Author: Dong Jing, Wall Street Insights
The AI infrastructure arms race is pushing the balance sheets of tech giants to historical limits, sounding alarm bells in the bond market.
Amazon launched a new round of $25 billion in dollar-denominated bond issuance on Tuesday. Immediately after the announcement, AI-related bonds experienced a massive sell-off. Spreads on bonds issued by hyperscale cloud computing providers such as Meta and Amazon widened significantly in a single day, with declines far exceeding the daily fluctuation range of the investment-grade market.
According to Bloomberg data, if this issuance is successfully completed, it will push global AI-related bond sales to a record $335 billion this year, more than double the total for 2025. Meanwhile, Goldman Sachs issued a warning to clients last weekend, recommending proactive hedging against the risk of a massive cloud computing debt bubble. The core of market concern lies in the fact that the "bottomless pit" nature of AI capital expenditure is eroding the credit quality of tech giants at an unprecedented rate. According to Morgan Stanley data, the overall leverage ratio of hyperscale cloud computing providers has surged from 0.9 times in the third quarter of 2025 to the current 1.8 times, doubling in just two quarters and surpassing the leverage level of the entire energy industry, with a growth rate of approximately 0.3 times per quarter. John Lloyd, Global Head of Multi-Credit at Janus Henderson, pointed out that portfolio holdings in hyperscale cloud computing providers are already quite concentrated, and new debt issuances must offer sufficient price concessions to attract investors to continue increasing their holdings. Amazon is once again "snap up" the bond market, with a single issuance reaching a new high for the year. This Amazon bond issuance is divided into up to eight maturities, ranging from 3 to 40 years, with the longest maturity being bonds maturing in 2066. The initial price guidance is a premium of approximately 145 basis points over US Treasury bonds. The underwriters are Barclays, Goldman Sachs, JPMorgan Chase, and Morgan Stanley. The proceeds will be used for general corporate purposes, including debt repayment, mergers and acquisitions, and capital expenditures. This is yet another massive financing round for Amazon this year. In March, the company issued $37 billion in dollar bonds, setting a record for the fourth largest single issuance in US corporate bond history. During the same period, it also issued €14.5 billion in euro bonds, and subsequently completed issuances in the Swiss franc and Canadian dollar markets. According to Bloomberg data, Amazon has raised approximately $82 billion through multi-currency high-grade bonds since last year, while the company's capital expenditures this year are projected to be close to $200 billion, leaving a significant funding gap. This issuance will complete Amazon's dollar financing plan for 2026; any subsequent dollar bond issuances will be opportunistic. However, analysts generally believe that this does not mean Amazon's overall bond issuance has come to an end. AI Bond Spreads Widen Sharply in One Day, Market Selling Pressure Erupts Following the announcement of Amazon's new bond issuance, the secondary market for AI-related bonds immediately came under pressure. According to MarketAxess data, the spread of Meta's 6.3% bond maturing in 2056 widened by approximately 8 basis points in a single day; the spread of Amazon's 5.95% bond maturing in 2066 widened by approximately 14 basis points. In contrast, the spread fluctuation of high-rated corporate bonds on normal trading days is typically only a few basis points. John Lloyd stated that the widespread weakness in ultra-large-scale cloud computing corporate bonds on that day primarily reflects investors shifting funds to participate in Amazon's new bond subscription, rather than a substantial change in credit fundamentals. However, he also pointed out that investors are increasingly questioning how much additional debt the industry will need to borrow in the coming years to finance AI infrastructure. SpaceX's $25 billion bond issuance in June was also not spared, with related bonds being dragged into this sell-off. According to BondCliQ, the six major players in the AI race—Amazon, Microsoft, Alphabet, Meta, Oracle, and Nvidia—have cumulatively issued over $460 billion in bonds in recent years, with approximately $100 billion issued this year. Echoing the volatility in the bond market, the credit default swap (CDS) market has also shown significant fluctuations. Oracle's CDS spreads are widening rapidly, and Coreweave and SpaceX's CDS are showing similar trends, indicating that the credit market is beginning to re-evaluate the risk pricing of AI-related names. Oracle's stock price has fallen nearly 50% from its June 2nd high, reaching its lowest level since 2026, and is considered a typical example of pressure on highly leveraged AI concept stocks. Morgan Stanley data shows that the overall leverage ratio of hyperscale cloud computing providers is rising at an alarming rate—from 0.9 times to 1.8 times in just two quarters, and the growth rate shows no signs of slowing down. This trend is highly similar to the credit tightening that occurred twice in late 2025 and March 2026, when tech-named CDS surged, temporarily halting the debt-driven capital expenditure wave. Goldman Sachs issued a warning, advising clients to proactively hedge against bubble risks. Faced with the aforementioned accumulating risks, Goldman Sachs trader Adam Crook released a report to clients last weekend, recommending proactive hedging strategies to cope with the market shocks that may be triggered by the continued expansion of hyperscale cloud computing debt. Crook pointed out in the report that a scenario exists where the stock market continues to reach new historical highs, while the credit market struggles to rise further from historically tight valuation levels—massive supply pressure will continue to suppress credit spreads, and higher interest rates will also hinder refinancing for existing issuers, especially low-rated issuers. Crook emphasizes that, **in the event of a macroeconomic shock—such as significant new information challenging the logic of the AI bull market or capital expenditure plans—the aforementioned credit market segment, currently at historically low spreads, will provide a risk hedge with asymmetric returns and convexity characteristics for macroeconomic or multi-asset portfolios.** It's worth noting that data from the public debt market may only represent the tip of the iceberg of the AI capital expenditure wave. According to analysis cited by ZeroHedge, **beyond disclosed capital expenditures and public debt, there is a potential off-balance-sheet exposure of up to $1.8 trillion, posing a systemic risk hazard.** Currently, spreads on AI-related bonds have been wider than those in the overall US investment-grade corporate bond market for several months. As the financing needs of hyperscale cloud computing providers continue to reach new highs, bond investors are facing increasing uncertainty about how to price this bottomless AI arms race.
Weatherly
Anais
Alex