Welcome to the home for real-time coverage of markets brought to you by Reuters reporters. You can share your thoughts with us at markets.research@thomsonreuters.com

HYPERSCALER CAPEX SPENDING BRINGS BACK MEMORIES OF 1990s TELECOM INVESTMENT BOOM

Hyperscalers are set to pour billions into AI capex spending, as they race to capture the next-generation's revolutionary technology, with Goldman Sachs noting consensus 2026 capex estimates have risen from $467 billion at the start of the earnings season to $533 billion today.

The Wall Street brokerage points out that in Q3 2025 alone, hyperscalers spent $107 billion, up 76% year-over-year, and while analysts now expect 34% capex growth in 2026, maintaining the recent 75% pace could push spending to $700 billion, rivaling the late 1990s telecom investment boom.

Four of the biggest U.S. technology companies, Alphabet GOOGL.O, Microsoft MSFT.O, Facebook-owner Meta META.O and Amazon AMZN.O flagged plans to accelerate capital spending over the next year, in their recent earnings reports.

Traditionally, hyperscalers have relied on cash flows for funding, but are increasingly tapping debt markets, though their balance sheets remain strong and credit ratings high.

However, such massive borrowing could strain the debt market itself. Supply bottlenecks, shifting from chips to power constraints, also threaten to slow growth but may drive further investment and capex inflation, they added.

If hyperscalers choose to fund more AI spending with debt, the size of the debt market - rather than their own financial strength - would be the main constraint. Raising debt on this scale would represent a significant portion of recent investment-grade bond issuance, Goldman said.A recent example is Oracle ORCL.N, whose bonds have fallen after reports that the company plans to add $38 billion to its already large debt to fund its AI infrastructure, according to analysts and investors.

Investor confidence hinges on clear links between capex and revenue growth, with risks from negative market reactions, macro shocks, and rising leverage among smaller players.

As AI adoption expands, attention is shifting to "AI Platform" and "Phase 4" companies poised to benefit from productivity gains, even as the ecosystem navigates growing financial and operational complexity.

Goldman cites names such as KeyCorp KEY.N, Bank of America BAC.N, Robert Half RHI.N, Labcorp LH.N and Accenture ACN.N among the names as part of its AI productivity beneficiaries screen.

With businesses embracing AI and concerns about infrastructure costs rising, investors are turning to the next wave of beneficiaries. These include platform providers gaining direct revenue from AI use and companies with high labor costs that are exploring AI automation for efficiency.

(Kanishka Ajmera)

*****

EARLIER ON LIVE MARKETS:

BNP SEES UK AUTUMN BUDGET BALANCING FISCAL CREDIBILITY WITH POLITICAL RISKS CLICK HERE

SUPERPOSITION SHOWDOWN: WHO WILL CRACK THE QUANTUM CODE FIRST? CLICK HERE

ANIMAL SPIRITS SLUMP, US STOCKS STRUGGLE CLICK HERE

GENERATIVE AI COULD MAKE MARKETS MORE EFFICIENT - DB CLICK HERE

FINANCIALS DRAG; FOOD, TOBACCO AND DRUGS GIVE SUPPORT CLICK HERE

BEFORE THE BELL: TENTATIVE STABILISATION, AWAITING NVIDIA CLICK HERE

HUSH FALLS OVER MARKETS AHEAD OF NVIDIA CLICK HERE

(Terence Gabriel is a Reuters market analyst. The views expressed are his own)