The artificial intelligence infrastructure boom will require more equity financing as lenders become increasingly wary of concentrated borrowing across the sector, Blackstone President and Chief Operating Officer Jon Gray said at the Global Financial Leaders’ Investment Summit in Hong Kong.
The growing demand for computing capacity, data centres. and energy infrastructure is creating financing needs that debt markets alone may struggle to accommodate, Gray said, pointing to concentration risks among companies borrowing heavily to fund the AI expansion.
The distinction between equity and debt investors is becoming increasingly important as the industry attracts more capital, he said.
Equity investors can benefit from the upside when a handful of technology companies dominate a market. Debt investors, however, are primarily concerned with receiving interest payments and recovering their principal, leaving them more exposed to the risks of concentrated borrowing.
“If you’re a debt investor and you see this huge concentration of borrowing, you don’t have upside,” Gray said. “You’re playing for repayment and coupons.”
He said the financing requirements of the AI industry would likely lead to more equity issuance, citing fundraising by Meta Platforms, Google. and Oracle, as well as the prospect of public listings by SpaceX and Anthropic.
“There’s going to be more equity needed across the space to keep up with these enormous demands,” Gray said.
The funding requirements extend beyond data centres and computing infrastructure. Energy supply and the expansion of defence-related capabilities are also driving demand for capital, he said.
Gray nevertheless expressed confidence in the underlying investment case for AI infrastructure, pointing to the widening gap between demand for computing power and the difficulty of bringing new capacity online.
Companies in Blackstone’s portfolio have seen spending on AI models increase 20-fold over the past year, reflecting the financial benefits businesses are getting from applications such as content creation, customer engagement and coding, he said.
On the supply side, securing suitable sites, obtaining permits, sourcing chips and accessing electricity have become increasingly difficult. Bringing together the capital needed to finance these projects is another challenge.
Gray contrasted the current investment cycle with the telecommunications infrastructure boom of the 1990s, when companies built capacity in anticipation of demand that had yet to materialise.
“Now you’re in an environment where the demand’s running well ahead of supply,” he said.
He also pointed to the financial strength of the companies backing much of the current infrastructure development. Blackstone’s AI infrastructure investments are largely tied to major cloud computing companies, or hyperscalers, through long-term contracts, some extending 15 to 20 years.
These agreements, alongside strong demand for computing capacity, provide a degree of visibility for investors financing the buildout, Gray said.
Still, he acknowledged that the investment boom would not be without setbacks, with some companies potentially lacking the financial resources or operational capabilities to deliver on their commitments.
“There will be some things that go awry,” he said, adding that investors would increasingly scrutinise whether companies were financing their expansion on sound terms.
