Fixed Income: The AI Issuance Boom and What it Means For Portfolios

AI-related borrowing is reshaping fixed income by flooding public and private markets with issuance, increasing dispersion and forcing investors to rethink benchmark-based portfolio construction.

• Four companies issued $187 billion over the past year across maturities and currencies, while corporate net issuance has exceeded Treasury net issuance this year.
• Apollo estimates $3 trillion to $5 trillion of AI-related issuance, with roughly $800 billion of capacity for large issuers in major indices; AI exposure in the Bloomberg Aggregate has risen from 0% to 4.5% in one year.
• The financing need could reach $5 trillion by 2030, but the speakers argue the market could reach the target by 2028 and that benchmark-aware investors may become saturated with public investment-grade corporate debt before 2029.
• Apollo’s Valor and xAI GPU financings were conservatively structured, assumed no residual value and still priced at wide spreads, implying that private-market pricing can pressure comparable public bonds.
• In high yield and leveraged loans, software trades at its widest level in five years while non-software is at its tightest; AI-related data-center bonds have shown sharply different structures and price moves of two or three points in a week.

Takeaway: Investors should broaden origination across public and private credit, compare structures and pipeline risk, and seek active sources of spread and alpha rather than relying on the traditional government-heavy Aggregate benchmark; the adjustment may also widen risk across financials, cable companies and other capital-intensive borrowers.