From Contracted Megawatts to Cash Flow: The Power Readiness Challenge for Neoclouds
Part 1 of 2 | AI Infrastructure Private Credit Series
AI infrastructure financing is expanding rapidly, but the ability to deliver electricity to new data centers is becoming an increasingly important factor in private credit underwriting.
In October 2026, S&P Global Ratings reported $105.3 billion in cumulative debt issuance across 32 rated North American digital infrastructure transactions, including project finance, asset-backed securities, and corporate debt structures.¹ Meanwhile, the International Energy Agency (IEA) projects global data center electricity consumption will nearly double from 485 terawatt-hours (TWh) in 2025 to approximately 950 TWh by 2030.²
For neoclouds and AI compute operators, this growth creates an important distinction between contracted power capacity, operational computing infrastructure, and revenue-generating assets.
The relevant underwriting question is not simply how much power an operator has contracted. It is when that capacity becomes available to support deployed computing assets, enforceable customer payment obligations, and dependable cash flow.
This two-part series examines how electricity availability influences AI infrastructure credit. Part 1 focuses on the operational and contractual risks facing neocloud operators. Part 2 considers how those risks should influence financing structures, advance rates, and lender protections.
The Gap Between Contracted Power and Operational Capacity
The scale of projected electricity demand from AI infrastructure is significant. However, announced power commitments and interconnection requests should not be confused with infrastructure ready to support operational computing capacity.
In April 2026, the Electric Reliability Council of Texas (ERCOT) reported a large-load interconnection pipeline approaching 410 gigawatts (GW) through 2030, with data centers accounting for approximately 87% of requested capacity.³
Importantly, approximately 294 GW of that projected pipeline represented projects for which interconnection studies had not yet been submitted.
This distinction matters for credit investors. A substantial portion of announced capacity remains at an early stage of development rather than representing electricity infrastructure committed and ready for delivery.
For neocloud operators that depend on third-party data centers, the maturity of a power commitment can be as important as the amount of power ultimately expected to become available.
The distinction is also visible in public company disclosures.
Core Scientific, a data center infrastructure provider supporting AI compute operators, reported approximately 1,120 megawatts (MW) of leased customer power capacity as of September 30, 2026, compared with 485 MW of billable capacity. The company identified equipment availability, permitting, interconnection sequencing, construction, and customer deployment schedules as factors influencing the conversion of contracted capacity into billable revenue.⁴
Similarly, CoreWeave, an AI cloud infrastructure operator, reported more than 4.6 GW of contracted power and approximately 1.7 GW of active power in its third-quarter 2026 results.⁵
These figures reflect different operating portfolios, contractual arrangements, and stages of infrastructure deployment. The difference between contracted and operational capacity does not necessarily indicate delays or credit deterioration. Much of the capacity may be progressing according to planned development schedules.
Nevertheless, these disclosures illustrate the execution requirements involved in converting future power commitments into operational computing infrastructure.
For lenders, contracted megawatts provide visibility into future capacity. An important financing consideration is when those commitments translate into operational GPU clusters, enforceable customer payment obligations, and cash flow available for debt service.
How Power Delivery Affects Neocloud Financing
Neocloud operators face an infrastructure development cycle that extends beyond securing GPUs and customer contracts.
Computing capacity must be supported by suitable data center facilities, electricity, cooling infrastructure, networking equipment, and operational readiness. The delivery schedules for these components do not always align.
The IEA's February 2026 electricity outlook estimates that new grid infrastructure can require five to fifteen years to plan and complete, compared with approximately one to three years for data center construction.⁶
For neocloud operators, this mismatch can become important when financing commitments precede the availability of operational computing infrastructure.
Consider an operator acquiring a new generation of GPUs to satisfy anticipated customer demand. The company may have negotiated equipment financing, secured hardware allocations, and entered into customer agreements. However, the associated facility may still require utility upgrades, commissioning, or additional electrical infrastructure before the GPUs can be deployed.
Depending on the financing structure, several issues may emerge:
Debt carrying costs. Interest expense or lease obligations may begin before financed assets generate sufficient revenue.
Liquidity requirements. Delays can increase working capital needs, consume cash reserves, and create equipment storage or infrastructure-related costs.
Collateral exposure. GPUs may experience technological depreciation or declining resale values while awaiting deployment.
Customer contract performance. Revenue commencement may depend on cluster availability, service acceptance, and contractual delivery milestones.
Refinancing risk. Delayed deployment may shorten the period of operating cash flow available before scheduled debt amortization or maturity.
The degree of exposure depends on how borrowers, lenders, infrastructure providers, and customers allocate these risks.
Moody's Asset Management Research highlighted related concerns in July 2026, emphasizing the distinction between infrastructure completion, operational readiness, and sustained revenue generation.⁷
For neocloud operators, another consideration is that power availability may depend on a third-party colocation provider rather than a direct utility relationship.
Lenders may therefore need to evaluate not only customer contracts and GPU collateral, but also the underlying colocation agreements, facility delivery commitments, and remedies if infrastructure is unavailable as scheduled.
Contracted Revenue Does Not Eliminate Deployment Risk
Customer contracts can substantially improve the credit profile of AI compute infrastructure financing.
Long-term commitments, minimum revenue obligations, investment-grade counterparties, and contractual payment protections provide lenders with visibility into future cash flow.
However, contracted revenue and immediately enforceable payment obligations are not necessarily equivalent.
CoreWeave reported approximately $132.5 billion in revenue backlog as of September 30, 2026. Its disclosures explain that future revenue recognition depends on the satisfaction of delivery and service availability requirements.⁵
For credit investors, this makes the underlying customer agreements particularly important.
Relevant considerations include when customer payment obligations begin, whether payments depend on hardware acceptance or operational availability, and what remedies apply if infrastructure delivery is delayed.
Contracts with strong payment commitments, creditworthy counterparties, and limited termination rights can materially reduce risk. Conversely, contracts with conditional commencement dates, performance requirements, or significant cancellation provisions can increase the period between capital deployment and revenue generation.
This does not mean contracted AI infrastructure revenue is inherently uncertain. Rather, the strength of the credit depends on the specific contractual obligations and the operator's ability to deliver computing capacity.
Electricity Costs Introduce Another Consideration
Power availability is only one component of the underwriting equation. The cost and contractual terms of electricity supply can also influence GPU infrastructure economics.
Electricity procurement may expose operators to wholesale price movements, transmission charges, demand-based tariffs, and infrastructure-related expenses.
The extent of this exposure varies by operating model.
Some neocloud operators benefit from arrangements that pass electricity costs through to customers. Others enter into fixed-price customer agreements while retaining a portion of variable power expense.
Long-term supply contracts, colocation agreements, and customer cost pass-through provisions can mitigate this exposure. Their effectiveness, however, depends on pricing formulas, contract duration, and the allocation of unexpected charges.
For lenders, the question is not simply the cost of electricity but whether the operator has sufficient pricing flexibility and contractual protection to preserve operating margins throughout the financing term.
From Power Commitments to Financeable Cash Flow
The growth of AI infrastructure creates substantial demand for debt capital, but not every contracted megawatt carries the same financial significance.
A borrower with confirmed power availability, operational computing infrastructure, and established customer billing presents a different credit profile from one whose projections depend on future facility delivery.
Similarly, a borrower may have strong customer contracts and financing commitments while still facing deployment risks associated with power, cooling, networking, or equipment availability.
Power-related delays do not automatically result in credit deterioration. Strong customer agreements, liquidity reserves, sponsor support, and appropriately structured financing can reduce risks.
For neocloud operators and their capital providers, the practical priority is aligning power readiness, GPU deployment, customer obligations, and debt repayment schedules.
The quality of that alignment should directly influence how AI infrastructure debt is structured and financed.
Coming Next: Part 2 — Structuring Private Credit for AI Compute Infrastructure
In Part 2, we examine how lenders can translate these operating risks into financing terms, including GPU-backed warehouse facilities, equipment financing, collateral advance rates, phased funding, liquidity protections, and debt amortization.
The discussion moves from identifying power and deployment risks to understanding how private credit structures can allocate and mitigate them.
About Axis Group Ventures
Axis Group Ventures is a boutique investment banking and strategic advisory firm focused on global debt capital markets and private-market secondaries for venture- and private equity-backed companies. The firm works with founders, C-suite executives, and institutional investors to develop and execute customized financing solutions, including corporate debt, asset-based financing, equipment financing, and other private credit structures.
For more information, visit www.axisgroupventures.com.
Sources
S&P Global Ratings, North America Data Center Dashboard, October 23, 2026.
International Energy Agency, Key Questions on Energy and AI, April 16, 2026.
ERCOT, ERCOT Update: Senate Committee on Business & Commerce, April 1, 2026.
Core Scientific, Form 10-Q, Quarter Ended September 30, 2026, November 5, 2026.
CoreWeave, Third Quarter 2026 Financial Results, November 10, 2026.
International Energy Agency, Electricity 2026: Grids, February 6, 2026.
Moody's Asset Management Research, Power Without Delivery: Emerging Credit Implications, July 31, 2026.
Disclosures & Disclaimers
This blog post is provided by Axis Group Ventures for informational and educational purposes only. It does not constitute investment, legal, accounting, or tax advice and should not be relied upon as such. Nothing contained herein should be interpreted as an offer to buy or sell any securities. Any actual offer or solicitation will be made exclusively through formal documentation provided by the relevant issuer or seller.
Axis Group Ventures is not a registered broker-dealer and does not execute, negotiate, or recommend the purchase or sale of securities. Any introductions or private-market support provided by Axis Group Ventures are conducted strictly in an advisory and consulting capacity. Readers should conduct their own due diligence and consult qualified professionals before making any financial decisions.
Investments in private securities involve significant risks, including the potential loss of the entire investment, and are typically illiquid. Past performance does not guarantee future results.
This blog post was developed using Axis Group Ventures’ proprietary market knowledge, professional insights, and years of industry experience, supplemented by large language model technology to assist with drafting and editorial refinement.
Comments