AI Boom Creates New Pressure on US Bond Yields
Economists have consistently expressed astonishment at the economy’s resilience amid the challenges posed by the Covid-19 pandemic, geopolitical conflicts, tariffs, inflation, and significantly elevated interest rates. A straightforward response has offered a comprehensive rationale: the expansion of artificial intelligence. However, the expansion of AI presents a potential threat, as evidenced by the yields on government bonds in the United States reaching levels not seen in nearly two decades this week, leading the Treasury Department to attempt to control borrowing costs. That risk arises from a significant borrowing spree undertaken by several of the largest technology firms globally. Until recently, companies primarily financed the construction of data centers and other infrastructure necessary for their technologies through self-funding. Currently, they are generating hundreds of billions of dollars through bond sales to address the capital requirements of advanced AI systems. “For most of the past decade, the large technology companies leading the AI build-out have funded their investment from operating cash flow,” said Lucas Baynes. “That era is ending.” Supercharged spending on AI financed in part by the surge in new bonds — over $200 billion so far this year among the largest AI companies — has prompted economists and investors to raise their forecasts for growth in the broader economy.
Higher economic growth typically encourages the Federal Reserve to keep interest rates elevated to prevent that growth from leading to higher inflation. Analysts indicated that the recent increase in Treasury yields was partially a result of expectations that growth driven by AI might compel the Fed to maintain elevated rates. This week, the yield on the 30-year US government bond, a benchmark for consumer loans such as mortgages, reached its highest level since 2007. The yield experienced a temporary decline following the Treasury’s intervention in the market on Wednesday, which involved an increase in the volume of its own debt eligible for purchase. This action contributed to heightened demand, elevated prices, and a subsequent reduction in yields. However, on Thursday, yields commenced their ascent once more. That presents a challenge for policymakers. The Trump administration has committed to reducing borrowing costs to enhance affordability, utilising Treasury yields as a measure of its effectiveness. AI borrowing is not the sole factor contributing to the upward pressure on Treasury yields and interest rates in a broader context. Analysts have also highlighted the federal deficit and the ongoing conflict with Iran. However, the present convergence of risks in financial markets has rendered the AI build-out a double-edged sword.
While AI has bolstered the economy through its expenditures, it has simultaneously led to an increase in interest rates, thereby diminishing affordability for borrowers, which encompasses consumers, businesses, and the government itself. The economy “has enjoyed a massive boost” from the five so-called AI hyperscalers — Alphabet, Amazon, Meta, Microsoft and Oracle — spending their enormous stockpiles of cash to develop AI infrastructure, said Matt King, founder of Satori Insights, a market research firm. But that has changed, he said. “Now that the hyperscalers are having to borrow, further capital expenditure is no longer ‘free,’” King said, meaning there is not only a cost to the company in the form of interest but also a broader cost to the economy. “Real yields are rising, raising costs for the rest of the economy,” he added. Between 2020 and 2024, the five hyperscalers collectively issued an average of under $30 billion in debt annually, as indicated by data. In 2025, that figure exceeded $100 billion. As of the current year, it has surpassed the $200 billion mark. Microsoft stands out as the sole hyperscaler that has not engaged in bond market fundraising in the past year.
Broader AI-related debt issuance — beyond the five hyperscalers — is anticipated to exceed $1 trillion annually from 2027 through 2030, as reported by Vanguard. Some analysts have posited that the influx of new debt is diverting investors from the Treasury market, consequently driving up yields on government bonds. However, some argue that this is expected to exert only a minimal influence, considering the vastness of the Treasury market and the more pronounced effect that AI debt supply is having on borrowing costs within the corporate bond market. Nonetheless, the substantial influx of supply has proven to be quite challenging for bond investors to manage. Some companies have begun to increase their interest payments to attract investors. The interest on corporate bonds is generally priced as a differential over government debt of equivalent maturity, referred to as the “spread.” Alphabet, the parent company of Google, enjoys some of the most favourable corporate borrowing costs. It issued 10-year bonds in April, offering a spread of 0.63 percentage points over 10-year Treasury notes. It issued another 10-year note this month, offering a spread of 0.85 percentage points.
