The bond market surge is creating significant pressure on AI infrastructure financing, with higher yields generally drawing more investment into bonds issued by large tech firms investing in AI infrastructure. According to Macquarie Group's global rates strategist Thierry Wizman, "The private sector wants to have the AI revolution. Who's going to take a step back? It's going to be the consumer. And higher yields are going to do that a little bit." This trend is particularly evident in the AI debt market, where companies are increasingly turning to junk bond investors for funding even for investment-grade debt. As reported by Business Standard, QTS Realty Trust Inc. sold $3.9 billion of bonds this week to fund a facility in Georgia tied to Microsoft Corp., with the bonds carrying high-grade ratings but yielding about 7.23%. Similarly, BlackRock Inc. is paying a 7.53% yield on blue-chip securities sold in July for a Texas data center project.
The bond market surge is being driven by unprecedented government borrowing costs, with the U.S. government already paying $931 billion in interest on its debt through the first 10 months of its fiscal year, according to Bloomberg Intelligence. This represents more than it spent on health, national defense or veterans benefits and is just behind Social Security and Medicare spending. The total U.S. government bond market reached $31.5 trillion as of July, creating intense competition for capital as the federal government's total debt has surpassed $40 trillion, a staggering record. The 10-year Treasury yield has risen to 4.74%, matching its highest point in more than a year, with Treasury Secretary Scott Bessent announcing plans to double bond buybacks to bring yields down. However, these measures have provided only temporary relief as global investors increasingly seek higher-yielding alternatives.
The trend reflects intense competition for capital as firms pour hundreds of billions of dollars into artificial intelligence infrastructure. As reported by Business Standard, companies have already borrowed more than $410 billion for data centers and other AI investments this year, according to data compiled by Bloomberg News. Steven Schweitzer, a high-yield portfolio manager at Advent Capital Management, noted that "high-yield investors become tourists in investment-grade technology debt" when they can buy fortress balance sheets at yields that look like double B-rated debt. The world's biggest bond market faces more competition from higher-yielding bonds overseas than in recent decades, with Japanese government bonds now paying more than 4%, U.K. bonds reaching 5.81%, and German bonds at 3.76%, versus 5.27% for comparable U.S. bonds. This global shift is forcing investors to reconsider traditional Treasury investments.
Big tech companies are borrowing at unprecedented levels, marking a sharp contrast to their traditional funding methods. According to Business Standard, for decades, tech companies could fund necessary investments first with equity and later with cash flow from profitable businesses. However, with vast upfront costs required for AI, they're now selling mountains of debt for projects where payoffs could be years away, making financing riskier. Mark Malek, chief investment officer at Siebert Financial, warned that "the tech guys are borrowing the money, but they're also borrowing money at high costs, which is going to change the weighted average cost of capital." The U.S. government's debt-to-GDP ratio continues to climb, with officials at the Federal Reserve and economists warning for years that the U.S. government is on an unsustainable path with spending versus revenue. This creates additional pressure on the bond market as investors increasingly question whether the government can service its massive debt burden.
The debt trend is expected to continue as AI spending projections remain massive despite rising borrowing costs. As reported by Business Standard, Vanguard Group Inc. wrote that tech companies known as hyperscalers could spend almost $800 billion this year on AI, and more than $1 trillion every year from 2027 to 2030. Most of this expenditure will be funded through debt markets. Andrew Keches, co-head of US high grade research at Barclays Plc, noted that "there's a big question every time you're underwriting a new deal of how much more is behind that, so investors are simply asking for more to feel comfortable underwriting these bonds." However, a measure in the bond market that shows how worried bond investors are about potential defaults by several big economies' governments has not risen excessively, according to strategists at Macquarie. This suggests that while yields are climbing, investors are not yet panicking about government solvency, though the US junk bond market's size of $1.5 trillion is less than a fifth of its high-grade counterpart, creating potential liquidity constraints for AI debt financing.
