Independent analysis of data centre power architecture, autonomous operations, and the infrastructure behind AI. By Adi Kumar.

Financing the AI Buildout: ten capital-structure essays

Financing the AI Buildout · ten-essay series

Financing the AI Buildout. Ten capital-structure essays

Ten essays on the specific capital structures funding the AI infrastructure buildout: from DDTL contract-collateralised debt through public-market REITs, sovereign wealth participation, sale-leaseback + green bonds. Each essay decomposes the structure, maps the lender + buyer universe, quantifies pricing evolution 2020-2026, and closes with a retrospective scorecard.

10essays
Retrospective scorecardsevery essay

01The essays

  1. I
    DDTL + contract-collateralised debt
    $95B by 2026. Blackstone/Ares/Sixth Street/KKR
  2. II
    Private-market equity structures
    $200B+ deployed 2020-2026. BREIT + Brookfield + DigitalBridge
  3. III
    Hyperscaler contracted revenue securitisation
    ABS structures. Estimated $5B → $115B 2026-2030
  4. IV
    Nuclear PPA financing structures
    Restart + SMR + traditional stacks compared
  5. V
    Sale-leaseback for hyperscaler DC
    $25B 2026 volume. Cap rates 5.5-7.5%
  6. VI
    Green bonds + sustainable finance
    $40B+ annual AI-DC green bond issuance
  7. VII
    Insurance + risk transfer
    30-80 bps of annual capex. Captives + reinsurance
  8. VIII
    Sovereign wealth + international capital
    $80-120B deployed. PIF + Mubadala + GIC + Temasek
  9. IX
    Public REITs + public-market vehicles
    ~$150B market cap. DLR + EQIX + REIT arbitrage
  10. X
    Hyperscaler buy-vs-build (series close)
    MSFT + AWS + META + GOOG capital-mode comparison

DFramework calibration case study

Live-deal application of the uncontracted-principal veto on a DDTL structure.

  1. D1
    CoreWeave $2.6bn DDTL 3.0 · the AMBER worked example
    SEC filing crwv-20250728 (Jul 2025) · SOFR+400 + 0.50% undrawn, Ba2/BB+, parent-guaranteed. The market-tested amber pattern.

Working on an AI infrastructure decision?

I advise PE sponsors + LP allocators + hyperscaler treasury on AI infrastructure capital-structure design, DDTL lender selection, and refinancing sequencing.

How I work →

Related: Case Studies, Financing hub, M&A hubFull series index: /tag/financing-the-ai-buildout/.

Definition Financing the AI Buildout is a ten-essay series covering the capital structures deployed to fund the 2025-2030 AI infrastructure buildout. Coverage runs from delayed-draw term loans backed by hyperscaler contracts through nuclear PPAs, sale-leaseback, green bonds, sovereign wealth commitments and REIT vehicles.