AI bond supply: Why the market can absorb the coming wave

The expected wave of AI-related investment-grade bond issuance should be manageable because it is modest relative to the market’s normal growth, issuer leverage is contained and demand remains strong.

• Hyperscalers are expected to spend around $6 trillion on AI infrastructure over the next five years, but only $1.5 trillion is forecast to be debt-funded and roughly $1 trillion to reach investment-grade bond markets.
• That implies about $200 billion of additional annual technology bond supply over five years, compared with more than $1 trillion of expected annual net new supply for the US investment-grade market and $1.75 trillion-$2 trillion of gross issuance including refinancing.
• Investment-grade EBITDA is expected to grow by more than 10% annually over the next couple of years, supporting the forecast of more than $1.8 trillion of net new supply between FY25 and FY27 despite modest deleveraging.
• Technology-sector leverage has fallen from 1.8 times earnings to 0.8 times over the past decade because earnings grew faster than debt; if borrowing had kept pace with earnings, the sector might already have issued about $1.1 trillion more bonds.
• Historical episodes show that bank issuance after the global financial crisis caused little spread disruption when borrowing plans were clearly disclosed, whereas telecom and energy issuance produced sharper widening amid faster leverage growth and less transparency.
• Demand from Asian savings, better-funded pension schemes, insurers matching long-dated liabilities and household inflows into investment-grade bond funds should help absorb the additional supply.

Takeaway: Investors should focus less on the headline volume of AI-related issuance and more on issuer fundamentals, leverage and transparency; clearly communicated and disciplined borrowing could make any resulting spread widening an opportunity rather than a major market threat.