Gotrade News - The AI capex trade is facing a financing test in the bond market, with Goldman Sachs Asset Management now underweight the biggest AI borrowers as a flood of new hyperscaler debt washes over credit markets.
According to Seeking Alpha, Lindsay Rosner, head of multi-sector fixed income at the Goldman Sachs unit, said: "We believe there will be a lot of hyperscaler issuance."
The call lands on the companies funding the AI build-out, including Alphabet (GOOGL), Meta Platforms (META) and Oracle (ORCL), alongside Microsoft and Amazon, all tagged in the Seeking Alpha report.
Because Rosner runs a fixed-income book, the stance concerns hyperscaler bonds rather than a rating on the shares. The worry is supply: as reported by Reuters, Goldman Sachs data shows hyperscaler gross debt issuance is expected to reach $420 billion next year, up 60% from 2026 estimates.
Key Takeaways
Goldman Sachs Asset Management is underweight the largest AI borrowers on expectations of heavy hyperscaler bond supply.
Goldman data points to $420 billion of hyperscaler gross debt issuance next year, up 60% from 2026 estimates.
Bonds from AI-related issuers trade at spreads around 115 basis points, versus 78 basis points for the broader investment-grade market.
AI Bond Spreads Hold Near 115 Basis Points
Reuters described the market for highly rated corporate credit as split in two: bonds from AI-related firms are being met with caution, while traditional issuers such as financial and industrial firms draw spirited bidding.
Portfolio managers told Reuters they are not concerned about defaults. Instead, the sheer volume and unpredictability of borrowing needed to finance data centers, chips and AI infrastructure are pushing buyers to demand generous concessions and rethink portfolio concentration limits.
Per the same report, spreads on AI-related issuers have stayed persistently wider at around 115 basis points, compared with 78 basis points for the broader investment-grade market.
Alphabet had to offer a large concession to complete its August debt sale, and Russell Brownback, deputy chief investment officer for global fixed income at BlackRock, characterized some deals as double-A credits pricing closer to triple-B spread levels.
Metric | Figure | Source |
|---|
Expected hyperscaler gross debt issuance, next year | $420 billion (up 60% vs 2026 estimates) | Goldman Sachs data, via Reuters |
Spread on AI-related issuers | About 115 bps | Reuters |
Spread on broader investment-grade market | 78 bps | Reuters |
US corporate issuance through August | $1.9 trillion | SIFMA, via Reuters |
Calendar 2026 hyperscaler capex guidance | Close to $800 billion | FactSet |
Incremental annual debt as share of capex | 9% (FY24) to 32% (LTM, mid-2026) | FactSet |
Lon Erickson, portfolio manager at Thornburg Investment Management, told Reuters that bonds from major AI spenders such as Meta and Alphabet have consistently traded wider than similarly rated peers, even though both firms generate lots of cash and have strong balance sheets. "Investors are only able to digest so much, so fast," Erickson said.
Wellington Management's Loren Moran added: "There are a lot of investors that just want something other than hyperscaler debt for now. The market is a bit starved for anything ex-hyperscaler."
Hyperscaler Capex Near $800 Billion Outruns Cash Flow
The borrowing tracks the scale of spending. According to FactSet, aggregate capex at Alphabet, Amazon, Meta, Microsoft and Oracle is expected to exceed $690 billion in fiscal 2026, while calendar 2026 guidance points to close to $800 billion including finance leases and customer pre-payments.
Alphabet raised its 2026 capex guidance to $180 billion to $190 billion from $175 billion to $185 billion, and Meta lifted its range to $125 billion to $145 billion from $115 billion to $135 billion.
FactSet noted that fiscal 2026 free cash flow is expected to move close to zero or turn negative for every company in the group except Alphabet and Microsoft. Incremental annual debt rose from 9% of capex in fiscal 2024 to 32% over the last twelve months by mid-2026, taking aggregate total debt to about $700 billion.
Equity is part of the funding mix too. Alphabet priced an $84.75 billion equity raise in June 2026, per FactSet, and Amazon recently placed a $25 billion bond that was 2.5 times oversubscribed, below the 3.2 times coverage on its March issue.
Compute Spending Steps Up to $380 Billion
On the other side of the trade sit the suppliers of chips, servers and data-center gear. FactSet said compute represents the majority of the investment, with a significant step-up to $380 billion in 2026, and that the high-bandwidth memory (HBM) shortage is also lifting prices of other memory types, feeding into storage and networking. In custom silicon, FactSet pointed to Google leading the way with its tensor processing units (TPUs).
Brown Advisory's Colby Stilson told Reuters his team is being very selective within hyperscaler debt: "Our degree of investment conviction needs to be very high because of the coming supply and because of the lack of visibility into that return on invested capital."
For investors tracking the AI capex trade, the signals to watch are whether new hyperscaler deals keep requiring concessions, whether AI-related spreads move back toward the broader investment-grade market, and how capex guidance shifts at the next round of earnings.
Open a Gotrade account today and start investing in US stocks from $1.
Sources
