AMZN Amazon.com Inc.
Dive Nov 3, 2026. Price at dive $250.69. Looks cheap Price at the dive was below our base value.
Researched Cases
- Bear
- $211.35
- Base
- $272.42
- Bull
- $338.27
The base case is the forecast believed most likely, one path. Bear and bull are the same path with the top three swing factors each moved by a researched step, all three in the same direction. Everything else is identical across cases.
Key Assumptions, Stated
- Discount rate (WACC)
10.01% - Terminal growth
3.0% - Shares
10.929B - Balance sheet
Net debt $80.2B
Identical in all three cases. Built from SEC filings. Equity risk premium: 3.70% (Damodaran, Oct 8, 2026).
Swing Factors
The three swing factors (KPIs: the few numbers the value hinges on) behind the cases. Each row shows the researched base path and the bear and bull flexes.
| Swing Factor | Base Path | Bear Flex | Bull Flex | Why It Matters |
|---|---|---|---|---|
| Capex intensity, capex as % of sales along the fade | 30.0% of sales in 2027 (~$290B) fading to 9.9% by 2036 | Base path +1.5pts of sales every year (31.5% in 2027 -> 11.4% by 2036) | Base path -1.5pts of sales every year (28.5% in 2027 -> 8.4% by 2036) | The dominant swing in the model: FCF is capex minus a guided peak, so each 1pt of sales on the capex path is worth roughly $20/share; the fade from ~30% of sales to the pre-surge 9-12% band is the single biggest judgment in the memo. |
| AWS revenue growth | +42% in 2027, +41% in 2028, fading to +8% by 2036 | 5% lower each year than base (x0.95), 2027: 42% x 0.95 = 39.9% | 5% higher each year than base (x1.05), 2027: 42% x 1.05 = 44.1% | AWS supplies most of the early growth points (~26% of 2027 sales growing 42% vs ~8% for the rest) and the highest-margin revenue; management guided 40%+ for both 2027 and 2028, so the guide's credibility is the second-largest lever on value. |
| Cash (OCF) margin path | 22.0% of sales in 2027 rising to 26.5% by 2036 | Base path -0.75pt every year (21.25% in 2027 -> 25.75% by 2036) | Base path +0.75pt every year (22.75% in 2027 -> 27.25% by 2036) | FCF is built OCF-first: the margin ramp off the filed 21.1% TTM level, driven by AWS/ads mix, decides how much of the revenue path survives the capex peak; a 0.75pt step is worth roughly $15/share. |
Run Your Own Version
Each workspace has three sliders, one per swing factor. The sliders need JavaScript; the researched values above work without it. Move the sliders past the bear and bull marks, or type any value you believe. The researched cases stay shown above. Your value updates as you go.
See the research
Summary
Base $272 vs price $250.69. The call rests on three findings: (1) management's own guide, AWS growing north of 40% in both 2027 and 2028 off a $578B backlog, verified on the Q3-2026 call, carries total revenue from ~$835B in 2026 to ~$962B in 2027 and ~$1,123B in 2028, reconciling with consensus ($949.7B / $1,101.6B) without borrowing it; (2) the capex surge is a peak, not a plateau, ~$220B in 2026 and 'meaningfully more' in 2027, then fading to 9.9% of sales by 2036, inside Amazon's own 9-12% pre-surge history, so FCF is negative through 2029 and inflects hard after; (3) the discount rate uses Damodaran's actual current headline implied ERP of 3.70%, not the stale 4.20%. What would change the call: AWS growth printing below ~35% in 2027, 2027 capex guided above ~$330B with no fade signal, or evidence the OCF margin cannot climb off its filed 21.1% TTM level.
Assumptions, Stated
| Input | Researched Value | Decision | Evidence |
|---|---|---|---|
| Risk-free rate (FRED DGS10) | 4.91% | kept | FRED DGS10 printed 4.91% on 2026-11-02 (4.94% on Oct 30), corroborated by multiple FRED mirrors; the 10-year touched 5.31% on Oct 5, so the rate environment is high and volatile. Source: https://equibles.com/economicdata/dgs10 |
| Equity risk premium (Damodaran implied ERP) | 3.70% | replaced | Damodaran's own start-of-October 2026 computation (his Oct 8, 2026 post 'Stock Prices, Earnings and Cashflows: The AI Effect', as reported): expected return on stocks 8.99% less the then 10-year Treasury 5.29% = 3.70% implied ERP, that IS his headline implied premium on the trailing-12-month cash-flow (dividends + buybacks) basis, not a lesser variant. The 4.20% figure matches his start-of-January 2026 level (4.23%), it is stale, carried forward and mislabeled as current. The premium fell below 4% for the first time in 2026 precisely because Treasury yields rose ~1.1pts while expected equity returns rose only ~0.6pt. Sources: https://infrastructurecapital.substack.com/p/three-different-yields-one-misleading and https://pages.stern.nyu.edu/~adamodar/New_Home_Page/datafile/implpr.html |
| Beta (multi-vendor) | 1.49 | kept | Managed market-data channel 1.517; StockAnalysis 5Y monthly 1.46 (https://stockanalysis.com/stocks/amzn/statistics/); MarketBeat 1.45; TipRanks 1Y 1.49. Dispersion ~0.07, tight for a multi-vendor beta; no vendor supports a materially different risk profile. |
| Cost of debt (observed issuance) | 5.66% pre-tax; 4.25% after tax at 25% marginal rate | kept | Observed 2026 issuance brackets the synthetic rate: Amazon's Sep 2026 sterling senior notes priced at 5.200% (2029), 5.550% (2032), 6.250% (2038), 6.650% (2045), blended 5.87% (SEC prospectus supplement, https://d18rn0p25nwr6d.cloudfront.net/CIK-0001018724/c81b5278-7074-4f7a-b3a1-0e8c68f978d8.pdf); the Nov 2025 USD deal carried 4.65% at 10 years and 5.45-5.55% at 30-40 years when Treasuries were lower; Mar 2026's $37B USD deal priced its 50-year at +130bp. Amazon sold $92B+ equivalent across currencies in 2026 to fund capex, marginal USD cost at today's 4.91% risk-free rate is consistent with ~5.6-5.9%. |
| Revenue path and AWS growth | Sales $962B 2027 (+15.3%), $1,123B 2028, fading to $2,431B by 2036; AWS +42% 2027, +41% 2028, fading to +8% 2036 | replaced | Q3-2026 verified: total sales $201.9B (+12%), AWS $46.0B (+39.3%, sixth straight acceleration quarter, $184B run rate, backlog $578B up triple-digits). Management on the Oct 29, 2026 call: 'we believe that AWS can grow north of 40% in each of 2027 and in 2028' (https://www.fool.com/earnings/call-transcripts/2026/10/30/amazon-amzn-q3-2026-earnings-call-transcript/). The base path takes the guide at face value (42%/41%, not above it). Cross-check only: S&P Global consensus via Simply Wall St, revenue $949.7B 2027, $1,101.6B 2028 (https://simplywall.st/stocks/pe/retail/bvl-amznus/amazoncom-shares/future), the researched path reconciles within ~1-2%. Non-AWS grows +8% fading to +4% (Q3: North America +6% at a 7.7% margin, International +5%/+8% ex-FX; Q3 growth was understated ~4pts by the Prime Day shift into Q2). |
| Capex path / capex intensity | 30.0% of sales 2027 (~$290B), 28.5% 2028, fading to 9.9% of sales by 2036 (~$240B) | replaced | Guidance verified on the Q3-2026 call: 2026 cash capex ~$220B, and 2027 'meaningfully more', the fade starts from that guided peak, never below it. Street 2027 capex estimates span ~$290B (this base) to ~$355B (Morgan Stanley), so the base sits at the guided end of the range, not the extreme. Terminal landing re-derived from Amazon's own history: capex ran 9-12% of sales in 2021-2023 pre-surge; the path lands at 9.9%, inside that band, at its low-middle, because AI infrastructure is structurally heavier than the old fulfillment build but the surge multiple cannot persist against a $2.4T sales base. |
| Cash (OCF) margin path | 22.0% of sales 2027 rising to 26.5% by 2036 | replaced | Today's level is filed, not assumed: TTM operating cash flow $168.3B on ~$798B TTM sales = 21.1%. Mix arithmetic for the ramp: AWS is 23% of Q3 sales at a 38.7% operating margin and rises to ~32% of sales by 2028 and ~52% by 2036 on the guided path; advertising (~$19.6B/quarter, ~10% of sales, high incremental margin) compounds in the high teens (Q3's +11% carried an 8.5pt Prime Day timing hit; underlying ~20%, Q4 guided high-teens). Shifting ~29pts of sales mix from retail (~12-13% OCF margin) toward AWS (~50% OCF margin incl. D&A add-back) mechanically adds ~+10pts before offsets; haircut for AI/model-resale mix diluting AWS margins (management flagged third-party model mix weighing on AWS margin), Leo expensing, and retail maturity lands the ramp at +5.4pts, to 26.5%, the level the arithmetic supports after offsets, neither the cautious flat-21% nor a 30% stretch. Street cross-check: consensus OCF $248.8B on $949.7B sales for 2027 (26.2%) is more aggressive than this base's 22.0% first year. |
| Terminal growth | 3.0% | adjusted | Re-decided on scale: the terminal year sits on a ~$2.43T revenue base, already the largest revenue pool in the market. Perpetual 3.0% is roughly two-thirds of long-run nominal GDP growth (~4-4.5%: ~2% real + ~2-2.5% inflation), meaning Amazon's share of the economy keeps edging up forever but at a slowing rate; AWS's runway (85% of global IT spend still on-premises, per management) supports at least that, while the sheer base argues against the tool's 3.5% default. Not the cautious number (2.5% is in the sensitivity span) and not the generous one. |
| One-time items (Anthropic marks) | Excluded from run-rate earnings and cash flow; stake excluded from the headline bridge (sensitivity in evidence) | kept | The $69.8B TTM non-operating income is identified: a $16.8B Q1-2026 gain when Anthropic convertible notes converted to preferred stock, plus $53.4B of Q2-2026 pre-tax non-operating income 'primarily from our investments in Anthropic' (Q2-2026 release/10-Q, reported https://beancount.io/blog/2026/07/31/amazon-fy2026-q2-earnings-analysis). It is non-cash and non-recurring, excluded from every forecast path. The stake itself (carrying value reported ~$190B at June 2026, per filing-based analyses) stays OUT of the headline EV-to-equity bridge per bridge discipline; sensitivity: counting it at carrying value would add roughly $15-17/share after tax and illiquidity haircuts, stated here as detail, never in the headline. Anthropic has also committed at least $110B of spend with Amazon through 2036 (per its S-1 reporting), which supports the AWS path independently of the stake's paper value. |
| Share count | 10.929B diluted shares | kept | SEC EDGAR: weighted-average diluted shares for the latest quarter (period ended 2026-09-30) = 10.929B; cover-page basic count 10.813B (2026-10-21). No buyback program is shrinking the count, shares creep up ~1%/yr with stock compensation (SBC $19.9B in the latest fiscal year, treated as a real cost; SBC-as-cash-cost lens in the sensitivity span). No outstanding awards are stacked on top of the filed diluted count. |
| Net debt and lease treatment | Net debt $80.2B | kept | Composed from filings at 2026-09-30: current debt $6.8B + long-term debt $173.1B + finance leases $14.5B = $194.4B, less cash $55.8B and marketable securities $58.4B = $114.2B. Long-term debt tripled from $65.6B at end-2025 to fund the capex surge, already inside this bridge. Operating-lease liabilities stay OUT: rent is paid inside operating cash flow, so adding the lease liability to debt would charge the same rent twice. |
| World context (backlog, ads, competition) | Supportive demand visibility; ads trajectory intact under the timing noise | adjusted | AWS backlog $578B (triple-digit YoY growth) gives the 40%+ guide contracted support; Anthropic's at-least-$110B commitment to Amazon through 2036 (its S-1, reported https://ppc.land/amazon-google-and-microsoft-hold-269bn-of-anthropic-compute-contracts/) anchors a visible slice of that demand. Advertising: +11% in Q3-2026 to $19.64B with an identified 8.5pt Prime Day timing drag, underlying ~20%, Q4 guided back to high-teens (https://ppc.land/amazon-ad-revenue-growth-drops-to-11-after-prime-day-moves-to-june/). Competitive position: AWS is re-accelerating (17% -> 20% -> 24% -> 28% -> 37% -> 39% over six quarters) while funding the build with $92B+ of 2026 bond issuance at investment-grade spreads, the balance sheet, not equity dilution, is carrying the surge. |
What Would Have to Be True
For the $250.69 price to be right on the researched base path, the market must be discounting at ~10.43% instead of the researched 10.01% WACC, or, equivalently at the researched WACC, pricing in perpetual growth of ~2.29% instead of 3.0%. In operating terms: the price is right if AWS undershoots its 40%+ guide, or the capex fade stalls several points above the pre-surge 9-12% of sales, long enough to erase roughly $22/share of present value.
Sensitivity Span (Detail Only)
$179.52 to $399.65. Full variant span, detail only, never the headline range. Low end: stacked extreme (all three KPI flexes adverse plus WACC +0.5pt and terminal growth 2.5%). High end: mirror stack (flexes favorable, WACC -0.5pt, TG 3.5%). Single-variant markers inside the span: WACC +/-0.5pt gives $247-$302; TG +/-0.5pt gives $257-$290; beta across vendors (1.45-1.60) gives $252-$280; keeping the stale 4.20% ERP instead of the current 3.70% gives $238; treating stock-based compensation (~$19.9B/yr filed) as an added cash cost gives ~$241. Terminal value is 78% of enterprise value because FCF is negative in 2027-29, the span is wide for that reason and is flagged, not hidden. The span is detail only, never the headline range.
Cross-Checks (Tool Only)
Tool cross-checks only, the headline is the researched DCF. Tool DCF $32.26 is rejected: its FCF-to-earnings conversion ratio collapsed to 0.12x because trailing FCF is negative at the capex peak, so it mechanically shrinks every forecast year. Tool blend $84.76 inherits that broken DCF at 50% weight. Graham $347.29 runs on GAAP EPS that includes the $69.8B Anthropic mark, so it is inflated as a cross-check. Multiples: no valid peer basket exists for AMZN (tool drops it; never substituted). Street average target ~$330 (Nasdaq $330; MarketBeat consensus ~$321) sits between the base and bull cases.
Smell Tests
- Implied P/E at the $272.42 base value: 22.0x trailing GAAP EPS of $12.37, inside the 12-40x band, but flattered: GAAP EPS contains the one-time Anthropic mark. On normalized EPS (~$7.60, TTM net income $83.0B after removing the $69.8B one-time item) the base value is ~35.8x, still inside the band, near its top, which fits a company whose cash flow is back-loaded by a capex peak.
- Market-implied discount rate on the base path at the $250.69 price is ~10.43% vs the researched 10.01% WACC, a 0.42pt gap, not a chasm: the market is charging modestly more for the capex risk than the researched rate does.
- Implied terminal growth at the price and researched WACC is ~2.29%, below the 3.0% assumed and far below the 5% cap, no terminal-growth heroics are embedded in the price.
- Terminal value is 78.4% of enterprise value, high, and disclosed as the model's main fragility: early FCF is negative (-$77B in 2027) because the forecast starts at the guided capex peak.
Educational only. Not financial advice.