The Shift from Silicon to Steel
For the past two years, the AI narrative has been dominated by equity valuations and chipset performance. As we close 2025, the story has fundamentally shifted to the balance sheet. The physical build-out of artificial intelligence—the data centers, cooling systems, and power grids—is no longer just a technology story; it is a credit story. According to recent data, AI project financing deals have surged to $125 billion this year, up nearly eightfold from $15 billion in 2024. While equity markets digest the promise of generative capacity, debt markets are now financing the steel and concrete required to deliver it.
This massive capital expenditure cycle, estimated by JPMorgan to require near $5 trillion in total investment, is testing the limits of traditional financing. Public markets are becoming crowded, and with banks restricted by capital requirements, a significant funding gap has emerged. This dislocation is driving borrowers toward private credit solutions to bridge the estimated $1.4 trillion shortfall needed to keep the infrastructure rollout on schedule.
The Great Risk Transfer
Beneath the headline numbers, a more complex—and potentially fragile—financial structure is emerging. Major technology firms, including Microsoft and Meta, are increasingly utilizing creative deal structures to keep massive capital expenditures off their primary balance sheets. By signing massive lease agreements with intermediaries rather than taking on direct debt, these hyperscalers are effectively engaging in a risk transfer, pushing exposure down to smaller developers and their lenders.
The market state currently exhibits signs of "irrational exuberance" typical of a Risk-On frenzy. Reports indicate that some aggressively positioned borrowers are requesting loan-to-cost ratios as high as 150%, justifying these leverage levels based on projected future valuations rather than current tangible assets. This behavior signals a market priced for perfection, where lenders are loosening covenants to win allocation in the hottest sector of the economy. When capital treats projections as collateral, the risk of a correction rises significantly.
The Bond Capital Perspective: Financing Assets, Not Hype
At Bond Capital, we view the AI infrastructure boom through the lens of asset-based security. While the technology is novel, the underlying collateral—industrial real estate, power infrastructure, and high-quality lease receivables—is familiar. However, the current lending frenzy requires a disciplined hand.
We are seeing terms in the broad market that ignore fundamental credit discipline. Lending against "future valuation uplift" without robust debt service coverage is a strategy for a bull market that never ends. Bond Capital takes a different approach. We recognize that AI infrastructure is the industrial revolution of our time, but we finance it based on today’s realities.
- Real Assets Over Speculation: We focus on the physical plant and the creditworthiness of the offtake tenants, not theoretical valuation multiples.
- Structure Matters: In a market utilizing synthetic leases and circular financing, we prioritize clean, senior-secured structures that protect capital if the "toothpaste tube" of risk bursts.
- Certainty of Execution: For borrowers navigating a volatile public market or a banking sector immobilized by regulatory caps, Bond Capital offers speed and certainty—structured correctly.
The infrastructure build-out is necessary, but how it is funded will determine who survives the inevitable cycle turn. We remain active, but vigilant.
