- BlackRock says AI infrastructure spending is far from peaking and will keep driving capital flows.
- Ben Powell highlights hardware, power and materials firms as the clearest winners, not model developers.
- Hyperscalers are racing to secure chips and electricity, pushing long-term procurement and debt issuance.
BlackRock: capex deluge shows no signs of slowing
Ben Powell, BlackRock’s chief investment strategist for APAC, told CNBC at Abu Dhabi Finance Week that the AI-related capital expenditure wave is still accelerating. He described a “picks and shovels capex super boom,” arguing that firms supplying the AI build-out — chipmakers, energy providers and materials companies — are best positioned to capture durable gains.
“The capex deluge continues. The money is very, very clear,” Powell said, noting that major tech firms are spending aggressively to avoid being left behind.
Why suppliers beat model developers for now
Powell said much of the incoming capital will flow to the physical infrastructure powering AI: GPUs, servers, power plants, transmission, and even copper wiring. Those companies provide the components hyperscalers need to expand data-center capacity, and their revenue streams are less tied to the uncertain economics of model monetization.
Investors have already seen how hardware winners can outperform. Nvidia’s GPUs are central to modern AI systems and helped the company reach record valuations this year. But Powell warned that the race to secure computing power and electricity is fueling long-term procurement commitments, not a short-lived spike.
Hyperscalers tapping credit markets
Powell also said that leading tech firms have only just started to tap debt markets to fund the next phase of AI expansion. “The big companies have only just started dipping their toes into the credit markets… feels like there’s a lot more they can do there,” he told CNBC.
That suggests additional capital could arrive via corporate bonds or other financing, supporting continued investment in data centers and chip supply agreements. Companies including Amazon and Meta have already budgeted tens of billions annually for AI-related investments, and some industry estimates point to a near doubling of data-center power demand by 2030.
Global infrastructure ripple effects
The AI build-out is reshaping energy and materials markets. Grid operators from the U.S. to the Middle East are planning upgrades to meet growing electricity demand. S&P Global expects data-center power needs to rise substantially through the decade, creating tailwinds for utilities, renewable projects, and commodity suppliers.
What investors should watch next
Powell expects positive surprises to favor suppliers: chipmakers, power producers, electrical equipment makers, and companies involved in data-center construction. For investors seeking exposure to AI’s long-term industrial impact, the “picks and shovels” trade offers a clearer risk-reward profile than backing model developers alone.
As hyperscalers accelerate spending and explore credit markets, the suppliers that enable AI at scale may remain the most reliable beneficiaries of the ongoing boom.
Image Referance: https://www.cnbc.com/2025/12/08/blackrock-ben-powell-bet-trade-as-ai-spending-frenzy-hyperscalers-investment-credit.html
