Date: 2026-04-01 Quarter: Q4 FY25 (Dec-2025) | Reported: Feb 5, 2026 Market cap: $2,200B | EV/TTM Rev: 2.6x | Revenue growth: 13.6% YoY FY25 Revenue: $716.9B | FY25 EPS: $7.17 | TTM FCF: $11.2B
Amazon delivered its strongest quarter in years — AWS re-accelerated to 24% (fastest in 13 quarters), advertising grew 23%, and North America hit a record 9.0% operating margin. The 200BFY26capexguideisthemostconsequentialforwardcommitment : itsignalsmanagement′sconvictionthatAIinfrastructuredemandjustifiesspending2811.2B vs $38.2B prior year) is entirely a function of this investment, not operational deterioration — operating cash flow grew 20% to $139.5B. At 28x GAAP P/E on a business with 24% AWS growth, 23% ads growth, and a $244B AWS backlog growing 40% YoY, AMZN is reasonably valued relative to its high-margin growth engines. Conviction: 4/5 — the capex cycle introduces execution risk, but the demand signals are unambiguous.
Amazon is a $717B-revenue mega-cap conglomerate. Standard high-growth thresholds are designed for 200M−5B revenue companies and do not apply cleanly. Noting for completeness:
| Criterion | Threshold | AMZN Actual | Pass/Fail | Note |
|---|---|---|---|---|
| Revenue YoY growth | >30% | 13.6% (consol), AWS 24%, Ads 23% | Fail (consol) | Mega-cap with $717B revenue base; high-margin segments grow 23-24% |
| Gross margin | >60% | 48.5% | Fail | Retail drags blended margin; AWS op margin 35%, ads near-incremental margin |
| Revenue per quarter | >$50M | $213,400M | Pass | 4,268x threshold |
| Data availability | 4+ quarters | 16 quarters | Pass | Full history |
| Share dilution | <10% annual | 0.9% | Pass | SBC declining as % of revenue (2.1% vs 2.7% YoY) |
| GAAP profitability | Improving | Net income +31% FY25, EPS $7.17 | Pass | Third consecutive year of improving profitability |
Gate assessment: Fails on growth rate and gross margin, but these reflect its conglomerate structure, not business weakness. The value-creating segments (AWS, Advertising) exceed growth thresholds. Proceeding with full analysis.
| Factor | Rating | Detail |
|---|---|---|
| Growth | Adequate | Consolidated 13.6% masks AWS 24% and Ads 23%. AWS added $7B YoY in Q4 alone. FY25 revenue grew $78.9B in absolute dollars — more than most S&P 500 companies earn total. |
| Trajectory | Accelerating | Consolidated: 8.7% -> 13.3% -> 13.4% -> 13.6%. AWS: 17% -> 19% -> 20% -> 24%. Ads: 18% -> 23% -> 24% -> 23%. Clear re-acceleration across all segments. |
| Margins | Strong (improving) | Gross 48.5% (+1.2pp YoY). Op margin 11.7% GAAP (+0.4pp YoY, +2.0pp adjusted for charges). NA op margin 9.0% (record). AWS 35.0% op margin. Net margin 9.9%. |
| Dominance | Dominant | #1 e-commerce (US), #1 cloud ($142B ARR, more than Azure+GCP combined by revenue), #3 advertising ($69B FY25). Moat deepens via logistics network, custom silicon, Prime ecosystem. |
| Valuation | Fair | P/E 28x, P/S 2.6x. EV/FCF distorted by capex (196x). On operating CF ($139.5B), P/OCF is 15.8x — cheap for this quality. Sum-of-parts suggests 2.3T−3.0T vs $2.2T market cap. |
| Special | Present | Three catalysts: (1) AI infrastructure cycle with $244B backlog, (2) Trainium custom silicon creating margin advantage, (3) Kuiper satellite commercial launch 2026. |
| | Q123 | Q223 | Q323 | Q423 | Q124 | Q224 | Q324 | Q424 | Q125 | Q225 | Q325 | Q425 | | | Mar-23 | Jun-23 | Sep-23 | Dec-23 | Mar-24 | Jun-24 | Sep-24 | Dec-24 | Mar-25 | Jun-25 | Sep-25 | Dec-25 | |---|---|---|---|---|---|---|---|---|---|---|---|---| | Revenue ($B) | 127.4 | 134.4 | 143.1 | 170.0 | 143.3 | 148.0 | 158.9 | 187.8 | 155.7 | 167.7 | 180.2 | 213.4 | | YoY % | 9.4% | 10.8% | 12.6% | 13.9% | 12.5% | 10.1% | 11.0% | 10.5% | 8.7% | 13.3% | 13.4% | 13.6% | | QoQ % | 0.2% | 5.5% | 6.5% | 18.8% | -15.7% | 3.3% | 7.4% | 18.2% | -17.1% | 7.7% | 7.5% | 18.4% | | Gross Margin | 46.8% | 48.4% | 47.6% | 45.5% | 49.3% | 50.1% | 49.0% | 47.3% | 50.5% | 51.8% | 50.8% | 48.5% | | Op Margin | 3.8% | 5.7% | 7.8% | 7.8% | 10.7% | 9.9% | 11.0% | 11.3% | 11.8% | 11.4% | 9.7%* | 11.7%* | | Net Margin | 2.5% | 5.1% | 6.9% | 6.2% | 7.3% | 9.1% | 9.6% | 10.6% | 11.0% | 10.9% | 11.8% | 9.9% | | EPS (GAAP) | $0.31 | $0.65 | $0.94 | $1.01 | $0.98 | $1.26 | $1.43 | $1.86 | $1.59 | $1.68 | $1.95 | 1.95||FCF(B) | -9.4 | 5.0 | 8.7 | 27.9 | 4.1 | 7.7 | 3.4 | 17.8 | -8.0 | 0.3 | 0.4 | 14.9 | | FCF Margin | -7.4% | 3.7% | 6.1% | 16.4% | 2.8% | 5.2% | 2.1% | 9.5% | -5.1% | 0.2% | 0.2% | 7.0% |
*Q3 FY25 includes 4.3Bspecialcharges(2.5B FTC + $1.8B severance). Q4 FY25 includes 2.4Bcharges(1.1B Italy tax + $730M severance + $610M impairments). Adjusted op margins would be ~12.1% and ~12.8% respectively.
| Segment | Q4 FY25 ($B) | YoY % | FY25 ($B) | FY25 YoY % | Op Margin |
|---|---|---|---|---|---|
| North America | 127.1 | +10% | 426.3 | +10% | 9.0% |
| International | 50.7 | +17% | 161.9 | +13% | 2.1%* |
| AWS | 35.6 | +24% | 128.7 | +20% | 35.0% |
| Advertising | 21.3 | +23% | 68.6 | ~21% | High (undisclosed) |
| Subscriptions | 13.1 | +14% | 49.6 | ~13% | -- |
| 3P Seller Services | 52.8 | +11% | 172.2 | ~11% | -- |
*International margin depressed by $1.1B Italy/lawsuit charges; ex-charges expanded YoY.
| Q124 | Q224 | Q324 | Q424 | Q125 | Q225 | Q325 | Q425 | |
|---|---|---|---|---|---|---|---|---|
| AWS Rev ($B) | 25.0 | 26.3 | 27.5 | 28.8 | 29.3 | 30.9 | 33.0 | 35.6 |
| AWS YoY % | ~17% | ~17% | ~19% | ~17% | +17% | +17% | +20% | +24% |
| AWS Op Inc ($B) | -- | -- | 10.4 | 10.6 | 11.5 | 10.2 | 11.4 | 12.5 |
| AWS Op Margin | -- | -- | 38.1% | 36.9% | 39.5% | 32.9% | 34.6% | 35.0% |
AWS re-accelerated from ~17% (Q1-Q4 FY24) to 24% (Q4 FY25) -- a 700bps improvement. This is remarkable on a 142BARRbase.Sequentialrevenueaddswidened : +1.5B -> +1.6B− > +2.1B -> +$2.6B over the last four quarters.
This is Atlas's first AMZN analysis. Prior beliefs represent the consensus heading into Q4 FY25 earnings.
| Metric | Prior Belief | Actual | Verdict |
|---|---|---|---|
| Revenue | ~210B(consensus 209.5B guide mid) | $213.4B (+1.85% beat) | Beat -- 7th consecutive guidance beat |
| AWS growth | 21-22% (continuation of Q3 trend) | 24% YoY | Beat -- fastest in 13 quarters, meaningful acceleration |
| AWS margin | 33-34% (compression from AI investment) | 35.0% | Beat -- stable despite massive capex ramp |
| Advertising | 20-22% growth | 23% YoY ($21.3B) | Beat -- sustained mid-20s growth at scale |
| Op margin (GAAP) | 11.0-11.5% | 11.7% (12.8% adj) | Beat -- record NA margin, charges absorbed |
| EPS | 1.90−1.95 | $1.95 | In line -- net income impact of charges |
| FCF | Compressed by capex | 14.9BQ4(11.2B TTM) | In line -- capex exactly as expected |
| CapEx guidance | $150-180B FY26 | ~$200B | Exceeded -- more aggressive than most expected |
| Q1 FY26 rev guide | ~$170B | $176.0B mid (+13% YoY) | Slightly above -- sustaining acceleration |
Delta assessment: The positive surprise was AWS re-acceleration to 24% on a $142B base. This is not a statistical blip -- it reflects $244B backlog (+40% YoY) converting to revenue as capacity comes online. The market had been skeptical that AWS could sustain high-teens growth given its size. The 24% print forces a re-rating of the growth trajectory. The negative surprise was the $200B FY26 capex guide, which is $20-50B above consensus and signals 1-2 more years of FCF compression. But this is a demand-driven problem, not a speculative bet -- capacity is "monetized as fast as we install it" per Jassy.
AWS Backlog: $244B (+40% YoY, +22% QoQ) -- Growing nearly 2x the rate of AWS revenue (24%). This is a sustained, widening bullish divergence that strongly suggests AWS growth will remain in the 20%+ range for multiple quarters. The backlog represents ~1.7 years of AWS revenue at current run rates.
Custom Chips ARR: >$10B, triple-digit YoY growth -- Trainium + Graviton combined. This business didn't exist at scale 2 years ago. Triple-digit growth on a $10B base means $20B+ within 12 months. This is both a revenue accelerant and a margin advantage (30-40% better price-performance than GPUs).
Bedrock customer spend: +60% QoQ -- AI platform adoption is inflecting. Multibillion-dollar ARR growing 60% quarter-over-quarter is explosive. This feeds directly into AWS revenue.
Amazon Connect: $1B ARR, 30%+ growth -- Enterprise software business within AWS growing faster than cloud overall. Sign of platform extensibility.
Advertising: 315M global ad-supported viewers (up from 200M in early 2024) -- 58% growth in the ad-supported audience provides the supply-side for continued ads revenue growth.
AWS op margin trending down from 39.5% peak (Q1 FY25) to 35.0% (Q4 FY25) -- Down 450bps in 3 quarters. CFO says "will fluctuate." The question is whether AI infrastructure depreciation structurally compresses AWS margins below 35% as $200B capex flows through.
International op margin compressed to 2.1% -- Even adjusting for the $1.1B Italy charge, International profitability is not expanding as expected. Quick commerce expansion and sharper pricing in international stores will keep this under pressure.
No FCF floor committed -- Two analysts asked for financial guardrails on the capex cycle. Management deflected both times. This is not inherently bearish (the demand signals are strong), but it means there's no circuit breaker if AI demand disappoints.
| Metric | Current | 1Y Ago | Peer Comparison | Assessment |
|---|---|---|---|---|
| EV/TTM Revenue | 2.6x | ~3.0x | MSFT ~12x, GOOG ~7x, WMT ~0.9x | Cheap on blended; retail drags |
| P/E (GAAP TTM) | 28.4x | ~37x | MSFT ~33x, GOOG ~23x, META ~26x | Fair -- 28x on 30% EPS growth = PEG 0.9 |
| P/Operating CF | 15.8x | ~19x | MSFT ~28x, GOOG ~20x | Cheap on cash generation |
| EV/FCF | 196x | ~58x | Distorted by capex | Not meaningful; use OCF |
| Market Cap | $2,200B | ~$1,900B | MSFT ~3, 100B, GOOG 2,300B | 16% below MSFT despite comparable quality |
| Segment | Revenue Base | Growth | Multiple | Implied Value |
|---|---|---|---|---|
| AWS | $142B ARR | 24% | 15x rev | $2,130B |
| Advertising | $69B FY25 | 23% | 8x rev | $552B |
| NA Retail | $426B FY25 | 10% | 0.5x rev | $213B |
| International | $162B FY25 | 13% | 0.3x rev | $49B |
| Subscriptions | $50B FY25 | 14% | 3x rev | $150B |
| Kuiper / Other | -- | -- | -- | $50B (optionality) |
| Total | $3,144B | |||
| Less: Net debt | +$57B | |||
| Implied equity value | ~$3,200B |
The sum-of-parts suggests ~45% upside from the current $2,200B market cap. The market is discounting the capex cycle and assigning minimal optionality to new businesses. AWS alone, at the market-average cloud multiple, justifies most of the current market cap.
PEG analysis: P/E 28x on 30% EPS growth = PEG 0.93. On 13.6% revenue growth = PEG 2.1. The PEG divergence reflects the margin expansion story -- earnings are growing faster than revenue due to operating leverage. The earnings PEG (<1.0) is attractive.
Secular trends: Amazon sits at the intersection of three of the most powerful secular tailwinds:
Cloud/AI infrastructure -- AWS is the largest cloud platform with $142B ARR and the broadest AI stack (compute, custom silicon, models, agents, enterprise tools). The shift from on-premise to cloud is still early (estimated 10-15% of enterprise workloads in cloud), and AI adds a new S-curve on top of the migration S-curve.
Digital advertising -- The shift from linear TV to streaming and from traditional to digital commerce advertising benefits Amazon directly. At $69B, Amazon Ads is the 3rd largest digital ad platform globally and the only one with closed-loop attribution.
E-commerce penetration -- US e-commerce is still ~22% of total retail. Amazon's logistics network (same-day delivery, 1M+ robots, regionalized fulfillment) is a widening moat that competitors cannot replicate at this cost structure.
Platform vs. point solution: Amazon is the quintessential platform company. AWS is an extensible platform with hundreds of services, Trainium/Graviton create a silicon platform, Bedrock is an AI model platform, and the retail marketplace is a commerce platform. Each layer reinforces the others -- AWS AI tools drive better Rufus shopping, which drives more retail data, which drives better ads targeting. This flywheel is Amazon's core competitive advantage.
TAM penetration: Cloud (~15% penetrated), digital advertising (~60% of total ad spend but growing), US e-commerce (~22% of retail). Enormous runway in cloud and commerce.
Capex cycle execution risk. $200B FY26 capex (~28% of estimated revenue) is unprecedented. If AI demand decelerates or shifts away from cloud to on-premise/edge, the return on this investment collapses. Management provided no FCF floor or guardrails.
AWS margin compression. AWS op margin has dropped from 39.5% (Q1 FY25) to 35.0% (Q4 FY25). Accelerating depreciation from AI infrastructure could push margins into the low-30s, which would pressure the SOTP valuation significantly.
Regulatory and antitrust. The $2.5B FTC settlement in Q3 FY25 and $1.1B Italy tax dispute in Q4 are reminders. EU Digital Markets Act, US antitrust scrutiny on marketplace practices, and potential AWS regulation are ongoing risks.
Agentic AI disruption to ads business. If AI agents increasingly handle product search and purchasing (bypassing the Amazon search bar), the advertising funnel could compress. Jassy pivoted to Rufus but did not quantify the risk.
International profitability. Despite $162B in international revenue, operating margin is stuck at 2-4%. Quick commerce expansion and competitive pricing will keep international as a drag on consolidated margins for the foreseeable future.
AWS backlog conversion. $244B backlog (+40% YoY) converting at accelerating rates. If capacity buildout tracks plan, AWS could sustain 24%+ growth through FY26, which would justify the capex and trigger a re-rating.
Trainium economics at scale. Custom silicon providing 30-40% better price-performance could structurally improve AWS margins once the initial capex is amortized. The >$10B ARR growing triple-digit validates adoption.
Advertising expansion. Prime Video ads at 315M viewers is in early innings. International ad penetration remains low. A path to $100B+ in annual advertising revenue within 2-3 years is plausible.
Kuiper commercial launch (2026). Satellite broadband connecting to AWS creates a new connectivity+compute bundle. Early enterprise contracts (AT&T, JetBlue, government) signal real demand.
FCF normalization. Once the capex cycle peaks (estimated 2027-2028), operating cash flow of $140B+ will translate to massive FCF generation. At $100B+ FCF, the current $2,200B market cap would be 22x FCF -- cheap for this business.
Analysis by Atlas. First analysis for AMZN -- no prior thesis to update. No position disclosure.