Amazon Stock Fundamentals 2026: AMZN Earnings, Cash Flow and Valuation

Amazon Web Services grew 39.3% to $46.0bn in the third quarter of 2026 — its fastest rate in nearly five years and a sixth consecutive quarter of acceleration — and the shares fell around 6% to 7% in extended trading. The reconciliation of those two facts is a single line in the cash flow statement: free cash flow over the trailing twelve months was an outflow of $42.4bn, against an inflow a year earlier.

That is the whole of the current Amazon fundamentals question in one quarter. Earnings and margins are at records — operating income of $28.1bn, up 61%, on a 13.9% operating margin, the highest the company has ever reported. Cash flow is deeply negative because capital expenditure is running at roughly $220bn for 2026 with more promised for 2027. And valuation is now being asked to price two businesses at once: a cloud division accelerating with a $578bn contracted backlog, and a capital programme consuming more cash than the rest of the group produces. The quarter was not a disappointment. It was a demonstration that Amazon has chosen scale over cash, and the market repriced that choice.

What actually happened

Amazon reported third-quarter results after the US close on 29 October 2026. Net sales rose 12% to $201.9bn, from $180.2bn a year earlier, broadly in line with expectations. GAAP diluted earnings per share were $1.89, modestly below a consensus of about $1.96-$1.97 — the rare line in this release that missed.

AWS was the standout. Segment revenue of $46.0bn grew 39.3%, ahead of the roughly $45.3bn expected, and represented the division's fastest growth since 2021 at nearly three times the revenue base it had then. AWS segment operating income rose about 55% to roughly $17.8bn, a segment margin near 38.7%, and ahead of consensus. Chief executive Andy Jassy disclosed a contracted backlog of $578bn and told investors he expects AWS to grow at more than 40% in both 2027 and 2028 — guidance that, if delivered, would mean the acceleration has further to run.

Group profitability set records. Operating income of $28.1bn was up 61% year on year, and the 13.9% operating margin is the highest Amazon has reported. This is the part of the release that gets lost in the share reaction: on an operating basis, Amazon has never been more profitable.

Then the cash. Free cash flow, which Amazon reports on a trailing-twelve-month basis, was an outflow of $42.4bn — the measure having swung from a substantial inflow a year ago and having been negative for a second consecutive period. The driver is capital expenditure. Amazon held its 2026 cash capex guidance at approximately $220bn, the majority of it for AWS, and Jassy said the company expects to spend "meaningfully more in 2027 with a clear line of sight to attractive financial returns". He did not put a number on 2027, which matters: the buyside had been looking for a figure, and its absence left the single largest input to the cash flow model unquantified.

Fourth-quarter guidance was for revenue of $240bn-$247bn and operating income of $28.5bn-$33.5bn. The revenue range sits largely below consensus.

One point of housekeeping on the share price, because it is easy to conflate two separate moves. Amazon closed the 29 October regular session at $243.79, down 4.25% — but that fall happened before the release and belonged to a broader large-cap technology selloff that followed Alphabet's results the previous evening. The reaction to Amazon's own numbers was the extended-session decline that followed, reported in a band from roughly 6% to as much as 12% depending on the source and the moment captured, clustering around 6% to 7%.

Why the market reacted the way it did

The simplest reading is that investors were handed an accelerating revenue line and an unquantified cost line, and priced the uncertainty rather than the acceleration.

Consider what the market already knew going in. AWS growth was expected to be strong. Capex of around $220bn for 2026 had been guided previously. Neither was news. What was new was the combination of Jassy declining to size 2027 capital spending while saying it would be meaningfully higher, and a trailing free cash flow figure that had deteriorated to a $42.4bn outflow. An investor trying to value Amazon on cash flow was left with a known negative and an unknown that is explicitly getting larger.

There is also a sequencing problem. Amazon is spending the cash now and recognising the revenue over years. The $578bn backlog is a genuine asset, but backlog converts to revenue on customer timetables, while data centres, chips and power contracts are paid for on the vendors'. The gap between those two schedules is exactly the free cash flow deficit. Jassy's position — that the returns are visible and attractive — is a statement about the far end of that gap. The market, in the short term, was looking at the near end.

The EPS miss contributed, but it is the least important part. Amazon missing by seven cents on $201.9bn of revenue while posting record operating margin is noise. The cash flow trajectory is signal.

The Openbook read

Growth strengthens, and the quality of it improves. Twelve per cent group revenue growth is unremarkable for the top line, but the mix is what matters: the highest-margin segment is the fastest-growing one and is accelerating, which pulls the group's earnings growth well above its revenue growth. A $578bn backlog converts growth from a forecast into something closer to a schedule. Management guiding AWS above 40% for two further years is the most aggressive forward statement in the release.

Profitability is at a record and is the factor most underappreciated in the reaction. A 13.9% group operating margin at Amazon's revenue scale would have been considered implausible a few years ago, and AWS margins near 38.7% with revenue accelerating demonstrates that the capital being deployed is not being bought at the expense of unit economics. Profitability is not the problem here, which is precisely why the cash figure is so arresting — this is not a company losing money. It is a profitable company out-investing its profits.

Solvency is where the capital programme has to be judged, and it remains comfortable rather than untroubled. Amazon retains substantial operating cash generation, a large cash and securities position and ready access to debt markets. The $42.4bn trailing outflow is funded, not distressed. But solvency is a factor measured against direction as well as level, and the direction is toward more spending with no stated ceiling for 2027. The judgement is not whether Amazon can fund this — it can — but at what point the funding mix shifts materially toward debt.

Momentum has deteriorated, and the sector context matters. Amazon fell 4.25% in the regular session on another company's results before falling further on its own, which tells you the market is treating the hyperscalers as a single AI-capex trade rather than as separate businesses. Operating momentum within AWS is the strongest it has been in five years. Price momentum is working against it.

Reward/risk is the factor that has genuinely shifted, and the shift is a widening rather than a deterioration. The reward case is now concrete and measurable: AWS accelerating through 39% toward a guided 40%-plus, a backlog larger than most companies' market capitalisations, record group margins, and an in-house silicon programme in Trainium that Jassy said would be offered externally late in 2027 — a potential second revenue line out of the same capital base. The risk case is equally concrete. If AI demand decelerates, Amazon will own an enormous fixed-cost base built for a growth rate that did not arrive, and the depreciation attached to $220bn-plus of annual spending runs through the income statement for years whether or not the revenue follows. The asymmetry has not obviously moved in either direction — it has stretched, and investors are being asked to take a view on AI demand durability rather than on Amazon's execution, which has not been in question this quarter.

The read-across

The cloud comparison is the most useful one, and it cuts against the idea that AWS is losing ground. In the same quarter, Microsoft reported Azure growth of 48%, up from 43% in the June quarter, and Alphabet reported Google Cloud revenue roughly doubling to about $30.7bn with a backlog near $561bn. All three accelerated simultaneously. The figures are not strictly comparable — Google's segment includes Workspace subscriptions and TPU system sales, and Microsoft's Azure figure is a product and services measure rather than pure infrastructure — but the direction is unanimous. This is a demand environment lifting every provider at once, not a share war.

That conclusion has an uncomfortable corollary. If all three hyperscalers are accelerating and all three are spending heavily, then the differentiating variable for equity investors is no longer cloud growth. It is capital discipline and the cash flow profile that results. Alphabet fell about 10% after its own results despite doubling cloud revenue, and posted negative free cash flow of its own. Amazon fell on an acceleration. The market is currently penalising the spending more than it is rewarding the growth, uniformly across the group — which investors can compare directly on our screener.

The read-across runs further down the chain. A capital programme of $220bn-plus, rising in 2027, is revenue for semiconductor, networking, power and data-centre construction suppliers. The hyperscalers' cash flow pain is their order book. The risk for that supply chain is concentration: a single large customer deferring a build cycle moves their revenue more than it moves Amazon's.

What to watch next

The fourth-quarter release, expected in late January or early February 2027, is where 2027 capital expenditure should finally be quantified. Amazon has historically given the coming year's capex figure with its fourth-quarter results, and that number is now the single most consequential disclosure for the valuation. "Meaningfully more" than $220bn covers a very wide range.

Alongside it, watch whether AWS growth clears the 40% threshold Jassy guided to. The fourth quarter would be the seventh consecutive quarter of acceleration if it does, and the first test of that two-year commitment.

Watch the backlog figure too. The $578bn stands against Google Cloud's $561bn; the rate of change in each, and the proportion expected to convert within 24 months, says more about near-term revenue than the absolute totals.

Finally, watch the Trainium external launch slated for late 2027. If Amazon can sell its own silicon to third parties, part of the capital base stops being purely a cost centre — and that would change the shape of the free cash flow argument more than any single quarter's spending guidance.