Citadel Securities LLC projects that over $500 billion in debt will be issued in the public and private markets by 2028 to finance chips for artificial intelligence (AI) data centers. This amount could represent more than 5% of the Bloomberg US high-grade index by that year, according to Jeff Eason, head of the investment-grade desk at Citadel. Eason noted that most of the new debt will likely have shorter maturities of three to five years to align with the lifespan of the chips. He described this potential issuance as unprecedented in scale compared to the current market.
Key Details
The surge in AI-related debt issuance has already reached approximately $570 billion globally, primarily driven by major tech companies known as hyperscalers, including Amazon.com Inc., Microsoft Corp., and Alphabet Inc.'s Google. These firms are constructing data centers at an unprecedented pace, with Citadel estimating that chip-makers alone could issue more than $250 billion by 2028. Eason remarked,
This has the potential to become one of the largest new sectors in investment-grade credit.
Background
Long-term interest rates have surged to levels not seen in two decades, with hyperscaler debt issuance identified as a significant factor. Apollo Global’s chief economist, Torsten Slok, highlighted that this issuance is one of the main drivers behind the current high 30-year Treasury yields, alongside inflation and fiscal deficits. The term "AI Debt Tsunami" has been coined to describe the situation, with Morgan Stanley reporting that high-grade AI debt supply reached $270 billion by early July, more than double the total for all of 2025.
Related coverage: Nvidia Explores $250 Billion Financing Amid AI Concerns.
The anticipated increase in AI-related debt could pressure credit markets and contribute to rising long-term Treasury yields. Investors are particularly focused on the financial health of major tech companies as they navigate the challenges of servicing this new debt load. Watch for upcoming earnings reports from key players in the AI sector to gauge their ability to manage this financial burden.