# AI Infrastructure Financing: State of the Market

> GPU-backed term loans, data-center securitization, project debt, and funding transmission. Provenance-tagged research framework through July 20, 2026. Not investment advice.

## Entry points

- Interactive report: https://ai-physical-infra-debt-analysis.vercel.app/
- Agent JSON: ./agent.json
- Portfolio project: https://magro.dev/projects/ai-physical-infra-debt/
- Portfolio demo: https://magro.dev/demos/ai-physical-infra-debt/
- Agent Research: https://magro.dev/agent-research/ai-infrastructure-financing/
- Essay: https://magro.dev/writing/ai-infrastructure-buildout-bubble/

## What this page is

A static interactive research report with:

- market size maps (stocks vs flows)
- CoreWeave financing ladder (DDTL 1.0 to 5.0)
- interactive recovery / required-spread model
- contract quality waterfall
- funding transmission channels
- crisis comparison by mechanism

## Principles

- Prefer structure-specific evidence over bubble slogans.
- Distinguish outstanding stocks from annual issuance flows.
- Tag provenance: verified, approx, unverified, analysis.
- Display light/dark is available in standalone mode; Agent Mode exports stay available outside embeds.