According to post-earnings coverage, sell-side analysts at multiple firms raised their price targets following Amazon’s Q2 2026 results on July 30, citing AWS’s accelerating growth and a nearly half-trillion-dollar services backlog. The stock jumped over 15% the next day. That reaction is the mood, meaning what investors are currently willing to pay based on sentiment and expectations. The math, meaning what the filed numbers actually justify, is more complicated, and the gap between them is the story.
In one breath
Amazon’s cloud business is growing at its fastest pace in four and a half years, and its profit margins are expanding sharply. IF the multiple were to revert toward its decade median of 41.5 and IF earnings hold near current filed levels, the arithmetic would point to a materially higher price range, though the $220 billion capex commitment is a live reason the discount may persist. The open question is whether a capital spending program larger than most countries’ annual budgets will consume the profits before they compound, or whether it builds a moat wide enough to justify the bill.
How a bookseller became an infrastructure company
Amazon launched AWS in 2006 as an afterthought, a way to rent out spare computing capacity. Within two decades AWS was generating tens of billions in annual revenue at margins that made the retail business look like a rounding error, and investors stopped thinking of Amazon as a retailer. That reframing took fifteen years. The current moment is a second reframing: from cloud provider to AI infrastructure backbone, and it is happening much faster.
The 10-Q filed July 30, 2026, for the quarter ended June 30, shows AWS revenue of $42.2 billion for the quarter alone, up 37% from a year earlier. That is the fastest growth rate in 18 quarters, meaning the business is speeding up, not maturing. AWS now generates roughly 60% of Amazon’s total operating income while representing about a fifth of its revenue. Think of it as a single engine powering most of the plane.
For investors analyzing Amazon’s segment economics, that concentration is central: the retail operation is large, but AWS is where the profit lives.
The retail floor holds, but logistics costs are climbing
Amazon’s North America retail segment nudged its operating margin slightly higher in Q2 2026, per the 10-Q filed July 30, 2026. That is steady, not spectacular. International retail slipped a little as the company absorbs the cost of building out fulfillment in newer markets.
The honest read from the filing is that fulfillment and shipping costs grew faster than online store sales. Automation helps at the margins, but the physical business of moving boxes remains expensive. The retail operation is a volume game with thin margins; it keeps customers inside Amazon’s ecosystem, where the higher-margin advertising and subscription businesses can reach them.
Advertising revenue hit $17.2 billion in Q1 2026, growing at a double-digit clip year-on-year, per Amazon’s Q1 2026 earnings release; gross margin figures for the advertising segment are derived from the segment disclosures in the 10-Q filed for the quarter ended March 31, 2026 (SEC EDGAR). That business barely existed a decade ago. It now quietly subsidizes the cost of cheap shipping.
The capex question that keeps analysts up at night
Here is the tension that the bullish price-target revisions gloss over. Amazon has raised its full-year 2026 capital expenditure guidance to $220 billion, up from a prior forecast, with the increase driven by rising memory chip costs. In Q2 alone, capital spending jumped by more than two-thirds from the same quarter last year. That spending rate exceeds Amazon’s entire annual revenue from just a few years ago.
The consequence shows up immediately: trailing twelve-month free cash flow, meaning cash left over after all investment spending, flipped to a meaningful outflow as of June 30. Management says demand already extends into 2028 and that capacity constraints will persist through 2027. The bet is that building now locks in customers who cannot easily leave once their systems run on Amazon’s custom Trainium and Graviton chips.
Whether that bet pays off depends on whether AI workloads keep growing fast enough to fill the infrastructure being built. The data cannot answer that yet, and confident forecasts in either direction outrun the available evidence.
The regulatory clock is ticking separately
One risk the earnings call did not linger on: the FTC antitrust trial, filed alongside 18 states in September 2023, is scheduled for October 2026. The suit alleges Amazon used its market power to inflate prices and suppress competition. A separate consumer class action certified in August 2025 targets Amazon’s anti-discounting policies for third-party sellers. Neither case has a clear outcome, and the filing discloses both as material risks. Regulatory proceedings of this scale rarely resolve quickly, but they can reshape business practices even before a verdict.
Reading the numbers
- Q2 2026 AWS revenue: $42.2 billion, up 37% year-on-year. AWS is the profit engine. At a 39% operating margin, roughly 39 cents of every dollar AWS earns stays as profit. A household earning $100,000 and keeping $39,000 after all expenses would be considered extraordinarily efficient. This is what makes AWS’s acceleration meaningful rather than just large.
- Full-year 2025 net margin: 10.8% (per the 10-K filed for fiscal year ended December 31, 2025), up from 9.3% in FY2024 and effectively zero in FY2022. Three years ago Amazon earned almost nothing on each dollar of sales. Now it keeps about eleven cents. That trajectory, sustained over three years, is the math the bulls are pricing.
- P/E of 32.7 versus a decade median of 41.5 (BullScope internal estimate, methodology forthcoming, computed from SEC EDGAR filings and FactSet closing prices, January 2016 through June 2026). A multiple is the price tag per dollar of annual profit. Amazon is currently priced at the 12th percentile of its own decade multiple, meaning it has traded at a higher multiple 88% of the time over that period. IF that multiple were to revert toward the decade median of 41.5, and IF trailing earnings hold near filed levels, the arithmetic would point to a materially higher price range. Whether the $220 billion capex overhang prevents that reversion is the open question the data does not yet resolve.
- Trailing twelve-month free cash flow: negative $7.6 billion. Free cash flow is what remains after the company pays for everything it builds. A negative number means Amazon is spending more than it earns in cash right now. A household earning $80,000 a year but spending $82,000 on home renovations is in the same position: not broke, but burning reserves on a bet that the house will be worth more later.
If the capex cycle peaks and AWS revenue continues compounding at even half its current rate, the free cash flow picture reverses sharply. If AI demand disappoints or competition compresses cloud pricing, the bill comes due without the offsetting revenue. Both scenarios are live.
For the standing yardsticks on Amazon: the price tag, the filed record, and the four gauges, refreshed with each edition, see the BullScope Evidence Sheet: Amazon.
Sources
- Amazon 10-Q, quarter ended June 30, 2026 (SEC EDGAR)
- Amazon 10-K, fiscal year ended December 31, 2025 (SEC EDGAR)
- Amazon Q2 2026 earnings release, July 30, 2026
- Q2 2026 earnings call transcript (Investing.com)
- Amazon raises 2026 capex to $220 billion (MLQ.ai)
- FTC antitrust trial set for October 2026 (Inc., citing AP)
- De Coster et al. v. Amazon consumer class action
- Amazon advertising and margin analysis (Forrester)
- AWS vs. Microsoft vs. Google Cloud Q2 2026 (CRN)
