Home โ€บ Blog โ€บ Who Pays for the AI Build-Out?
AI Compute Stack ยท Part 14 ยท Finance

Who Pays for the AI Build-Out? Capex, Debt, Neoclouds and the Circular Money Machine

In 2026 four companies alone will spend roughly $700 billion on data centres, more than the annual GDP of most countries. Part of it comes from the most profitable businesses ever built. A growing share comes from bonds, private credit, GPU-backed loans, off-balance-sheet joint ventures and deals in which chipmakers invest in their own customers. This guide follows the money layer by layer, asks whether the revenue is there to pay it back, and shows who is left holding the risk if it is not.

FL
FrontierAGI Team
โš ๏ธ How this was researched This draws on web searches run up to 23 October 2026, before Microsoft, Alphabet, Amazon and Meta report third-quarter results, so capex figures are earlier guidance and may change. Company-level capex, bond totals, lab revenue run rates and the size of "circular" commitments come largely from secondary reporting and trackers that disagree; ranges are shown. Revenue figures for OpenAI and Anthropic are press reports of private companies, and the two count revenue differently. Michael Burry's depreciation figure and Sequoia's 2026 gap estimate are reported second-hand. All calculators are illustrative. See Reading the numbers. This is an explainer, not investment advice.

๐Ÿ’ฐ About $700 billion in one year

The earlier parts of this series followed electricity, factories, chips, packaging, networks and memory. Every one of them is paid for by someone. In 2026 the four biggest spenders, Amazon, Alphabet, Microsoft and Meta, are on track for combined capital expenditure of roughly $700โ€“745 billion, up from about $410 billion in 2025, according to trackers compiled from company guidance. Add Oracle, the neoclouds, the AI labs' own projects, sovereign programmes and Chinese spending, and the total is far larger.

For most of the boom this was paid for out of profits. That is changing. One estimate has capital spending absorbing about 94% of the tech sector's operating cash flow in 2025โ€“2026, and the gap is increasingly filled with borrowed money, much of it in structures designed to keep it off the biggest companies' balance sheets.

~$700โ€“745BBig Four capex in 2026 (trackers; varies by lease treatment)
$121BHyperscaler US bond issuance in 2025, vs ~$28B a year in 2020โ€“24 (BofA)
$27.3BBonds sold for one Meta data centre, kept off Meta's balance sheet
9.12%Interest on CoreWeave's latest GPU-backed loan (Aug 2026)

๐Ÿ—บ๏ธ The money map

The money moves in three stages: sources of capital, the builders who spend it, and the suppliers who receive it. Some arrows loop back: a supplier invests in a builder, who spends the money with that supplier.

๐Ÿ”€ Follow a dollar

Click any box for details. Toggle "Show circular loops" to highlight money that returns to where it started.

Simplified. Flows are illustrative of structure, not to scale.

๐Ÿฆ Layer 1: paying from cash flow

The hyperscalers' advantage is that they own some of the most profitable businesses in history: search and YouTube ads, Office and Azure, Instagram and WhatsApp ads, AWS. Their operating cash flows run to well over $100 billion a year each, so they can fund most of their build-out themselves. But capex is now growing faster than cash flow, squeezing free cash flow and buybacks, and pushing even these companies to borrow.

๐Ÿ“Š 2026 capex guidance by company

Midpoint of reported guidance before Q3 2026 results. Ranges in the labels; sources differ.

Compiled from capex trackers (Introl, CreditSights, others). Amazon may have raised to ~$220B; Alphabet reported as $180โ€“190B or $195โ€“205B; Microsoft ~$175โ€“190B depending on how finance leases are counted; Meta $125โ€“145B. Check Q3 earnings releases (late October) for current figures.

๐Ÿ“œ Layer 2: the bond wave

Until 2024, the biggest tech companies barely borrowed. Bank of America counts US corporate bond issuance by the five AI hyperscalers (Amazon, Alphabet, Meta, Microsoft, Oracle) at about $28 billion a year on average in 2020โ€“2024, then $121 billion in 2025. Big deals came one after another: Oracle $18 billion (September 2025), Meta $30 billion (October 2025), Alphabet $17.5 billion and Amazon $15 billion (November 2025), and Alphabet and Oracle together nearly $60 billion in February 2026. By August 2026 Big Tech had issued roughly $200 billion of bonds, and Goldman Sachs counted $489 billion of AI-related debt worldwide.

๐Ÿ“ˆ Hyperscaler bond issuance

US corporate bonds, $ billions. The 2026 bar is a partial year and uses a different, broader count.

BofA via Reuters (2020โ€“24 average, 2025); Business Today (2026 to August, "Big Tech"). Not strictly comparable.

๐Ÿšฉ Signs of strain Investors increasingly hedge AI borrowers with credit default swaps (insurance against default). MUFG reported that Oracle's five-year CDS spread more than tripled after its September 2025 bond sale, and spreads have also risen for Broadcom, Meta and Nvidia. Oracle, which has committed to build huge capacity for OpenAI, was sued by bondholders who say it failed to disclose how much more debt it would need. Most analysts still see the big four as very strong credits; the stress is concentrated in more leveraged players.

๐Ÿงฉ Layer 3: off the balance sheet

The biggest private-credit deal ever done shows the new playbook. To build its Hyperion AI campus in Richland Parish, Louisiana, Meta formed a joint venture with Blue Owl Capital. The venture, a special-purpose vehicle (SPV), sold $27.3 billion of fully amortising senior secured notes due 2049, anchored by PIMCO and structured by Morgan Stanley. Blue Owl owns about 80% and Meta about 20%. S&P rated the notes A+, yet they paid about 6.58%, well above Meta's own bonds.

๐Ÿ—๏ธ How the Hyperion deal works

Click the numbered steps.

Why do it? The debt does not appear as Meta's borrowing; rating agencies treat Meta's commitment as a lease. That protects Meta's credit metrics and spreads the funding across investors hungry for long, high-quality paper. The catch: through the lease and a residual value guarantee, Meta still carries most of the risk. If it walks away and the campus is worth less than promised, Meta pays the difference. Critics note that S&P's rating leaned heavily on Meta's own AA- credit. Other hyperscalers and labs have been reported to be pursuing similar joint ventures.

โ˜๏ธ Layer 4: neoclouds and GPU-backed debt

Neoclouds are companies that rent GPUs, not general cloud services: CoreWeave, Nebius, Lambda, Crusoe, Fluidstack and hundreds of smaller operators. Their model is simple and highly leveraged. Sign a multi-year contract with a big customer (Microsoft, OpenAI, Meta, Google), borrow against that contract and the GPUs themselves, buy the GPUs, and repay the loan from the rent.

CoreWeave
The largest. Its Q2 2026 filing showed $35.6 billion of debt and $16.3 billion of operating-lease liabilities, against $103.7 billion of contracted future revenue (41% due by 2028). Its GPU loans sit in separate borrowing entities secured by chips and customer contracts. The August 2026 "DDTL 5.5" facility cost SOFR + 5.5% (about 9.12%), up from about 8.05% in May, and was rated one notch below investment grade. Goldman Sachs estimated it still needs up to ~$23 billion of new debt to deliver its backlog. Its bond yields and CDS spreads jumped between 15 and 20 October 2026.
Lambda
Closed a $1.0 billion fixed-rate loan at 6.78% in October, rated Baa1 (investment grade) because it is backed by two investment-grade customers.
Nebius
Funds its Microsoft and Meta capacity largely with asset-backed debt tied to those contracts, including a $775 million GPU-backed term loan.
Crusoe ยท Fluidstack
Crusoe has leaned on equity (a reported $3.9 billion Series F). Fluidstack has arranged roughly $10 billion of GPU-collateralised financing.
๐Ÿ’ก The key idea A neocloud loan is only as good as two things: the customer contract behind it, and what a GPU is worth when the contract ends. That is why loans backed by investment-grade customers price near 6โ€“7% while others cost 9% or more, and why the depreciation debate below matters so much to lenders.

๐Ÿงฎ Does a GPU pay for itself?

๐ŸŽ›๏ธ GPU loan payback model

One GPU bought with borrowed money and rented out. All defaults are illustrative, not any company's figures.

Simple model: straight-line depreciation over the useful life, interest on the average outstanding balance, operating cost per rented hour. Ignores taxes, residual value and contract terms.

๐Ÿ” Layer 5: circular deals

The most debated financing comes from the suppliers themselves. In several deals, a company that sells chips or cloud capacity invests in, or extends credit to, a customer who then spends the money on that same company's products.

๐Ÿ”„ The deals

Click a deal.

Estimates of all the interlocking commitments range from about $800 billion to $1.5 trillion, and the counts overlap. Defenders call this ordinary vendor financing, which industries from aircraft to telecoms have long used, and note the chips are real and in use. Critics point out that it inflates the suppliers' revenue with money they supplied, ties many companies' fortunes to one customer (OpenAI), and resembles the telecom-equipment financing that unwound badly in 2001.

๐Ÿ“‰ Is there enough revenue to pay it back?

In June 2024, Sequoia partner David Cahn asked "AI's $600B question": how much revenue would AI need to justify the GPUs being bought? By his 2026 update, as reported, about $1.5 trillion of infrastructure spending implies roughly $3 trillion of revenue. Goldman Sachs uses a different method and puts breakeven at about $308 billion a year of AI revenue on 2026โ€“27 capex. The answer depends on assumptions, which you can set below.

Revenue is growing very fast. Bloomberg reported that OpenAI was at roughly a $50 billion annualised run rate in late September and expects $70 billion or more by year-end, though its numbers are disputed because it and Anthropic count cloud-partner sales differently. Anthropic was reported at a $65 billion run rate at the end of July, with investors reportedly projecting $100โ€“110 billion by year-end. Cloud AI revenue at the hyperscalers adds more. But these figures are still small next to the spending.

๐Ÿงฎ The revenue gap calculator

How much annual AI revenue does a year of capex need? Illustrative model in the spirit of Sequoia's calculation.

Big Four โ‰ˆ $700B; Cahn's broader count โ‰ˆ $1.5T
Clouds and AI apps must earn a margin on top of their compute cost

Required revenue = (capex รท years) ร— (1 + operating share) รท (1 โˆ’ margin). Comparison bar: OpenAI's expected ~$70B plus Anthropic's projected ~$100B year-end run rates (press reports, not audited, partly overlapping definitions).

โณ The depreciation fight

How long does a GPU last? The answer decides how much profit the build-out shows. Hyperscalers generally depreciate servers over five to six years. Investor Michael Burry has argued the real economic life of AI chips is closer to two or three years, because Nvidia releases a much better generation every year, and that the gap overstates profits by about $176 billion in 2026โ€“2028 (a figure critics say is directionally plausible but unproven). Amazon has already cut the life of some servers from six years to five.

๐ŸŽš๏ธ What useful life does to the books

$100 billion of GPUs bought in one year. Drag the life assumption.

Straight-line depreciation, no residual value. Real hardware is often redeployed to cheaper work (inference, smaller models) late in life, which supports longer lives.

โš–๏ธ Who holds the risk?

If AI revenue falls short, the losses do not land evenly. They flow down a ladder from the most protected investors to the least.

๐Ÿชœ The risk ladder and three scenarios

Pick a scenario to see who gets hurt. Qualitative and illustrative.

๐Ÿ•ฐ๏ธ Timeline

Jun 2024 ยท Sequoia's David Cahn publishes "AI's $600B question".
2024 ยท Hyperscaler capex climbs sharply; CoreWeave raises billions in GPU-backed loans.
Jan 2025 ยท $500B Stargate venture (OpenAI, SoftBank, Oracle and others) announced.
Mar 2025 ยท CoreWeave goes public.
Sep 2025 ยท Oracle's ~$300B OpenAI contract reported; Oracle sells $18B of bonds; Nvidia announces up to $100B investment in OpenAI.
Oct 2025 ยท AMDโ€“OpenAI 6 GW deal with warrants; Meta and Blue Owl's $27.3B Hyperion financing; Meta's $30B bond.
Nov 2025 ยท Alphabet ($17.5B) and Amazon ($15B) bond deals; depreciation debate goes mainstream.
Early 2026 ยท Reports that the Nvidiaโ€“OpenAI investment had stalled; Huang says it was "never a commitment". Alphabet and Oracle issue ~$60B of bonds in February.
Aug 2026 ยท Big Tech bond issuance passes ~$200B for the year; CoreWeave's 9.12% GPU loan.
15โ€“20 Oct 2026 ยท CoreWeave bond yields and CDS spike. Q3 hyperscaler earnings due at month-end.

๐Ÿ”ญ What to watch

๐ŸŽฏ The bottom line The AI build-out is no longer paid for only out of Big Tech's profits. Bonds, private credit, off-balance-sheet joint ventures, GPU-backed loans and supplier financing now carry a growing share, and much of it ultimately rests on a few customers, above all OpenAI, turning fast-growing revenue into profit. If they do, today's borrowing looks like a smart bet on a new utility. If revenue stalls, losses flow first to the most leveraged players, neoclouds and their lenders, then to suppliers whose sales depended on financing their customers, while the hyperscalers absorb a painful but survivable write-down.

๐Ÿ”ข Reading the numbers

๐Ÿ“š References