Great Rotation Happening Now: This Sector ‘Is Over’ | Thomas Hayes

Summary

Thomas Hayes argues that a major rotation is underway: money is moving away from crowded AI infrastructure and semiconductor trades toward bonds, defensive sectors, small caps, value stocks, and international equities. He remains bullish on AI’s long-term productivity effects, but says investors should favor companies that use AI to improve margins rather than companies financing or supplying the buildout.

Main market and investment views

• AI and semiconductors: Hayes says the near-term “commodity trade” within AI—especially memory and commodity chips—is over. He argues that high prices will attract capacity from Chinese, Korean, and Japanese competitors, leading to oversupply. He says semiconductor stocks have historically suffered sharp downturns after capacity expands, even while the underlying technology theme continues.
• AI’s likely beneficiaries: He favors businesses applying AI to increase productivity and margins over chipmakers, hyperscalers, and other capex-heavy suppliers. He expects companies to adopt lower-cost open-weight AI models for many tasks.
• Market rotation: Hayes expects the next 6–12 months to bring broader gains beneath the headline indexes, with the S&P 500 potentially rising more modestly as a handful of former leaders consolidate or fall. He sees potential outperformance in small caps, value, international equities, and defensive sectors.
• Defensives and bonds: Utilities, consumer staples, and healthcare had reportedly begun attracting material buying over the prior five sessions, alongside bonds. Hayes says the comparison between a roughly 5.2% 10-year Treasury yield and an approximately 5% equity earnings yield is increasingly compelling for pension funds and other investors with fixed return targets.
• Rates and catalysts: He links rising global yields to a combination of economic growth, capital demand, inflation expectations, and elevated oil prices related to the Iran war. He says an end to the war could push the U.S. 10-year yield toward 4.75%. If nominal growth slows, he sees 4.25%–4.50% as possible, versus roughly 4.75%–5% if growth remains stronger. He says yields in the 5.75%–6% area could intensify competition between bonds and equities.
• Small caps and consumer sectors: Hayes is bullish on small caps over roughly 9–16 months, citing their relative underperformance, operating leverage, and ability to adopt AI quickly. He also sees potential recovery in consumer-discretionary areas such as restaurants, clothing, and sneakers, arguing that weak sentiment may not reflect actual spending.
• Housing: Hayes says he would not invest in housing through the specific heavily sold-off housing-related stock discussed in the interview. Instead, he favors homebuilding and building-supply exposure through Brad Jacobs’ acquisition strategy. He also cites a U.S. housing shortfall of about 4 million homes, but notes that rate relief and smaller starter homes could be important to a recovery.

Stocks, sectors, assets, and instruments mentioned

• Semiconductors and AI: Micron (MU), Nvidia (NVDA), Intel (INTC), SanDisk, commodity memory and semiconductor chips, GPUs, CPUs, large language models, open-weight AI models, and AI infrastructure.
• Natural gas: Comstock Resources (CRK) and natural gas reserves. Hayes links its Louisiana location to Gulf export terminals and data-center development in East Texas.
• China and international AI: Alibaba (BABA) and Moonshot AI; Hayes also discusses international equities and Asian AI companies broadly.
• Telecom infrastructure and real estate: Crown Castle (CCI), cell towers, Starlink Mobile, spectrum, and REITs.
• Logistics and AI adopters: GXO Logistics (GXO), United Rentals (URI), XPO (XPO), and Waste Connections (WCN); Waste Management (WM) is also mentioned.
• Defensive companies and sectors: Diageo (DEO), Baxter International (BAX), Coca-Cola (KO), Procter & Gamble (PG), Disney (DIS), utilities, healthcare, consumer staples, and real estate.
• Housing and building products: Taylor Morrison (TMHC), Lennar (the transcript’s “LAR” reference is unclear), and Brad Jacobs’ building-products platform, likely QXO (QXO). The transcript also references a Freddie Mac-related stock, but the exact security is unclear.
• Markets and rates: S&P 500, Nasdaq, Russell 2000, KOSPI, U.S. Treasuries, the 10-year Treasury, investment-grade debt, private credit, and CCC-rated credit.
• Other macro references: Oil, diesel, money-market funds, and the ISM manufacturing index.

Hayes’ cited figures and company claims

• Hayes says that since July 31, before the latest four days of trading, 89% of S&P 500 stocks were down more than 10%, 54% were down more than 20%, 29% were down more than 30%, and 7% were down more than 50%, while the index was held up by a small number of AI leaders.
• He says the semiconductor trade fell roughly 35%–60% in July, depending on the stock, and says he sold Intel at $96–$97 after first buying it around $19.
• Hayes describes Comstock Resources as having a market value of about $3.7–$3.8 billion and estimates its proved and unproved reserves are worth about $20 billion. He gives a prospective share-price range of $40–$60 over 3–5 years, but the quoted current share price is garbled in the subtitles.
• He says Alibaba owns about one-third of Moonshot AI, which was reportedly seeking a $50 billion public valuation. He characterizes Alibaba’s implied stake as about $15 billion and says he believes that value is not reflected in Alibaba’s share price.
• He cites Crown Castle’s 5.3% dividend yield and says the stock rose about 20% over two or three days after investors reassessed the competitive threat from Starlink Mobile.
• For Brad Jacobs’ past deals, Hayes cites 200x returns for investors in United Rentals and 50x for XPO. For the current building-supplies effort, he cites roughly $4 billion of debt and EBITDA growth from $2.5 million to $7.5 million; the units in the subtitle may be mistranscribed.
• He says small caps had their second-worst quarter relative to the S&P 500 this century. He cites ISM manufacturing at 54.5, nine consecutive months of expansion, and $8 trillion in money-market funds.
• Other economic figures cited include 4.2% unemployment, nominal growth of about 8%, and gasoline at $7 per gallon in Newport Beach.
• Hayes says the S&P’s valuation had compressed from about 23x to 19x earnings; at 20x, the implied earnings yield is about 5%. He contrasts that with a roughly 5.2% Treasury yield and mentions company earnings expectations of about 30% for the current quarter and 15% for the following year. He cautions that such optimistic estimates recall the high expectations preceding the 2022 downturn.

Framework and portfolio approach

Hayes’ stated approach is to:

• Distinguish the AI theme from the AI trade: Remain constructive on AI’s long-term impact while avoiding crowded, capex-intensive commodity suppliers.
• Favor AI adopters: Look for businesses using AI, robotics, or automation to lift productivity and margins.
• Screen for out-of-favor investments: Focus on companies that have sold off or are underappreciated, rather than simply following popular market leaders.
• Assess debt maturity, not just debt levels: He says he checks when portfolio companies’ debt comes due; he cites holdings with debt that generally does not need refinancing until 2030.
• Consider relative yields and investor constraints: Compare a bond’s contracted yield with the uncertain earnings yield from equities, especially for pension funds with roughly 6% discount or hurdle rates.
• Look for potential catalysts: A pause or reversal in yield increases, lower oil prices, or a positive resolution to the Iran conflict could support interest-sensitive stocks and other beaten-down areas.

Explicit cautions and disclaimers

• Hayes says his views and price targets are “opinion, not advice.”
• He cautions that semiconductor and AI infrastructure stocks could fall sharply even if AI adoption and productivity gains continue.
• He notes that small-cap and other contrarian opportunities may take time to work, and that volatility over a 1–3 month period is difficult to predict.
• The video’s disclaimer says it is for informational and educational purposes only and is not financial, investment, legal, or tax advice.
• The host also promotes a prediction-market sponsor and cites an advertised 22% chance that the S&P 500 finishes the year between 8,000 and 8,200. This is presented as a promotional market quote, not Hayes’ forecast.

Presenters and sources

David Lin (host/interviewer); Thomas Hayes, Managing Member of Great Hill Capital and publisher of Hedge Fund Tips; Paul Tudor Jones, whose remarks are played in a clip.