Amazon reports third-quarter results Thursday, Oct. 29, after the close.
At Epochedge, I watched this kind of print from the desk. Readers searching Amazon AWS pénzügyi kilátások 2025 want one signal.
That signal is the AWS backlog. The backlog stands at $496 billion. It is orders waiting to become revenue.
Fast conversion would validate the spending. Slow conversion would raise questions. Bloomberg analysts track this metric closely.
Last quarter, revenue reached $200.61 billion. That is up 19.6 percent year over year. Operating income rose 43 percent to $27.46 billion.
AWS revenue grew 37 percent to $42.23 billion. That was the fastest pace in 18 quarters. The SEC 10-Q filing confirms the figure.
GAAP EPS was $5.75. That total carried a $53.4 billion pre-tax gain from Anthropic. Wall Street Journal flagged that non-core boost.
GAAP means standard accounting rules. EPS means profit per share. You should separate that gain from core earnings.
Q3 guidance calls for net sales of $197.0 billion to $202.0 billion. Guidance points operating income to $22.5 billion to $26.5 billion. The midpoint suggests a softer sequential step.
- Prime Day moved into Q2 this year.
- That pulled retail sales into the prior quarter.
- Q3 retail gets a calendar headwind.
- Foreign exchange is about an 80 bps drag.
- A basis point is one hundredth of a percentage point.
- 80 bps equals about 0.8 percent.
The stock sits at $248. It is up 7.44 percent year to date. Bloomberg shows the recent trading range.
| Metric | Value |
|---|---|
| Annualized Run Rate for AWS | $169 billion |
| Operating Income | $27.46 billion |
| GAAP EPS | $5.75 |
| Q3 Guidance Range | $197.0 billion – $202.0 billion |
| Advertising Revenue | $19.81 billion |
| Q3 Revenue Growth Rate | 19.6 percent |
AWS hit a $169 billion annualized run rate. A run rate is a yearly pace. That is the fifth straight quarter of faster growth.
CEO Andy Jassy said demand will outstrip capacity. Jassy said, “we will still not have enough capacity.” That means 2026 demand remains high.
Capacity is the binding constraint. New data centers must keep growth rising. Chip, power, and land can delay the buildout.
You should watch permits and utilization comments. Management raised 2026 cash capex guidance to about $220 billion. That is up from roughly $200 billion earlier.
Cash capex is cash for long-term assets. The company cited higher memory costs. Memory costs affect AI servers and data center gear.
The Federal Reserve inflation data shows broader cost pressure. A SEC filing discusses capital spending, but not every input cost. The AWS margin expanded 650 basis points year over year.
CFO Brian Olsavsky warned margins can vary with AI mix. You should treat that warning as a real variable. AI workloads come with different hardware and pricing.
Bedrock offers lower-cost model access to developers. That could pressure cloud price levels. The Wall Street Journal noted competitive pricing pressure.
You want management to define pricing and growth. The earnings call transcript will reveal the stance. Advertising grew 26 percent to $19.81 billion.
Thursday Night Football inventory sold out. NBA inventory also sold out. Ad revenue helps because it is high-margin.
Sold-out inventory gives Amazon more pricing power. Bloomberg highlights advertising momentum. Retail still faces cost inflation in logistics.
Amazon expects to more than double its robotic arm fleet in 2026. That could offset fuel and transport cost pressure. You should read robotics as a margin defense.
Faster warehouse automation can reduce labor and transport strain. The guidance should explain the deployment timeline. Operating income is the direct earnings signal.
GAAP EPS can be noisy after large gains. Anthropic moved the recent GAAP number upward. You want core earnings growth and clear accounting.
The backlog conversion rate is the main test. A fast rate would justify higher spending. Trailing free cash flow turned negative at minus $7.6 billion after Q2.
Free cash flow is cash left after real investment. Jassy said returns are visible and strong. He repeated that capital spending will pay off.
Negative cash flow can coexist with strong earnings. Amazon has beaten EPS estimates 11 of the last 12 reports. This quarter, the backlog test matters more.
A beat on EPS would not prove the thesis. You can watch five items on Thursday. AWS revenue shows the demand conversion speed.
- Operating income shows the core margin story.
- Capital spending guidance shows the spending appetite.
- Advertising growth shows the retail engine strength.
- Data center commentary shows the supply limit.
- These five items form the practical checklist.
- You should compare the results against the backlog.
A faster revenue conversion would reduce investor worry. A slower conversion raises a harder question. Does the $496 billion backlog become real revenue?
The plan must earn enough cash to justify itself. The backlog is more than a sales list. It is a pending conversion to revenue.
You should ask when that conversion happens. AI demand can create supply limits before revenue grows. Amazon adds capacity as it sells more capacity.
That balance will appear in AWS revenue. The Bedrock price point matters for cloud margins. Lower-cost models can attract more developers.
They can also soften premium pricing. You want the pricing stance to stay disciplined. The transcript can reveal any wording change.
Small wording changes can move valuation ranges. Prime Day shifting to Q2 is a real detail. It moves promotional retail volume forward.
It also lifts Q2 and pressures Q3. The FX headwind is manageable but visible. It trims growth by about 0.8 percentage points.
You should ask whether operations offset it. The company also faces hardware cost inflation. A Federal Reserve report shows broader price stickiness.
The SEC filing may note capex not input prices. Advertising is a useful margin tool. It grew 26 percent to $19.81 billion.
It can improve margins without new warehouses. Sold-out sports inventory proves demand. It gives Amazon a clearer pricing path.
Bloomberg and WSJ reports support this read. The robotic arm fleet is a logistics answer. More arms can cut handling time.
That can reduce dependence on costly transport. Fuel inflation still squeezes transport cost. Amazon expects to double the fleet in 2026.
The company needs the automation to scale fast. You can read past the GAAP EPS number. Use operating income as the core score.
GAAP can move with gains outside the main business. The Anthropic gain made last quarter look stronger. The core question is whether AWS converts backlog fast.
A clean operating income beat would be more persuasive. Amazon has beaten EPS 11 times in 12 reports. The pattern is good, but the spending test is bigger.
Wall Street will ask about 2026 demand again. Capacity comments should be concrete. If Jassy repeats the constraint, the backlog story stays live.
The stock is up only 7.44 percent year to date. That leaves room for a reaction on the numbers. The guidance range is narrow enough to test management.
You can frame the quarter in two moves. First, see how fast backlog turns into revenue. Second, see if margins hold during AI expansion.
Fast conversion plus stable margins would be constructive. Slow conversion plus lower margins would be a warning. The middle case is the most likely.
That middle case asks for patience. It still requires the backlog to become real revenue. The Q3 report should give one more clue.
A strong backlog can still create cash strain. The spending happens before the revenue arrives. That timing can hurt free cash flow temporarily.
You should read capital spending as investment for future revenue. The investment must land on profitable demand. The Q3 report will test that investment.
Investors often ask about AI monetization. Amazon has the backlog, platform, and sales motion. The answer appears only in quarterly revenue.
The Wall Street Journal notes the capital cycle. Bloomberg consensus expects AWS growth to remain above market. A missed backlog conversion would force a reset.
You can look past the GAAP beat issue. Use operating income as the cleaner lens. Then compare it with the $496 billion backlog.
Q3 guidance implies a slower quarter than Q2. The net sales range centers near $199.5 billion. A top end near $202.0 billion would be good.
You want the guidance to land near the upper half. That would suggest retail demand remained firm. It would also leave room for AWS acceleration.
AWS acceleration is the important test. If capacity comes online fast, revenue should follow. If capacity stays tight, revenue may delay.
