🔍 Read the full analysis: How Leases And Bonds Help Pay For AI Data Centers on ThorstenMeyerAI.com
Get audio and creator gear delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment from July through September 2026, according to figures compiled from their filings. Their financing includes cash, bonds and loans, leases, partnerships and share sales; the scale of future lease payments and the duration of power and debt commitments leave questions about who carries the risk if AI revenue grows more slowly than expected.
Amazon, Microsoft, Alphabet and Meta spent about $202 billion on property and equipment from July through September 2026, more than twice their combined spending a year earlier, according to calculations based on company filings. To fund the data-center and AI build-out, the companies are drawing on operating cash, bonds and loans, leases, partnerships and share sales—a shift that spreads some costs and risks beyond their own balance sheets.
Recent financing moves illustrate the scale of the funding need. Meta said it sold $25 billion of its own stock and $12.5 billion of preferred shares. Microsoft borrowed $25 billion, including bonds with maturities reaching 40 years. Amazon disclosed that it had arranged a loan of up to $80 billion, secured by servers it will purchase with the proceeds. These transactions differ in structure and repayment terms, but each helps fund investment without relying solely on current cash generation.
The companies’ quarterly capital spending was reported as $77.3 billion for Amazon, $51.4 billion for Alphabet, $39.4 billion for Microsoft and $33.9 billion for Meta. Microsoft’s figure rises to $45.2 billion when new finance leases are included. Amazon said its spending increase primarily reflected AI investments; the companies’ broader capital budgets also cover other assets. Amazon expects about $220 billion of cash capital spending in 2026, while Alphabet has forecast $195 billion to $205 billion and Meta $135 billion to $145 billion. Microsoft expects about $170 billion for calendar 2026.
Cash flow figures help explain why funding methods are diversifying. In the quarter, Amazon generated $42.4 billion in cash from operations, Alphabet $42.3 billion, Meta $32.8 billion and Microsoft $58 billion. Those operating-cash figures are not the same as free cash flow, which accounts for capital spending. The source’s calculations put Amazon’s trailing-12-month free cash flow at negative $42.4 billion, and Alphabet’s and Meta’s quarterly figures at negative $9.1 billion and negative $2.1 billion, respectively. These measures and time periods differ, so they should not be treated as directly comparable.
Infographic 01 · One quarter, four companies
The bill doubled. The cash did not.
What Amazon, Microsoft, Alphabet and Meta paid for buildings, servers and chips from July to September 2026, set against the cash their businesses brought in.
AmazonQ3 2025: $35.1B spent
Spent $34.9B more than came in
AlphabetQ3 2025: $24.0B spent
Free cash flow −$9.1B
MicrosoftQ3 2025: $19.4B spent
$18.6B left over
MetaQ3 2025: $18.8B spent
Free cash flow −$2.1B
Cash purchases of property and equipment as reported; not all of it is for AI (Amazon also builds warehouses). Gaps for Amazon and Microsoft are my arithmetic; free cash flow for Alphabet and Meta as the companies report it. Microsoft’s quarter is its fiscal Q1 2027 (July–September 2026).
Sources: Amazon results, 29 Oct · Microsoft 10-Q, 28 Oct · Alphabet results, 28 Oct · Meta results, 21 Oct. Debt raised: cash flow statements, nine months 2026.
Powered by Thorsten Meyer AI
Who Carries the Data-Center Costs
The financing mix matters because the physical infrastructure is expensive and long-lived, while demand for AI services and the revenue they may generate remain uncertain. Bonds and loans put repayment obligations on borrowers; leases commit companies to future payments; equity sales raise money without the same fixed repayment promise but dilute existing shareholders. Partnerships and companies that own buildings can place some investment outside a technology company’s direct balance sheet, though the underlying costs and contractual obligations remain relevant.
The source calculates that the four companies had about $1.02 trillion in future payments on leases that had not yet started as of September 30, 2026, based on their quarterly filings. That figure is a sum of reported commitments, not an estimate of what is due immediately. It points to obligations that could continue even if a facility is delayed or demand changes. The source also notes that servers are typically written off over five to six years, while some data-center leases run as long as 20 years and certain bonds extend to 40 years. That difference in timelines is a central financing risk: revenue and equipment value may change faster than the commitments do.
The exposure is not limited to the largest firms. The source identifies newer GPU cloud providers, building-owning entities, investors and pension funds as possible holders of risk. Long-term electricity contracts can also affect utilities and, indirectly, electricity customers. The precise allocation depends on contract terms, which are not fully detailed in the figures presented.
Products Worth Considering
Anker SOLIX S2000 Portable Power Station, 2,010Wh, 1,500W
Bose QuietComfort Ultra Wireless Noise-Cancelling Headphones with Spatial Audio, Black
Sonos Era 100 SL Compact, Microphone-Free Speaker – Black
As an affiliate, we earn on qualifying purchases.
From Cash-Funded Builds to Financing Mix
For years, major technology companies could fund data centers largely from profits earned through advertising, software and online commerce. The 2026 investment wave has changed the scale of the bill. The source’s calculation puts combined property-and-equipment spending by Amazon, Microsoft, Alphabet and Meta at about $202 billion in the third quarter, compared with about $97 billion in the same quarter a year earlier. Those figures come from cash-flow statements and are the source author’s sum, rather than a single company-reported total.
Financing now combines several methods. Companies can pay from operating cash, issue bonds or take loans, lease facilities and equipment, enter partnerships, or sell shares. Each method has a different effect: debt adds interest and repayment obligations, leases create contracted payments, and stock sales exchange a portion of ownership for capital. Power agreements add another layer because data centers need electricity over time. No single method captures the full cost; reported capital spending and future commitments describe different parts of the build-out.
Meta stated on October 26 that it intended to fund AI investment “through a balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital.” That statement describes the company’s intended approach, not a guarantee that each source will be available on unchanged terms. The figures cited here are based on company filings and disclosures identified in the October 30 source article.
““through a balanced mix of operating cash flow, investment-grade debt, strategic partnerships, and equity capital.””
— Meta, in a statement reported October 26
The Unsettled Cost and Revenue Balance
It is not yet clear how quickly AI services will generate revenue sufficient to support the scale of new infrastructure spending. The figures establish spending levels, financing transactions and reported lease commitments; they do not establish the eventual return on each data center or how much capacity will be used.
The source raises a question about where risk sits if AI revenue grows more slowly than obligations, but the material does not provide a complete breakdown of guarantees, lease terms, partnership structures or power-contract costs. Nor does it establish that the industry is in a bubble. It identifies financing exposure at newer GPU cloud providers and other entities as a concern, but the specific likelihood or size of losses is not quantified. Future spending plans are company guidance and may change.
Upcoming Spending and Financing Disclosures
The next evidence will come from companies’ subsequent earnings reports and filings, which can show whether capital spending tracks current guidance and how cash flow, debt and lease commitments evolve. Investors and readers can compare actual results with stated plans, including Amazon’s roughly $220 billion 2026 cash-spending outlook, Alphabet’s $195 billion to $205 billion range, Microsoft’s approximately $170 billion expectation and Meta’s $135 billion to $145 billion range.
Further disclosures may clarify how much planned construction is financed directly, through leases or partnerships, and what commitments extend to power suppliers and building owners. Until then, the central issue is not simply how much data-center capacity companies plan to build, but how the bills are divided—and whether future demand can support them.
Key Questions
How much did the four companies spend in the third quarter of 2026?
Amazon, Microsoft, Alphabet and Meta spent a combined about $202 billion on property and equipment from July through September, based on the source author’s calculation from company cash-flow statements. Microsoft’s reported cash figure was $39.4 billion; including new finance leases, its capital spending was $45.2 billion.
Why are technology companies using bonds and loans?
Their data-center and AI investment has grown faster than the cash available after other business needs and capital spending. Debt supplies money now but creates repayment and interest obligations. The financing transactions do not by themselves show whether an investment will earn a sufficient return.
What does the $1.02 trillion lease figure represent?
It is the source’s sum of future payments on leases that had not yet started for the four companies, as of September 30, 2026. It is a reported commitment total, not an amount immediately payable or a measure of likely losses.
Who could bear the risk if AI revenues fall short?
Risk depends on each contract. Borrowers owe debt payments, companies remain bound by lease terms, and shareholders bear the effects of equity issuance and business performance. Partners, building owners, lenders, investors and potentially electricity customers may also be affected, but the source does not quantify their exposure.
Source: ThorstenMeyerAI.com
