Why Amazon’s Cash Spending Matters As AWS Growth Speeds Up
AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: Why Amazon’s Cash Spending Matters As AWS Growth Speeds Up on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on tech for your team

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

TL;DR

Amazon reported 39% third-quarter AWS revenue growth and a $578 billion contracted-revenue backlog, while spending $75.9 billion in cash on capital investments during the quarter. Trailing 12-month free cash flow fell to negative $42.4 billion, and Amazon shares dropped more than 6% in after-hours trading as investors weighed the cost of its AI and data-center expansion against earnings and forecast shortfalls.

Amazon reported 39% third-quarter revenue growth at AWS, its fastest rate since late 2021, even as it spent $75.9 billion in cash on capital investments during the quarter. Its trailing 12-month free cash flow turned to negative $42.4 billion, and shares fell more than 6% in after-hours trading on Thursday as investors weighed the AI build-out against softer results and spending needs.

AWS revenue reached $46.0 billion for the three months ended in September, while operating income increased 55% to $17.8 billion, according to Amazon’s October 29 results. The source account says AWS growth accelerated for six consecutive quarters, from 17% to 39%, excluding currency effects. CEO Andy Jassy said AWS’s contracted future revenue, or backlog, rose to $578 billion from $496 billion three months earlier.

The spending is running ahead of the cash generated after investment. Amazon reported $168.3 billion in operating cash flow over the past 12 months, up 29%, and capital spending of $210.7 billion, up 82%. The resulting trailing 12-month free cash flow was negative $42.4 billion, compared with positive $14.8 billion a year earlier. The source account separately estimates quarterly free cash flow at about negative $33.5 billion; that is its calculation, not a figure identified there as a company-reported metric.

Amazon executives said most of the planned investment is tied to AWS. Chief financial officer Brian Olsavsky said quarterly cash capital spending “primarily relates to AWS and AI,” while Jassy said the majority of the roughly $220 billion Amazon expects to spend in 2026 will go to AWS. The company expects “meaningfully more” spending in 2027, according to the source account. Amazon’s property and equipment grew from $357.0 billion at the end of 2025 to $508.9 billion at the end of September.

At a glance
reportWhen: Amazon reported third-quarter results o…
The developmentAmazon’s October 29 results paired accelerating AWS growth with heavy capital spending and negative free cash flow, prompting a sharp after-hours share decline.
Thorsten Meyer AI
Results · 29 Oct 2026

Infographic 01 · The scorecard

Fastest cloud growth in five years, deepest cash gap.

Amazon’s third quarter (July to September 2026), scored against the four numbers from my 28 October preview.

+39%AWS revenue, July–September, to $46.0 billion · fastest since Q4 2021
$578BAWS backlog: contracted, not yet delivered · up from $496B in Q2
$75.9Bcash capital spending in one quarter (gross purchases: $77.3B)
−$42.4Bfree cash flow, last twelve months · a year earlier +$14.8B

The four numbers from my preview, scored for Amazon

Status chips: ok = healthy, hold = watch, urgent = warning sign. My assessment.

  • 1 · Capex planAbout $220B for 2026, unchanged holdRaised from $200B in July, citing memory prices. 2027: “meaningfully more”, no number.
  • 2 · Cloud vs. capexAWS +39%, operating income +55% holdGross capex +120% vs. a year earlier ($77.3B vs. $35.1B, my arithmetic). Revenue grows fast; spending faster.
  • 3 · Free cash flow + fundingDeeply negative, financed with bonds urgentAbout −$33.5B in the quarter (my arithmetic). $48.1B of new long-term debt in Q3.
  • 4 · CapacityCustomers take it as fast as it is installed okBacklog +$82B in one quarter. Big blocks reserved well in advance; visibility into 2027–2028.

What beat

  • AWS revenue $45.98B vs. $45.29B expected
  • AWS operating income $17.77B vs. $16.97B expected
  • Operating margin 13.9%, Amazon’s highest ever

What missed

  • Earnings per share $1.89 vs. $1.96 expected
  • Advertising $19.64B vs. $20.82B expected
  • Q4 revenue guide midpoint $243.5B vs. $244.5B expected
  • Shares down more than 6% after hours (CNBC; GeekWire: more than 7%)

Sources: Amazon Q3 2026 release, 29 Oct 2026; Amazon Q3 call, third-party transcript; CNBC, 29 Oct; CNBC, 30 Jul; GeekWire, 29 Oct. Analyst expectations as reported by CNBC (LSEG, StreetAccount). Quarterly free cash flow and capex growth: my arithmetic from Amazon’s cash flow statement. Status chips: Thorsten Meyer.
Powered by Thorsten Meyer AI

The Cost of Building AI Capacity

The results show the financial trade-off in Amazon’s effort to expand cloud and AI infrastructure: demand is translating into orders, but the buildings, chips and equipment must be paid for before all the contracted revenue arrives. The backlog is a measure of future contracted business, not cash already collected, and its value depends on customers taking capacity and services over time.

For investors, the key question is whether AWS’s growth and operating profits will eventually justify the pace of investment. Amazon’s $17.8 billion in AWS operating income shows the cloud unit remains profitable, but company-wide free cash flow reflects capital spending across the business. Amazon shares fell after the report, while the source account also points to earnings per share below analysts’ expectations, advertising results that missed expectations, and a holiday-quarter revenue outlook slightly below forecasts. The share move cannot be attributed to spending alone.

For AWS customers, the build-out may add capacity for AI workloads, but the source account says large blocks of AI capacity are being reserved well in advance and some short-term capacity on older chips is selling at a premium. That suggests availability and cost may differ by product and timing. Amazon has not provided enough detail in the supplied material to quantify the impact for customers broadly.

Products Worth Considering

Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation

Apple AirPods Pro 3 Wireless Earbuds with Active Noise Cancellation

DJI Mic 3 Wireless Microphone System (2 Transmitters, 1 Receiver, Charging Case)

DJI Mic 3 Wireless Microphone System (2 Transmitters, 1 Receiver, Charging Case)

As an affiliate, we earn on qualifying purchases.

How the Spending Accelerated

Amazon’s investment plan has grown during 2026. The source account says the company had planned about $200 billion in capital spending for the year in February, then raised that figure in July, citing higher memory prices. The current expectation is about $220 billion. This is a company forecast, not a final tally, and Amazon has said spending will be meaningfully higher in 2027.

The planned assets include data centers, networking and computing equipment, as well as Amazon’s own chips, Trainium and Graviton. Jassy said Amazon is installing an AI accelerator about every 14 seconds and that commitments for its chips have risen more than 20-fold in a year. Amazon also said customers saved more than $2 billion in the quarter by using its chips; that savings figure is the company’s claim.

Investment affects results beyond the initial cash outlay. Amazon’s depreciation and amortization expense rose 25% to $20.9 billion in the quarter, according to the source account, as the cost of assets is recognized over time. Amazon’s long-term debt also increased from $65.6 billion to $173.1 billion over nine months, though the supplied material does not establish how much of that increase directly funded the capital program.

““Meaningfully more.””

— Andy Jassy, Amazon chief executive

What the Backlog Does Not Show

The $578 billion AWS backlog indicates contracted future revenue, but the supplied material does not specify how quickly it will convert to recognized revenue or cash, or how much is tied specifically to AI services. It is not equivalent to current-quarter sales or a guarantee that every order will produce the same margin.

Amazon has not given a detailed breakdown of its planned spending by data centers, chips, power, memory, or other equipment. Jassy said memory costs were elevated, and Amazon’s guidance cited energy prices and supply volatility, including for memory chips, as risks. The precise effect of those constraints on delivery schedules and spending remains unclear.

The after-hours share decline followed a report with several pressures, including spending, an earnings-per-share miss, advertising results and the holiday-quarter forecast. The available source material does not establish the relative weight investors assigned to each factor. It also does not provide enough detail to determine how much additional borrowing, if any, will be needed to support future investment.

Spending and Capacity to Watch

Investors will be watching Amazon’s next results for updates on AWS growth, operating income and backlog, alongside capital spending and free cash flow. The key test is whether new capacity becomes revenue quickly enough to narrow the cash gap, while operating costs and depreciation continue to rise.

Amazon’s stated plan is to spend about $220 billion in 2026 and meaningfully more in 2027. The timing and details of later investment remain subject to change. Customers and competitors will also be watching data-center capacity, memory and energy supply, and the availability and pricing of AI computing. The source account says Amazon plans to sell Trainium chips outside AWS beginning in late 2027; further details on that plan were not provided.

Key Questions

Why did Amazon shares fall after AWS growth accelerated?

AWS revenue grew 39%, but Amazon also reported negative trailing 12-month free cash flow and $75.9 billion in quarterly cash capital spending. The source account also reports an earnings-per-share miss, advertising results below expectations and a holiday-quarter revenue outlook slightly below forecasts. The decline followed that mix of results; the available information does not isolate one cause.

What does Amazon’s $578 billion AWS backlog mean?

It is the value of contracted future AWS revenue reported by CEO Andy Jassy, up from $496 billion three months earlier. It does not mean Amazon has already received that money, and the supplied material does not detail when the contracts will turn into revenue or cash.

How much does Amazon expect to spend in 2026?

Amazon expects about $220 billion in capital spending in 2026, according to the source account. Jassy said most of it would go to AWS, and the company expects spending in 2027 to be meaningfully higher.

Does negative free cash flow mean AWS is losing money?

No. Amazon reported $17.8 billion in AWS operating income for the third quarter. Negative free cash flow is a broader cash measure affected by operating cash generation and investment spending; it does not mean AWS itself reported an operating loss.

Source: ThorstenMeyerAI.com

You May Also Like

Signal: Four Frontier-Class Open Models in Eight Weeks — China’s Release Cadence Is the Story

Chinese labs released four frontier-class open models in just eight weeks, signaling a rapid production line that challenges Western dominance in AI.

Blackwell Surges In Global Coverage

Blackwell experiences a significant surge in international coverage, with mentions increasing 17-fold in recent monitoring reports, highlighting growing global interest.

Choosing The Most Effective AI Model Over Sovereign Constraints

Analysis of why organizations should prioritize model capability over sovereignty restrictions in AI deployment, based on recent expert insights.

AI’s Signal Power: Preventing A $425 Billion Economic Drain

Google delays Gemini 3.5 Pro, leading to a $425 billion market cap drop amid investor concerns over AI competitiveness.