Big 5 AI ‘hyperscalers’ have quadrupled their use of debt to fund operations, Bank of America says

Bank of America analyst Yuri Seliger says five hyperscalers—Amazon, Google, Meta, Microsoft and Oracle—have issued $121 billion of debt this year, more than four times their five‑year annual average of $28 billion, including $27 billion for Meta’s Richland Parish data center and Amazon’s $15 billion issuance on Nov. 17. The flood of investment‑grade supply has materially widened hyperscaler credit spreads versus the broader market (Oracle +48bps, Meta +15bps, Google +10bps from Sept. 1 to Nov. 14), translating to 27–49% underperformance vs. the IG index. Seliger expects roughly another $100 billion of issuance next year; although the companies generate ample cash flow, market observers warn that debt-financed AI capacity buildouts are complicating the investment case for mega-cap tech stocks.

Analysis

Bank of America analyst Yuri Seliger reports that five hyperscalers—Amazon, Google, Meta, Microsoft and Oracle—have issued $121 billion of debt year-to-date, more than four times their five‑year annual average of $28 billion; that total includes Meta’s $27 billion data‑center financing in Richland Parish and Amazon’s $15 billion issuance on November 17. All issuances are investment‑grade and Seliger notes the companies continue to generate sufficient cash flow to cover operations.

The surge of investment‑grade supply has materially widened hyperscaler credit spreads versus the broader market: from Sept. 1 to Nov. 14 Oracle spreads moved +48bps, Meta +15bps and Google +10bps, translating to 27–49% underperformance versus the IG index. This reflects investor difficulty absorbing concentrated supply even with intact near‑term cash flows.

Seliger expects another ~$100 billion of debt issuance next year, implying persistent funding‑driven volatility in credit markets. Morgan Stanley’s Lisa Shalett warns that debt‑funded AI capacity buildouts complicate the prior, simpler tech investment case, so investors should reprice both credit risk and equity multiples for higher leverage and funding cost uncertainty.

Key Decisions for Investors

• Consider selectively increasing exposure to hyperscaler IG bonds to capture wider spreads but size positions conservatively, prefer shorter maturities or laddered buys because Seliger expects roughly $100 billion more issuance next year
• Monitor Oracle credit closely given a +48bps spread widening from Sept. 1–Nov. 14 and treat ORCL as a higher‑convexity credit risk versus other hyperscalers
• Reassess equity valuation models for AMZN, GOOGL, META, MSFT and ORCL to incorporate higher leverage for AI capex and require clearer free cash flow or ROI evidence before adding to exposure
• If portfolio sensitivity to spread widening is material, implement defensive measures such as reducing duration in corporate bond allocations or using targeted hedges rather than increasing outright exposure