− Our rating is **HOLD**, with a 12-month price target of $223.51 against a current price of $284.02 — a deliberately below-consensus stance. At the prevailing price, Amazon is, in our judgment, discounting near-flawless execution of the conversion of its AI capital-expenditure supercycle into free cash flow. We hold conviction at Medium: the business quality is not in question, but the price alread

− | Report Date | 2026-08-04 |

− | Current Price | $284.02 |

− | Rating | HOLD |

− | 12-Month Price Target | $223.51 |

− | Conviction | Medium |

− **§1.6.4 Verdict.** Amazon possesses one of the widest and most durable competitive moats in the market — scale, a subscription flywheel, a lock-in cloud platform, and a data-advantaged ad business that few peers can match. Our reservation is not about the moat's width but about margin durability at the current price: near-term consolidated and AWS margins are being pressured by the AI-capex depre

− | Revenue | 11.7% | 13.2% | —% |

− | EBITDA | 39.1% | 24.8% | —% |

− | Net Income | —% | 29.5% | —% |

− | Diluted EPS | —% | 27.9% | —% |

− | FCF | —% | -21.6% | —% |

− | **Z-Score** | **2.01** | **2.30** | **2.25** |

− | Zone | Gray (1.81–2.99) | Gray (1.81–2.99) | Gray (1.81–2.99) |

− **Interpretation — a gray-zone score that materially understates Amazon's true credit strength.** The Z-Score reads 2.01 in FY2023, 2.30 in FY2024, and 2.25 in FY2025 — the whole span sits in the model's "gray" zone (1.81–2.99), technically below the >2.99 "safe" threshold. Taken at face value that looks cautionary, and per analytical discipline it should not be softened: the score *is* in the gra

− | Risk-Free Rate (10Y UST) | 4.75% | US Treasury daily yield curve |

− | Beta (Levered) | 1.45 | Raw five-year monthly beta, not relevered — Amazon carries no net debt, so there is no leverage adjustment to unwind |

− | **Cost of Equity (Ke)** | **10.88%** | CAPM: Rf + β × (ERP + CRP) |

− At 10.88%, Amazon's cost of equity screens above the broad-market average, and the reason is not the risk-free rate or the equity risk premium — both are middle-of-the-range, Damodaran Tier-1 inputs — but the beta. A raw five-year monthly beta of 1.45 is the single dominant lever in this entire valuation. It is also demonstrably high relative to Amazon's own through-cycle history, which has typica

− | **WACC** | **10.88%** |

− At 100.0% equity and 0.0% debt, Amazon's capital structure is effectively all-equity for valuation purposes. This is not a modelling convenience — it is the economic reality that cash and marketable securities exceed gross debt, so the company carries net cash. On a market-value basis the debt weight collapses to zero, and the WACC therefore equals the cost of equity: 10.88%. The practical consequ

− | WACC | 10.88% | 10.88% | 10.88% |

− *Note: Bear and Bull use the standard adjustments above (per the valuation-agent methodology), applied to every projection year. The resulting fair values appear in Section 6.2 ($193.92 / $256.39).*

− | PV of UFCF ($M) | — | -$21,892.1M | $5,592.1M | $32,014.7M | $50,030.5M | $60,336.1M |

− The defining feature of this projection is that Year-1 unlevered free cash flow is **negative** at -$24,274.7M, and so is its present value at -$21,892.1M. This is not a modelling error — it is the AI-capex supercycle rendered in cash. In the first projection year, capital expenditure modelled at 22.0% of revenue exceeds the operating cash the business throws off, so the franchise consumes cash ev

− | Sum of PV of UFCFs ($M) | $126,081.3M |

− | Terminal Value — Perpetuity Growth ($M) | $1,417,717.5M |

− | PV of Terminal Value ($M) | $2,256,224.5M |

− | **Enterprise Value ($M)** | **$2,382,305.8M** |

− | **Equity Value ($M)** | **$2,436,938.8M** |

− | Shares Outstanding (M — current count used in the per-share bridge) | 10,903 |

− | **DCF Fair Value / Share (Base)** | **$223.51** |

− | Upside / Downside vs. Current Price | -21.3% |

− The two terminal methods do not disagree at the margin — they disagree by a factor of well over two, and understanding why is the whole story. The **perpetuity-growth** method values the terminal cash flows at $1,417,717.5M; the **exit EV/EBITDA** method, at the chosen 14.0x multiple, values them at $3,781,960.0M. The gap is not a rounding artifact — it is the direct consequence of Amazon converti

− We adopt the **exit multiple** as the selected terminal method — $3,781,960.0M, discounted to $2,256,224.5M — and it is this choice that produces the base-case fair value of $223.51, or -21.3% against the current price. The rationale is that a mechanical perpetuity on trough-conversion free cash flow structurally penalizes a business whose capital intensity is cyclical and whose reinvestment is bu

− The price target is sensitive to both the discount rate and the terminal growth rate, and the grid below isolates the **perpetuity-growth** terminal leg — varying WACC and terminal growth affects only the perpetuity terminal value. This is an important labelling point: the bolded central cell of the grid is the perpetuity leg evaluated at base WACC and base terminal growth, **not** the reported ba

− | WACC \\ TGR | 2.0% | 2.5% | 3.0% | 3.5% | 4.0% |

− |---|---|---|---|---|---|

− | 9.9% | 92.0 | 97.5 | 103.8 | 111.0 | 119.5 |

− | 10.4% | 85.7 | 90.5 | 95.8 | 102.0 | 109.1 |

− | **10.9%** | 80.2 | 84.3 | 88.9 | **94.2** | 100.2 |

− | 11.4% | 75.2 | 78.8 | 82.9 | 87.4 | 92.6 |

− | 11.9% | 70.7 | 73.9 | 77.5 | 81.5 | 85.9 |

− The grid geography is as follows: WACC increases **down** the rows and terminal growth increases **across** the columns, so the highest fair values sit in the top-right corner (lowest discount rate, highest terminal growth) and the lowest in the bottom-left. The striking result is that the perpetuity leg does not reach the current share price of $284.02 anywhere in the grid — not even in the most

− The critical point in this table is which return metric to trust. ROE and ROA are contaminated on the Alphabet side by a large, non-operating, largely unrealised equity-securities gain that lifts its reported ROE to 35.7% and ROA to 25.3% — the same class of mark-to-market distortion under current fair-value accounting that has periodically flattered and depressed Amazon's own bottom line. Ranking

− | Interest Coverage | 35.2x | 10.8x | 50.9x | —x |

− | EV/EBITDA | 21.0x | 22.4x | 18.5x | 30.0x |

− | P/E | 39.6x | 46.9x | 27.1x | 34.6x |

− Read with those caveats, Amazon does not look expensive within its cohort. Its EV/EBITDA of 21.0x sits below Alphabet's 30.0x and roughly in line with the group, on an EBITDA that — unlike the software peers — is struck after absorbing a far heavier depreciation load from the fulfilment and data-centre build. The headline P/E comparison is the trap in this table: Amazon's 39.6x looks high on an ab

− | Days Inventory Outstanding | 37 days | 40 days | 5 days | N/A days |

− ## 6.1 Valuation Framework

− Our fair value for Amazon rests on four independent methods, equal-weighted (25% each) within every scenario to produce the composite fair value of $184.29. Each method captures something the others cannot, and for a business in the middle of an AI-and-data-center capital-spending supercycle the disagreement between them is not noise — it is the investment case. We present all four transparently r

− | P/E Relative | 25% | Earnings-based and market-oriented, anchoring to the large-cap technology cohort — but credible only on *normalized* EPS, with the non-cash equity marks stripped out (see 6.3). Applied to reported EPS it is meaningless this year. |

− Our headline price target is anchored on the discounted-cash-flow analysis under an **exit-multiple** terminal treatment. The DCF is deliberately built from operating income (EBIT → NOPAT → unlevered free cash flow), clean of the non-cash Anthropic and Rivian valuation gains that inflate reported earnings (see the RED flag disclosed in 6.3). The scenario range below is the exit-method output.

− | Scenario | Fair Value / Share | Upside / Downside vs. Current Price ($284.02) |

− | Bear | $193.92 | — |

− | Base | $223.51 | -21.3% |

− | Bull | $256.39 | — |

− The critical structural feature of this DCF is that the terminal value does essentially all of the work: discounted explicit-period free cash flow sums to only $126,081.3M, and Year-1 unlevered FCF is *negative* at -$24,274.7M because modeled FY2026 capex exceeds operating cash generation. The present value of the terminal value ($2,256,224.5M) accounts for 95% of enterprise value. When the termin

− That is why the two standard terminal treatments diverge so violently. Under a **perpetuity-growth** treatment (terminal growth 3.5%), the terminal value is $1,417,717.5M — the base cell of the perpetuity grid in 6.5 implies a fair value of roughly two-fifths the exit-method figure, an outright bearish result of roughly two-thirds downside. Under the **exit EV/EBITDA** treatment (14.0x), the termi

− For a franchise of Amazon's quality we regard the exit method as the fair anchor. A perpetuity treatment applied to a company whose near-term unlevered FCF is negative capitalizes trough — indeed negative — cash flow into perpetuity, which is not a credible representation of a durable AWS-and-advertising earnings engine; the implied terminal multiple of roughly 5x is far below where any large-scal

− The enterprise-to-equity bridge adds back the net-cash position of -$54,633.0M (cash and marketable securities net of debt), lifting the $2,382,305.8M enterprise value to $2,436,938.8M of equity value — a rare case where the balance sheet is a tailwind to per-share value rather than a drag.

− Each relative method applies a through-cycle target multiple to the company's metric to derive a fair value per share. The implied fair values below are the base-case outputs that flow into the composite in 6.4 (the three relative methods plus the DCF).

− | P/E | 39.6x | 40.0x | $226.00 |

− | EV/EBITDA | 21.0x | 16.0x | $218.87 |

− | FCF Yield | 0.3% | 3.0% | $133.44 |

− *The current multiple shows where the stock trades today; the target multiple is the analyst's through-cycle anchor; the implied fair value is the target multiple applied to the company's metric. These three relative fair values and the DCF are equal-weighted into the composite (6.4).*

− For a business of this mix — retail marketplace plus a high-margin cloud and advertising engine — **EV/EBITDA is the most reliable relative yardstick**: it is capital-structure-neutral, comparable across the platform peer cohort in Section 5, and less distorted than P/E by the depreciation surge running through this year's income statement. We anchor the target of 16.0x to a modest discount to the

− The P/E line requires a specific and material caveat, and it is where the year's central **RED flag** bears directly on valuation. Reported diluted EPS of $7.17 is substantially inflated by non-cash, Level-3 mark-to-market gains on the Anthropic and Rivian equity investments — gains that swung "other income" from a loss to a very large gain, that far exceed the increase in operating income, and th

− The FCF-yield method is disclosed for completeness but should be read as a floor rather than a fair estimate. This is the second **RED flag** with direct valuation consequence: free cash flow collapsed to $7,695.0M in FY2025 from $32,878.0M a year earlier — a drop of roughly two-thirds — as cash capex nearly doubled to fund the AWS/AI build, with management guiding spending higher again in 2026. T

− | DCF | $133.29 | $158.87 | $188.65 |

− | P/E Relative | $192.10 | $226.00 | $259.90 |

− | EV/EBITDA Relative | $186.79 | $218.87 | $250.95 |

− | FCF Yield | $116.03 | $133.44 | $156.99 |

− | **Composite Fair Value** | **$157.05** | **$184.29** | **$214.12** |

− *Each scenario's composite is the equal-weighted average of the four methods. The Base case is the central reference; Bear and Bull bound the range.*

− **Composite Fair Value (Base):** $184.29

− *Source: Valuation sheet of the workbook (approved assumptions); peer anchors per Section 5; every figure traceable via lineage.json.*

− **Current Price:** $284.02 | **Implied Upside / Downside:** -35.1%

− Two points on how to read this composite against the $223.51 headline target. First, the DCF input into the composite ($158.87) is deliberately the *average* of the perpetuity and exit terminal methods, not the exit-only figure that anchors our target. The composite therefore blends in the bearish perpetuity result and the trough FCF-yield method, and by design sits below the exit-anchored target

− The base-case DCF fair value is most sensitive to two levers: the discount rate (driven by an equity beta that is elevated versus Amazon's own history) and the terminal treatment. The grids below hold everything else constant and flex WACC against terminal growth (perpetuity method) and against the exit multiple (exit method); the central, bolded row is our base-case WACC. The message is unambiguo

− | WACC \\ TGR | 2.0% | 2.5% | 3.0% | 3.5% | 4.0% |

− |---|---|---|---|---|---|

− | 9.9% | 92.0 | 97.5 | 103.8 | 111.0 | 119.5 |

− | 10.4% | 85.7 | 90.5 | 95.8 | 102.0 | 109.1 |

− | **10.9%** | 80.2 | 84.3 | 88.9 | **94.2** | 100.2 |

− | 11.4% | 75.2 | 78.8 | 82.9 | 87.4 | 92.6 |

− | 11.9% | 70.7 | 73.9 | 77.5 | 81.5 | 85.9 |

− | WACC \\ Exit EV/EBITDA | 12.0x | 13.0x | 14.0x | 15.0x | 16.0x |

− |---|---|---|---|---|---|

− | 9.9% | 202.7 | 218.2 | 233.6 | 249.1 | 264.6 |

− | 10.4% | 198.3 | 213.4 | 228.5 | 243.6 | 258.8 |

− | **10.9%** | 194.0 | 208.8 | **223.5** | 238.3 | 253.1 |

− | 11.4% | 189.8 | 204.2 | 218.7 | 233.2 | 247.6 |

− | 11.9% | 185.7 | 199.9 | 214.0 | 228.1 | 242.3 |

− **Rating:** HOLD | **12-Month Price Target:** $223.51 | **Conviction:** Medium

− We rate Amazon **HOLD** with **Medium** conviction and a 12-month price target of $223.51, the exit-method DCF fair value, which we frame as a range spanning the exit-method bear-to-bull scenarios of $193.92 to $256.39 — with the four-method composite ($157.05 to $214.12) sitting lower still as a conservative cross-check. At $284.02, close to all-time highs, Amazon is priced for near-flawless exec

− We stop at HOLD rather than a Sell for three deliberate reasons, and this is the crux of the judgment. First, a mechanical DCF structurally *under*-credits the real-option value embedded in AWS and Amazon's AI positioning — the large contracted AWS backlog and the scale of the committed build are demand signals the model cannot price. Second, the verified equity beta is elevated relative to Amazon

− The bull case requires several things to be true together: AWS revenue growth holds firm *and* its operating margin re-expands as the depreciation of the new capacity is absorbed (reversing the current margin compression), cash capex intensity peaks in 2026 and normalizes without eroding incremental returns, and beta reverts toward historical levels to lower the discount rate. The bear case is sim

− **Key Catalyst:** —

− **Key Risk:** —

− Beyond the headline catalyst and risk, three watch items carry direct valuation weight and are flagged for monitoring: the cash-tax rate is currently depressed by 2025-Act bonus depreciation and will normalize upward as those deductions reverse, pressuring future FCF; the very large block of signed-but-not-commenced leases and unconditional purchase obligations will convert into on-balance-sheet l

− | 1 | Free-cash-flow inflection as capex intensity peaks | FCF yield recovers sustainably above ~2% (vs. 0.3% in FY2025) | FY2026–FY2027 |

− | 2 | AWS operating-margin re-expansion as new-capacity depreciation is absorbed | AWS operating margin rises year-over-year for two consecutive quarters | 12–18 months |

− | 3 | Equity-beta reversion toward long-run levels, lowering the discount rate | Realized beta compresses toward ~1.2, lifting exit-grid fair value toward upper cells | Ongoing |

− ### Downgrade Triggers (conditions that would lower the rating or target)

− | 1 | Capex overrun with no cash-flow recovery | Cash capex stays elevated relative to revenue while free cash flow yield remains negligible | FY2026 |

− | 3 | AWS deceleration or continued margin compression | AWS revenue growth falls below ~15%, or operating margin compresses further | 6–12 months |

− **HOLD**

− | **Action** | HOLD |

− | **Reason** | The operating engine strengthened decisively — AWS operating income rose +63.6% YoY to $16,621M with segment margin expanding to 39.4% (+6.4 pp) — but free cash flow swung to -$8,821.0M on record $54,208.0M of capex, and part of the AWS reacceleration is tied to Anthropic/OpenAI, entities Amazon itself funds. |

− | **Thesis intact?** | PARTIALLY — the AWS-profit-engine thesis is confirmed and arguably reinforced (operating income +43.2% consolidated), but headline net income of $62,647.0M and EPS of $5.75 are dominated by a non-cash Anthropic mark — ~96% of the quarter's non-operating gain, and the large majority of net income — and negative FCF plus circular AI demand are new quality caveats. |

− | **Trigger to revisit** | Upgrade to ADD if FCF inflects positive as the capex cycle peaks; downgrade to REDUCE if AWS growth decelerates below ~20% ex-circular demand or the $92.0B Anthropic gain now parked in AOCI begins to reverse. |

− **Why HOLD:**

− The operating quarter was genuinely strong and thesis-confirming: consolidated operating income rose +43.2% YoY to $27,461.0M with EBIT margin up +2.3 pp, and AWS — the entire bull case — reaccelerated to +37% revenue growth while *expanding* margin 6.4 pp to 39.4%, driving AWS operating income +63.6% to $16,621M (10-Q p. 22). That reverses the AWS-margin-compression worry from the FY2025 review a

− - Upgrade condition: FCF inflecting back to positive on a trailing basis as the capex cycle peaks, *combined with* AWS sustaining ≥25% growth on demand that is demonstrably not Anthropic/OpenAI-circular — would move the action to ADD.

− - Downgrade condition: AWS growth decelerating toward the high-teens, capex continuing to outrun operating cash flow into 2027, or a downward Anthropic mark that begins reversing the $92.0B AOCI gain — would move the action to REDUCE.