analysts hold a consensus "Moderate Buy" rating with an average price target of $163.17, implying 9.2% upside potential.

3. 3

   Broadcom's AI chip boom masks margin squeeze despite record cash flow

   Broadcom's AI semiconductor revenue surged 143% year-over-year in Q2 FY2026, reaching 49% of total revenue, while its traditionally high-margin software business grew only 9% and now represents 32% of revenue. The shift in product mix caused consolidated gross margins to drop to 77.1%, down 230 basis points from a year ago, with further decline to 74% expected in Q3. Broadcom is generating $10.3 billion（約1.6兆円） in free cash flow per quarter and $30 billion（約4.8兆円） in quarterly AI bookings, making it a cash-generating business rather than a speculative bet. However, the rapid growth of lower-margin AI chips relative to the company's historically high-margin software (which operates at 93% gross margin) is mechanically compressing overall profitability, even as the company's software business itself continues to accelerate—infrastructure software revenue is expected to reach $8.9 billion（約1.4兆円） in Q3, up 31% year-over-year.

   Management emphasized that gross margin compression is a mix-shift effect rather than structural erosion, and operating margins have remained steady at a record 67.3%. CEO Tan Hock stated the company does not expect long-term impact on its software business, though the diverging growth rates between the two segments suggest the composition of Broadcom's earnings will continue to shift materially.

4. 4

   AMD launches Helios AI rack servers to challenge Nvidia

   AMD introduced Helios, a new AI rack server platform designed to compete directly with Nvidia's offerings in the data center market. Nvidia dominates AI infrastructure sales, and AMD's entry into purpose-built AI rack servers signals an intensifying push to capture market share in a category where customers are investing heavily.

   The Helios platform's adoption rate among cloud providers and enterprises will indicate whether AMD can meaningfully penetrate Nvidia's stronghold in AI compute infrastructure.

5. 5

   AI capex projected to exceed $1 trillion（約160兆円） by 2027

   Raymond James Investment Management chief market strategist Matt Orton predicts that artificial intelligence capital expenditure will top $1 trillion（約160兆円） by 2027, citing strong cloud monetization and capacity constraints as drivers. The forecast signals that Big Tech companies view AI infrastructure investment as essential to sustain revenue growth. If those companies fail to raise spending guidance, chip stocks could face a 15% to 20% sell-off, making AI capex guidance a critical signal for semiconductor valuations.

   Upcoming Big Tech earnings calls and capital allocation announcements, particularly whether companies maintain or increase AI infrastructure spending commitments.

6. 6

   Google's Frozen v2 chip boosts Alphabet 3%, revives semiconductor bull case

   Alphabet stock climbed about 3% on Monday after The Information reported that Google is building a new AI chip called Frozen v2 designed to run its Gemini models up to 10 times more efficiently by 2028. The chip bakes parts of the Gemini model directly into silicon, reducing data movement and power consumption per answer. Semiconductor stocks just closed their worst week in more than a year—the VanEck Semiconductor ETF (SMH) fell 8.9% and the Philadelphia SE Semiconductor Index (SOX) dropped nearly 10%—driven by fears of an AI slowdown. Google's announcement suggests major tech companies are doubling down on AI infrastructure investment, which several Wall Street analysts see as a buying opportunity. Google also faces a compute shortage that reportedly forced Google Cloud to turn away business and led Google to agree to pay SpaceX nearly $1 billion（約1600億円） a month to fill the gap.

   Alphabet's earnings report on Wednesday will signal whether Google is backing its chip ambitions with increased spending. The Frozen v2 chip launches in 2028 and will support TPUs Google has built since 2016 rather than replacing them. JPMorgan analyst Mislav Matejka expects new chip supply will not arrive before 2028, while Evercore's Mark Lipacis predicts another 10% to 15% of downside, though he notes the SOX typically bounces 36% within 20 weeks of corrections like this one.