Uncategorized

Concerns Over Hidden Debt Arise While Global Investments Drive US AI Advancements

According to Charles Wu of JPMorgan Asset Management, foreign investors are increasingly supporting the growth of artificial intelligence computing throughout North America, as he shared with the audience at the SuperReturn conference held in Singapore on Wednesday.

Institutional investors and foreign banks—including pensions, sovereign wealth funds, and insurance companies from regions like Asia and the Gulf—are playing a significant role in financing this sector, which has nearly $1 trillion in capital expenditures, Wu noted, who is also the head of Asia Pacific alternatives institutional client strategy at the bank.

“Compute training assets remain predominantly in North America,” he remarked during a panel discussion. “Both in banking and asset management, the capital facilitating these assets is increasingly becoming global.”

The strategies of global investors regarding the AI sector, particularly the substantial funding requirements for the world’s data centers, have been a prominent topic at the SuperReturn conference this week. While most panelists conveyed an optimistic viewpoint, some expressed caution.

Read: Big tech’s costly AI race is fuelling hundreds of billions of dollars in debt deals

“We are very cautious in how we evaluate risk,” stated Jean-Christophe Aubert, a senior director of infrastructure investments at PSP Investments.

ADVERTISEMENT

CONTINUE READING BELOW

“Concentration is a significant issue, but for us, given the capital we need to deploy, it’s about understanding how we can ultimately realize those assets in the future.”

While US hyperscalers are addressing much of the AI capital expenditure through their operating cash flows, they are also securing approximately $250 billion from the bond markets and roughly the same from bank loans, according to Wu.

He cautioned that his firm was monitoring several concerns, including lease obligations and concentration risk, as well as the possibility that some financing is being sourced off-balance sheet.

“The truth is many funding securities—whether investment-grade bonds, 144a offerings, private placements, or even private credit structured deals—seem to pivot around five key hyperscaler entities,” he explained. “These can act as guarantors, tenants, funding sources, or customers.”

Asia emerges as a key AI market

Although the US continues to be the leading market, several panelists concurred that Asia is likely to follow closely. The nature of inference—practical use of AI models—benefits from proximity to end users, thus boosting the demand for infrastructure in the region.

“Asia holds significant importance as it plays an intriguing role in both the demand and supply sides of the market,” stated Mohsin Pirzada, head of funds at Qatar Investment Authority.

ADVERTISEMENT:

CONTINUE READING BELOW

“We are witnessing a decentralization driven by consumer demand, alongside factors like data sovereignty and security,” he mentioned.

Read:
AI frenzy draws stock traders to Southeast Asia’s likely winners
Chatbots, AI news readers: Asia embraces automation despite job loss fears

The panelists also agreed that energy remains a significant bottleneck for various reasons. With gas as a common energy source, a shortage of gas turbines poses challenges for data centers striving to expand. I Squared fund partner Chenhua Shen indicated that Asia’s power markets are less liberalized, making utility access more difficult.

“This presents a substantial hurdle for Asia’s leapfrogging,” she commented. “Nonetheless, we remain hopeful, as Asia benefits from an abundance of renewable energy sources.”

© 2026 Bloomberg

Leave a Reply

Your email address will not be published. Required fields are marked *