Amazon's core growth story is real: cloud (AWS) growth just hit a 15-quarter high of 28%, and profit margins are expanding in both the US and international retail businesses.
Reported free cash flow looks weak mainly because Amazon is spending roughly $200 billion in 2026 building AI data centers; once that heavy spending is adjusted for, the picture looks much better.
Even so, today's price sits about 27% above our normalized fair-value estimate, requiring better-than-planned cloud growth and margin recovery just to be justified.
Overvalued, but only moderately so, the gap between price and value is modest, not extreme.
Amazon earns money from online retail (in North America and internationally) and from Amazon Web Services (AWS), its cloud-computing division. AWS is the smallest segment by revenue but by far the most profitable; the international retail business is the biggest drag on overall margins.
We'd revisit this call if the price falls below $94.00 or rises above $323.00.
These are review triggers for us to take another look, not price targets or a promise the stock will reach them.
| Metric | AMZN (subject) | Peer median |
|---|---|---|
| EV/Sales | 3.58x | 2.31x |
| EV/EBIT | 32.11x | 31.39x |
| EV/EBITDA | 17.62x | 23.57x |
| P/B | 6.30x | 13.16x |
| Gross Margin | 50.3% | 30.0% |
| ROIC | 16.5% | 20.1% |
| ROE | 18.9% | 22.4% |
| Rule of 40 | N.M. | 9.6% |
Q1 2026 revenue $181.5B, +17% YoY (10-Q, 3/31/26). AWS +28% (fastest in 15 quarters), NA +12%, Intl +19%. Consolidated operating margin expanded to 13.1% from 11.8% YoY, a real, verifiable inflection. Key debate: can the reported $364B+ AWS backlog (+93% YoY) convert to revenue fast enough to justify $200B FY2026 capex before the FCF trough becomes structural, especially as AWS segment margin compressed YoY (37.7% vs 39.5%).
Operating income $23.85B (13.1% margin) vs $18.41B (11.8%), genuine leverage. But net income $30.3B / diluted EPS $2.78 (vs ~$1.64 consensus) was heavily lifted by $15.6B of non-operating 'other income,' mostly non-cash Anthropic preferred-stock mark-to-market gains plus AFS reclassifications, and a $4.1B discrete tax charge tied to those same gains, the GAAP beat is partly a paper-gain artifact, not pure operating outperformance.
TTM (3/31/26) OCF $148.5B; gross capex $151.0B. Company-defined FCF collapsed to $1.2B TTM from $25.9B a year earlier (FY path: $36.8B FY23 → $38.2B FY24 → $11B FY25 → ~$1.2B TTM). Per plan formula (OCF−capex−SBC, SBC TTM $19.8B), FCF_ttm is negative, roughly −$18.6B to −$22.3B. FY2026 capex guide (~$200B, ~56% above FY2025's ~$128B) exceeds plausible OCF, implying FCF stays negative-to-breakeven through at least FY2026. Finance-lease PV also grew (PV $12.3B→$13.3B QoQ), and $106.3B of 'leases not yet commenced' plus $103.8B of unconditional purchase obligations sit outside the reported capex line, the true forward commitment exceeds the headline guide.
Cash+marketable securities $143.1B vs long-term debt $119.1B (net cash ~$24B), but debt nearly doubled in one quarter (+$53.4B new notes plus €14.5B Euro notes) to fund capex. $20B undrawn revolvers, $9.4B unused L/Cs, liquidity ample, not a solvency issue. Buyback capacity is minimal ($6.1B left on a stale 2022 $10B authorization; no repurchases in Q1'25 or Q1'26); dilution negligible (shares 10.731B→10.757B).
AWS margin 37.7% (down ~180bps YoY), still the highest-margin segment but compressing even as growth reaccelerates. NA margin 7.9% (+160bps), Intl 3.6% (+60bps). ROIC ≈ 16% (our research estimate: TTM EBIT×0.79 ÷ [debt+equity−cash]; not company-disclosed). FCF yield ≈ 0.05% (company basis) or negative (plan formula) vs $2.63T mcap. the automated screen g_implied not supplied in this task's snapshot; our research reverse-engineering (Section 6 logic) implies current EV requires ~$85-130B of normalized annual FCF, versus a $1.2B trailing print.
AWS's 28% growth trails Azure (+31%, MSFT FY26 Q1) and Google Cloud (+63%, margin nearly doubled to 32.9%). AMZN is the slowest-growing of the big three by rate, yet trades at ~20% EV/EBITDA premium to peer average (~14-17x vs ~12x) and a high-20s/low-30s forward P/E. The premium reflects AWS scale and backlog durability, not a growth-rate edge.
10Y UST ~4.5% (July 2026), rising Fed-hike odds, pressures AMZN's larger debt stack and premium multiples broadly. Amazon's own risk factors cite memory-chip supply volatility and tariff/trade policy, one macro channel hitting retail COGS (electronics) and AI-server input costs together.
Bezos is executing a pre-set Rule 10b5-1 plan (established ~May 2025), selling up to 25M shares through May 2026; ~95M shares sold since 2024 (~$18.2B), but he retains 900M+ shares (~$220B+). Formula-driven, non-discretionary, small versus total stake, read as routine diversification, not a signal.
Catalysts. AWS backlog conversion (incl. ~$100B/$38B OpenAI compute commitments); continued NA/Intl margin expansion; Amazon Leo/Kuiper commercial launch (targeted mid-2026, 375+ satellites in orbit, FCC deadline relief granted June 2026). Risks. AWS segment margin already compressing despite reacceleration; $106.3B 'leases not yet commenced' plus $103.8B unconditional purchase obligations suggest capital intensity persists past FY2026; circular AI-financing exposure (Amazon has invested/committed up to $50B into OpenAI equity while separately selling OpenAI large AWS compute commitments, and holds a large Anthropic stake whose mark-to-market gains currently inflate GAAP EPS).
(Q6) AWS growth genuinely reaccelerated to a 15-quarter high and beat consensus; NA/Intl margins expanded even in the heaviest capex quarter on record; the balance sheet remains net-cash with ample undrawn liquidity despite the debt-funded capex ramp. (Q7) The overvalued call is falsified toward 'wrong' if TTM company FCF exceeds ~$40B within 12 months while AWS growth stays ≥25%; it is falsified toward 'even more overvalued' if AWS segment margin fails to stabilize above ~39% within 3 quarters.
reassess_low ≈ $106 (0.55× our research Base normalized value of $192), approach toward Bear-case value. reassess_high ≈ $211 (1.10× Base), note the current price ($244.16) is already above this mechanical trigger, reinforcing the overvalued read and arguing for closer monitoring rather than a definitive AVOID.
Mechanical model output, a screen, not a fair value.conservative screen, understates growth/AI-capex names
A capex-normalized bear/base/bull FCF scenario (not the crashed TTM print) puts fair value at roughly $94/$192/$323; the $244.16 price sits ~27% above the Base case and requires above-base AWS-backlog conversion and margin recovery merely to be justified, though it remains below the Bull case, so the mispricing is directional, not extreme.
Research flagged that one or more mechanical-screen inputs look invalid or distorted (e.g. a data-extraction issue or a capex/FCF definition mismatch); the research verdict was independently re-derived from primary filings, not just a check on the same numbers.
Confidence = 0.7 × research sub-score + 0.3 × mechanical sub-score. Mechanical sub-score rewards the Layer 1 screen for agreeing with Layer 2's direction (corroboration), not just for being conviction=NORMAL in isolation.