Ben Powell, chief strategist for Middle East and Asia Pacific at BlackRock Investment Institute, during a Bloomberg Television interview at the Abu Dhabi Finance Week (ADFW) conference in Abu Dhabi, AD, United Arab Emirates, on Monday, Dec. 9, 2024.
Bloomberg | Getty Images
The wave of capital pouring into artificial intelligence infrastructure is far from peaking, said Ben Powell, chief investment strategist for APAC at BlackRock, arguing the sector’s “picks and shovels” suppliers — from chipmakers to energy producers and copper-wire manufacturers — remain the clearest winners as hyperscalers race to outspend one another.
The surge in AI-related capital expenditure shows no sign of slowing as tech giants push aggressively to secure an edge in what they see as a winner-takes-all contest, Powell told CNBC Monday on the sidelines of the Abu Dhabi Finance Week.
“The capex deluge continues. The money is very, very clear,” he said, adding that BlackRock is focused on what he called a “traditional picks and shovels capex super boom, which still feels like it’s got more to go.”
AI infrastructure has been one of the biggest drivers of global investment this year, fueling a broader market rally, even as some investors question how long the boom can last.
Nvidia, whose GPU chips are the backbone of the AI revolution, became the first company to briefly surpass $5 trillion in market capitalization amid a dizzying AI-fueled market rally that sparked talk of an AI bubble.
Microsoft and OpenAI also reached a restructuring deal in October to support the ChatGPT developer’s fundraising efforts. OpenAI has reportedly been preparing for an initial public offering that could value the company at $1 trillion, according to Reuters.
The build-out has set off long-term procurement efforts across the tech sector, from chip supply agreements to power commitments. Grid operators from the U.S. to the Middle East are racing to meet soaring electricity demand from new data centers. Companies, including Amazon and Meta, have budgeted tens of billions of dollars annually for AI-related investments.
S&P Global estimates data-center power demand could nearly double by 2030, mostly driven by hyperscale, enterprise and leased facilities, along with crypto-mining sites.
‘Dipping toes into credit market’
Powell also noted that leading tech firms have only begun to tap capital markets to fund the next phase of AI expansion, suggesting additional capital is on the way.
“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 said.
The “hyperscalers” are behaving as if coming second would effectively leave them out of the market, Powell said. That mindset, he added, has pushed firms to accelerate spending even at the risk of overshooting.
Much of that capital, Powell noted, is likely to flow to the companies powering the AI build-out rather than model developers, reinforcing a growing view among global investors that the most durable gains from the AI boom may lie in the hardware, energy and infrastructure ecosystems behind the technology.
“If we’re the recipients of that cash flow, I guess that’s a pretty good place to be, whether you’re making chips, whether you’re making energy all the way down to the copper wiring,” Powell noted, expecting “positive surprises driving those stocks in the year ahead.”
