Amazon Reports Tonight: AWS Up 31% Expected, $207 Billion of Capex Priced, and Microsoft Just Set the Bar

TL;DR

• Amazon reports Q2 2026 after the close on Thursday July 30, the same evening as Apple.
• Consensus: revenue of $196.97 billion, up 18%, with adjusted EPS of $1.82, up about 8%.
• AWS is expected at about $40.5 billion, up 31% from $30.8 billion. KeyBanc, Goldman Sachs and Bank of America all raised their AWS numbers into the print, with BofA going to 33% on Anthropic and OpenAI demand flowing through Bedrock.
• AWS margin consensus is 33.8%, with estimates spanning 30.9% to 38.2%. That 730 basis point range is the widest disagreement on any line in tonight's report.
• The market has already crept its 2026 capex expectation to roughly $207 billion, up about $2 billion since Alphabet reported. So a capex raise is partly priced. What is not priced is a raise without the contracted backlog Microsoft just produced.

What Time Does Amazon Report Earnings?

The short answer: after the close on Thursday, July 30, alongside Apple.

Two megacaps on the same evening splits analyst attention and makes after-hours pricing unreliable in both. It also lands after a morning in which core PCE and the first estimate of Q2 GDP print at 8:30am, which for a long-duration growth stock matters more than most people allow. Our full timetable is in the July 30 what-to-watch guide.

The Board

The revenue line is close to agreed. The margin line is a 730 basis point argument.

What the Street Expects

| Line | Consensus | Comparison | |---|---|---| | Revenue | $196.97B | +18% year over year | | Adjusted EPS | $1.82 | +about 8% | | AWS revenue | ~$40.5B | +31% from $30.8B | | AWS margin | 33.8% | range 30.9% to 38.2% | | 2026 capex expectation | ~$207B | up ~$2B since Alphabet reported |

Note the shape of that table. Revenue growing 18% with EPS growing only 8% is the whole Amazon story in two numbers: the top line is compounding and the spending is eating the difference.

Microsoft Just Changed What Counts as an Answer

This is the most important context for tonight, and it happened 24 hours ago.

For two weeks the bear case on AI has not been about demand. It has been: you are spending tens of billions a quarter against a forecast, not a contract. Alphabet made that case for the bears by posting the first negative free cash flow since its 2004 IPO on $44.9 billion of quarterly capex, and that print is what tipped the Nasdaq 100 into a correction.

Microsoft answered it. Not with a growth rate, with a backlog: commercial remaining performance obligation of $678 billion, up 84%, and up 25% excluding OpenAI, with all sequential growth coming from customers outside the frontier model companies. Then it guided capex higher and the stock rose about 8%. Full detail in the Microsoft breakdown.

So the bar for Amazon is no longer "grow AWS." It is show the contracts.

Amazon does disclose backlog. Tonight, that disclosure is the single most consequential thing in the release, and it will be read against Microsoft's $678 billion whether Amazon likes the comparison or not. AWS growing 31% with a backlog that confirms it is a re-rating. AWS growing 31% with a capex raise and no backlog colour is Alphabet's problem in a different ticker.

The AWS Margin Argument

Consensus 33.8%. Range 30.9% to 38.2%.

A 730 basis point spread on the margin of Amazon's most important segment tells you analysts genuinely do not know what AI infrastructure does to cloud profitability. That is the honest state of the debate.

The context: this consensus has come down from 37.7% expected last October, but up from 32.9% in February 2026. So the street got much more pessimistic, then partly recovered.

Why margin is contested. AI workloads require enormous upfront capital in GPUs and data centres, which depreciates. Depreciation hits AWS margin immediately, while the revenue from that capacity arrives over years. Faster growth funded by heavier capex can therefore lower reported margin while raising long-run value. The market has to decide which one it is looking at, quarter by quarter.

And there is a new cost variable nobody modelled in February: memory prices. Qualcomm guided light this week explicitly blaming rising memory, wafer, packaging and test costs, and Microsoft attributed roughly $25 billion of its calendar-2026 capex to component inflation alone. Amazon buys memory at the same scale. See the Qualcomm breakdown.

AWS margin at 36%-plus with 31% growth is an outstanding quarter and the stock works. At 31% it confirms that AI cloud growth is being bought rather than earned, and the capex raise becomes a problem.

The Bull Case and the Bear Case

Bull case. AWS reaccelerating to 31%, with three major brokers raising estimates into the print and BofA at 33% on Anthropic and OpenAI demand through Bedrock, is a genuine inflection in the segment that produces most of Amazon's operating income. Retail is a cash machine funding it, advertising compounds quietly at high margin, and the market has already moved its capex expectation to $207 billion, so a raise is not the shock it would have been a month ago. If Amazon pairs 31% growth with a credible backlog, it gets Microsoft's reaction rather than Alphabet's.

Bear case. Revenue +18% against EPS +8% is negative operating leverage, and the gap is capex and depreciation. Amazon's capex has a long history of arriving ahead of the returns, and CEO Andy Jassy is widely expected to raise the full-year guide tonight. If that raise lands alongside AWS margin at the low end of the range, Amazon becomes the third megacap in two weeks to tell the market that AI spending is compressing cash conversion, after Alphabet's negative free cash flow and Meta's $784 million of free cash flow on $31.86 billion of operating cash. That pattern is what the correction is about. See the Meta breakdown.

Our read: own Amazon into this on the strength of AWS reacceleration, but expect the reaction to be decided by margin and backlog rather than by revenue. The asymmetry is reasonable: the capex raise is broadly priced, the backlog upside is not. What would genuinely change our view is AWS margin below 32% alongside a capex raise, because that combination says the growth is being purchased at a price nobody has quantified.

The Options Angle

• Elevated implied volatility into a double megacap night means buying options outright pays full event premium, then surrenders it to the vol crush.
• Amazon's post-earnings moves are historically larger than Apple's, so if you want to express a view on tonight in premium, Amazon is the better vehicle. That is an argument about which instrument, not about whether to.
• Cash-secured puts below the current price are the cleanest bullish structure: you get paid to set a lower entry, and you profit from the volatility collapse rather than fighting it.
• Do not hold short-dated premium through both Apple and Amazon on one expiry. Two correlated prints, no chance to manage until Friday, and a core PCE print that morning that can move both regardless of what they report.

The One-Line Read

Amazon is expected to grow revenue 18% with AWS up 31%, and neither number will decide the stock, because Microsoft just reset the bar from growth to contracts by disclosing $678 billion of backlog: tonight is about AWS margin inside a 730 basis point range and whether Amazon can show the commitments behind roughly $207 billion of capex.