LSEG: Morgan Stanley collected $2.3B in debt and equity capital market fees in H1 2026, up from $1.4B in H1 2025, driven by AI infrastructure financing deals
Context & Ripple Effects
Morgan Stanley’s fee increase is a bank-level measure of a broader shift from AI spending plans to financed projects. The firm had already forecast AI-linked debt issuance approaching $570B in 2026, following its estimate that hyperscalers would fund only part of projected infrastructure costs themselves.
The financing backdrop has widened beyond conventional corporate borrowing: AI-linked companies had become the largest segment of the investment-grade market in prior coverage, while smaller infrastructure borrowers faced more investor scrutiny over rates and business risk.
First-order effects
- Morgan Stanley captures substantially more underwriting and advisory revenue as AI infrastructure financings generate debt and equity capital-markets activity.
- Data-center developers gain another potential funding channel as Morgan Stanley pitches leveraged loans rather than relying solely on bonds; the bank estimates roughly $15B of such loans in 2026.
Second-order effects
- The fee pool gives rival investment banks a stronger incentive to compete for AI-infrastructure mandates and to build loan, bond and equity structures around data-center projects.
- A larger role for leveraged loans broadens the investor base for data-center finance, but also shifts more project exposure toward credit investors rather than public bondholders.
Third-order effects
- If issuance continues to migrate across bonds, leveraged loans and equity, AI infrastructure will increasingly be financed as a distinct capital-markets asset class rather than as ordinary technology capex.
- The pattern could deepen the link between AI buildout and credit-market conditions: access to capital may increasingly shape which developers and AI-related companies can scale.
The trend: AI infrastructure is becoming a major fee-generating finance market, with banks packaging compute and data-center expansion across multiple forms of private and public capital.