INTRODUCTION TO AMAZON

Amazon is a retail-and-cloud compounder whose profit engine has quietly shifted to AWS. On 28 June 2026, at $232.69 per share, the market is pricing Amazon below our estimate of value, a rare gap for a business of this quality.

The Seven-year-old Explanation

Amazon does three big things. First, it sells almost everything online and ships it to the customer’s door. That is the store most people picture when they hear the name. Second, and far more profitable, it rents out computers over the internet through Amazon Web Services, or AWS. AWS is the invisible backbone that runs a huge share of the world’s apps and websites.

Third, Amazon sells advertising. Brands pay to appear at the top of Amazon’s search results, the way a supermarket charges cereal companies for the shelf at eye level. This third business is small compared to the store but earns almost pure profit because Amazon already owns the customers’ attention.

The way to think about the three engines is simple. The store brings the customers. The ads earn money from that attention. And AWS quietly earns most of the profit. Amazon owns the warehouses, the delivery vans, the data centres, and pays for all the trucks and servers itself. That is why the business is so expensive to run in the short term.

The reason Amazon matters right now is that the world is spending an unprecedented amount of money on artificial intelligence. Training and running AI models requires an enormous amount of computing power, and that computing power runs on the data centres AWS operates. Amazon’s FY26 capital expenditure of roughly $200 billion is aimed at expanding that capacity, and the payoff arrives from FY28 onward as the new capacity earns operating income.

The Investor Explanation

Amazon is a three-segment business headquartered in Seattle, Washington. It reports North America, International, and AWS as its three reportable segments per its FY25 10-K Note 8 Segment Information. North America and International contain the retail, subscription, and third-party marketplace operations, and AWS is the cloud-infrastructure franchise that carries the group’s margin.

Amazon reached about $716.9B of consolidated revenue in fiscal 2025 (fiscal year ending 31 December 2025), with operating income of about $80.0B at an 11.2% blended margin. Andy Jassy, who built AWS from scratch during his 24 years at the company, has been Chief Executive Officer since July 2021. The defining financial feature of this business is capital intensity. Capital expenditure of roughly $131.8B in fiscal 2025 was aimed largely at AWS data centres and logistics infrastructure. That spending depresses near-term free cash flow while building the earning power of the next decade.

PILLAR 1: CIRCLE OF COMPETENCE

Before a company can be valued, one must understand where its revenue comes from, who its competitors are, and what forces shape the economics of its industry. This pillar is the anchor: it defines the circle of competence within which every subsequent pillar operates.

Revenue Mix

Revenue mix is the share of a company’s total revenue that comes from each of its business lines. Understanding it matters because a company’s blended margin and growth rate are the weighted average of its segments. Change the mix and everything downstream shifts, from operating margin to return on capital to the DCF answer.

In Amazon’s case the revenue mix is doing something quiet but important. North America has been around 60% of consolidated revenue for years, International around 22%, and AWS around 18%. Looking only at revenue misses the point. AWS generates more than half of consolidated operating income on less than a fifth of revenue. Every percentage point of the mix that shifts toward AWS or toward advertising inside North America lifts the blended margin measurably.

For context, Microsoft is more than half enterprise-software revenue at very high margin. Google is more than 80% advertising, also very high margin. Walmart is nearly all retail. Amazon sits somewhere in between: retail scale of a Walmart with the cloud and advertising profit engines of a Microsoft or Google layered on top.

The Three Business Segments

North America. ~60% of consolidated revenue. The largest segment. Online store, physical stores, third-party marketplace, Prime subscriptions, and the fast-growing advertising business in the US and Canada. Retail gross margin is thin; the advertising layer is near-pure profit and lifts blended North America margin every year.

International. ~22% of consolidated revenue. Same retail model as North America but outside the US, in Europe, Japan, India, and Brazil. Earlier in its margin curve. Turned sustainably profitable for the first time in Q1 FY26 as the India business crossed its profitability inflection.

AWS. ~18% of revenue but >50% of consolidated operating income. The profit engine. Sells cloud computing (compute, storage, databases, networking, AI infrastructure) worldwide. Q1 FY26 revenue grew 28% YoY at a 37.7% operating margin while absorbing peak AI capex costs.

How Amazon Earns a Dollar of Profit

Not every dollar of Amazon’s revenue is equally profitable. The highest-margin dollar comes from an AWS compute contract, where hardware amortisation is spread across enormous customer volumes and the software layer is essentially fixed cost. The second-highest-margin dollar comes from a sponsored ad placed on an Amazon search results page, because Amazon already owns the customer and the ad slot is nearly zero marginal cost.

The lowest-margin dollar is a retail item shipped from a warehouse. The physical costs of the product, the pick and pack, the delivery, and the returns all subtract from the sale price. This is why the mix shift toward AWS and toward advertising matters so much. Every percentage point of revenue that moves from first-party retail to AWS or to advertising expands the consolidated operating margin.

Market Share

Market share is the percentage of a defined market that a company captures, either by revenue or by unit volume. It matters because market-share direction reveals whether a business is winning against its competition or losing to it.

Amazon’s market-share story is different by segment. In North American e-commerce, Amazon is the dominant platform with roughly 40% share and has held it for years while defending against Walmart, Target, and Costco. In global cloud infrastructure, AWS holds roughly 31% share, ahead of Microsoft Azure at 24% and Google Cloud at 12% per the most recent Synergy Research data. AWS lost some share to Azure in 2023 to 2024 as enterprises consolidated on Microsoft’s AI story, but AWS growth reaccelerated to 28% in Q1 FY26 as the Anthropic capacity ramp kicked in.

In digital advertising, Amazon has grown from a rounding error to the third-largest US advertising platform behind Google and Meta, with roughly $56 billion of ad revenue in FY25 growing more than 20% year-on-year. This is a share-gain story from Google’s search and Meta’s social slots.

The Competitive Landscape

Amazon competes against three main sets of rivals in three specific ways.

Microsoft Azure and Google Cloud. Direct hyperscale rivals. Azure competes hard for enterprise workloads, bundling AI (via OpenAI partnership) with Windows and Office. Google Cloud is aggressive on price and AI tooling. All three are running the same play: build hyperscale AI infrastructure fast enough to capture the next decade of workloads. AWS share: ~31%, Azure ~24%, Google Cloud ~12% (Synergy Research).

Walmart, Target, Costco. Physical-first retail challengers. Walmart is closing the online grocery gap fastest and its scale in fresh food gives it a structural advantage Amazon has not yet matched. Target and Costco defend narrower positions in general merchandise and bulk value.

Google and Meta. Digital advertising incumbents. Amazon Advertising has a decisive structural advantage: it owns retail traffic and purchase intent. When a shopper searches "running shoes" on Amazon, the platform knows they are minutes from buying. Google and Meta must infer intent; Amazon owns it. Amazon is now the #3 US ad platform behind Google and Meta at ~$56B FY25.

Customer Concentration

Customer concentration is the share of a company’s total revenue that comes from a small number of large customers. Investors watch it because concentration is a source of risk: losing one big customer can move quarterly numbers materially.

Amazon’s customer concentration is unusually low across the retail business, because it serves hundreds of millions of individual consumers. In AWS the picture is different. The top ten AWS customers are estimated to account for ~20% of AWS revenue, and the Anthropic capacity commitment of >$100B through 2027 is by itself the single largest customer contract in AWS history. Concentration is rising as the AI franchise scales.

Three Main Customer Relationships

Anthropic. Single largest committed AWS customer. $100B+ multi-year capacity commitment (announced November 2024, expanded through 2025). Anthropic pays cash for AWS compute. Amazon also holds an equity stake in Anthropic.

Meta and Apple. The two other named large AWS commitments. Meta uses AWS for parts of its infrastructure alongside its own data centres. Apple uses AWS as one of three cloud providers under a diversification strategy. Neither is disclosed at contract level.

US federal and state government. Growing AWS customer under GovCloud. Serves classified workloads. Slower-growing but very high-margin franchise with switching costs measured in years.

The Economics of This Industry

The retail-plus-cloud-plus-advertising industry has a distinctive economic profile that matters for how the numbers in Pillar 4, Pillar 5, and Pillar 6 come out.

Cyclicality. Moderate. Cloud is secular growth anchored by AI. Retail is consumer-sensitive: a sharp recession pressures NA and International revenue. Advertising sits between: more secular than retail, not immune to enterprise budget cycles.

Capital intensity. Very high and rising. Amazon owns its data centres, fulfillment centres, and delivery fleet. FY26 CapEx guidance of ~$200B at 24% of revenue is the largest single-year capex commitment in corporate history. Depresses near-term FCF (Pillar 5) while building the earning power Pillar 6 discounts.

Switching costs. Very high in AWS. Once an enterprise builds production workloads and data pipelines on AWS, migrating to another cloud takes months or years and carries operational risk. Anchors sticky recurring revenue and is the foundation of the AWS moat (Pillar 2).

Supplier concentration. Manageable, concentrated in AI accelerators. Amazon buys Nvidia GPUs, designs Trainium AI chips through Annapurna Labs, and buys wafers from TSMC. Nvidia is the largest supplier by dollar spend in FY25 and FY26. Trainium at a $20B run rate is Amazon’s hedge against Nvidia pricing power.

Regulatory environment. Active. FTC antitrust case on the third-party marketplace remains open. European Commission is reviewing cloud bundling. Consumer-protection agencies are reviewing Prime cancellation flows. None fatal on its own; together they define the tail risk in Pillar 10.

Verdict

Amazon’s circle of competence is hyperscaler cloud, scaled North American retail, and a fast-growing advertising engine layered on top. AWS is the durable high-return franchise; retail is thinner but essential to the flywheel; advertising is the mix-uplift that keeps the blended margin climbing. The concentration risk sits in AWS with the Anthropic commitment and a handful of hyperscaler-scale customers.

PILLAR 2: BUSINESS QUALITY AND DURABLE MOAT

Every serious equity investor is trying to answer the same question: does this business have some durable advantage that lets it earn a return above its cost of capital for long enough to matter? This pillar answers that question two ways: quantitatively by comparing ROIC to WACC, and qualitatively through the four moat categories.

Return on Invested Capital versus the Hurdle

The single cleanest test of business quality compares two numbers: the return on invested capital (ROIC) and the weighted average cost of capital (WACC). ROIC is what the business actually earns on every dollar of capital its owners and lenders have put in. WACC is the blended return those investors require.

If ROIC is above WACC, the business is creating value on every dollar it reinvests. Its owners are getting more than the market would give them at similar risk elsewhere. If ROIC is below WACC, the business is destroying value: reinvestment leaves shareholders poorer.

Amazon’s blended ROIC in FY25 was ~14%, following the FY24-FY25 operating margin inflection. Its WACC is 9.05%. That is a spread of ~5pp of value creation on every reinvested dollar. This is the mathematical core of the compounder thesis: as long as Amazon can keep reinvesting at 14% while paying 9% for capital, equity value compounds at the spread.

For context, Microsoft’s ROIC sits near 25% (a 15-point premium to its WACC), Google’s near 22% (a 13-point premium), and Apple’s near 45% (30-point premium). Amazon’s 5-point spread is smaller than these peers but real and improving. Walmart’s ROIC sits near 11% (a 4-point premium to its WACC).

Gross Margin

Gross margin is the share of each sales dollar that remains after paying the direct cost of the product. It is the single cleanest read on pricing power and mix.

Amazon’s consolidated gross margin was ~50% in FY25, up from 44% in FY22, driven by the mix shift toward AWS (very high margin) and advertising (very high margin) as a share of the revenue base. This expansion is not the result of cost cutting but of favourable mix, which is the higher-quality form of margin lift.

For peer context, Microsoft sits near 70% gross margin (pure software), Google near 58% (advertising plus cloud), Walmart near 24% (grocery-heavy retail). Amazon at 50% sits between the pure-software leaders and the pure-retail peers, reflecting its hybrid business mix.

The Four Kinds of Moat, Applied to Amazon

The financial analyst Pat Dorsey codified competitive advantage into four categories, and this framework is a useful lens through which to inspect Amazon’s durability.

Intangible assets. Wide in AWS through the software stack (proprietary services, developer tooling, Trainium custom silicon designs). Wide in Prime through the brand and the customer relationship. Advertising benefits from Amazon’s unique first-party purchase data, which no rival can replicate.

Switching costs. Very high in AWS. Once an enterprise’s production workloads, databases, and data pipelines live on AWS, migrating to another cloud takes months of engineering time and carries real operational risk. This is the deepest source of the AWS moat.

Network effects. Moderate. The third-party marketplace exhibits network effects: more sellers attract more buyers, which attract more sellers. Advertising has a mild network effect (more brands compete for ad slots, raising prices). AWS has weak network effects at the marketplace level.

Cost advantage. Extreme in retail through the fulfillment network built over two decades. Each new fulfillment centre lowers per-unit cost and speeds delivery, widening the gap against Walmart online and against smaller specialty retailers. Cost advantage in AWS comes from data-centre scale and from custom silicon (Trainium and Graviton).

Moat Trajectory

The direction of a moat matters as much as its current width. Amazon’s moat is widening on three fronts. AWS switching costs deepen every year as workloads mature and data-gravity accumulates. The Prime flywheel is compounding through Prime Video ad monetisation and international expansion. Advertising is scaling as a bigger share of the North America mix, and its margin is essentially pure profit.

The one moat that has narrowed slightly is grocery retail: Walmart has closed part of the online grocery gap and Instacart has taken share in some verticals. Amazon’s response is Fresh and Whole Foods integration, which is progressing but not yet decisive.

Verdict

Wide multi-source moat. AWS switching costs, Prime flywheel, logistics density, and the advertising engine all compound. Required MOS for Amazon is 30% for a wide-moat compounder with fortress balance sheet (see Pillar 7 for the requirement in dollars).

PILLAR 3: MANAGEMENT QUALITY AND CAPITAL ALLOCATION

Two things separate a great business from a merely good one over decades: the durability of its competitive advantages, and the quality of the people making the capital decisions. Pillar 2 covered the first. This pillar covers the second.

Operational Track Record

When investors talk about management quality, what they usually mean is track record: the sum of the decisions a CEO has made over years, weighed against what actually happened. Track record is more informative than any prospective narrative because it reveals the CEO’s judgment in the face of uncertainty.

Andy Jassy became Chief Executive Officer of Amazon in July 2021, succeeding founder Jeff Bezos. He had joined Amazon in 1997, one year after graduating from Harvard Business School, and founded AWS in 2003. By the time he became CEO of the parent company, he had already scaled AWS from a technical idea to a $60 billion revenue business at 30-plus percent operating margin. He inherited the CEO role knowing the business intimately, and knowing exactly how the cloud franchise had been built.

For contrast, most large-cap successor CEOs (Satya Nadella at Microsoft, Tim Cook at Apple) also came from inside long-tenured operational roles, and both have delivered compounding returns. External-CEO transitions in mega-cap tech have generally underperformed. Jassy’s internal succession is the base rate for what works.

Four Operational Tests Across Four Years

The track record since Jassy took over breaks into four specific tests, each of which speaks to the durability of the compounder thesis.

Cost programme execution. FY23-FY24 cost programme removed layers of middle management, closed under-performing physical stores, and reduced fixed cost. Consolidated operating margin recovered from 2.4% in FY22 to 11.2% in FY25, the highest in Amazon’s history.

AWS operating margin defence. AWS operating margin held above 33% through the depths of the FY23 cloud spending pullback, when peers saw margin compress. This defended the profit engine while revenue growth temporarily slowed.

Anthropic capacity commitment. The $100B Anthropic multi-year capacity commitment announced in November 2024 anchored the AWS Bull case. Jassy was directly involved in structuring the deal, which combines cash revenue with equity in Anthropic.

Trainium scaling. Trainium, Amazon’s custom AI accelerator built by Annapurna Labs, reached a $20B run rate by Q1 FY26. This is Amazon’s hedge against Nvidia pricing power and its answer to Google’s TPU strategy. Building custom silicon at hyperscaler scale is a decade-long strategic bet that is now paying off.

Capital Allocation

Capital allocation is how management deploys the cash the business generates. Every dollar of cash a business produces has to be spent somewhere: reinvested internally, spent on acquisitions, paid out as dividends, used to repurchase shares, or added to the cash pile. The CEO’s job is to allocate each dollar to its highest-return use.

Amazon reinvests almost everything it earns. Capital expenditure reached roughly $131.8B in FY25 and is guided to ~$200B in FY26, aimed at AWS data centres, AI infrastructure, and fulfillment. Amazon does not pay a dividend and has historically avoided share buybacks. Capital is retained and reinvested at the ROIC the business earns internally, which has been 14% in FY25 versus 9% WACC.

The M&A record under Jassy has been disciplined. The MGM acquisition ($8.5B, closed 2022) added premium content to Prime Video and has scaled advertising monetisation faster than the market expected. The One Medical acquisition ($3.9B, closed 2023) was a foothold in healthcare, still small relative to Amazon’s scale. The iRobot deal was abandoned in 2024 after regulatory pushback and was arguably a bullet dodged.

Insider Ownership and Compensation

Andy Jassy holds roughly 0.05% of Amazon’s shares outstanding, worth roughly $1.3B at the 28 June 2026 spot price. This is small in percentage terms for a founder-scale company because he is not the founder. Jeff Bezos still holds approximately 9% of shares outstanding, worth roughly $225B, and remains executive chair of the board. The founder alignment is unusually deep.

Executive compensation is primarily restricted stock units that vest over multi-year schedules. Jassy’s package is smaller than the CEO packages at peer companies (Microsoft, Google, Meta) but the incentive structure ties his wealth to Amazon’s long-run equity value.

Verdict

Grade-A capital allocator with a long-termist culture inherited from Bezos. Jassy built the largest cloud franchise in the world; that base rate supports the current AI cycle. Founder alignment via Bezos’s 9% stake still deep. The one unmodelled risk is the eventual succession beyond Jassy.

PILLAR 4: FINANCIAL HEALTH AND BALANCE SHEET

The balance sheet reveals whether a business can survive the downside cases under consideration, and whether it can fund the upside cases without diluting shareholders. For Amazon in the middle of the largest capex cycle in corporate history, this pillar answers the question directly.

How Much Debt versus How Much Cash

The starting question for any balance-sheet analysis is simple: how much does the company owe, and how much cash does it hold against that debt? Amazon’s Q1 FY26 10-Q shows $123.0B of cash and short-term investments against $153.0B of total debt. Net financial debt is only about $30B for a company with a $2.5T market capitalisation. In percentage terms, net debt is 1.2% of the equity value.

The professional shorthand for this comparison is net debt to EBITDA. Amazon’s FY25 EBITDA was ~$146B (operating income $80 billion plus D&A $66B). Net debt of $30B divided by $146 billion EBITDA gives a ratio of 0.21 times. In plain words: Amazon could pay off all its net debt with less than three months of operating cash flow.

Amazon has the second-strongest balance sheet in the mega-cap technology peer group, behind Microsoft (net cash of $70B). Google sits similarly to Amazon with modest net debt. Apple holds meaningful net cash. Meta is more leveraged. Walmart carries roughly 1.5 times EBITDA in net debt. Amazon’s position enables the AI cycle without any need for external financing.

Coverage of Interest, and the Debt Maturity Schedule

A related question is whether the company earns enough operating profit to comfortably cover the interest it pays on its debt. Amazon’s FY25 operating income was $80 billion against interest expense of ~$2.4B, giving an interest coverage ratio above 33 times. Interest costs are essentially rounding error against operating earnings.

The debt maturity schedule shows when a company has to repay or refinance the debt it currently owes. Amazon’s debt is laddered across a series of maturities running out to 2062. There is no meaningful maturity wall in the next five years. Any refinancing can be done at rates close to the 10-year US Treasury yield given Amazon’s AA credit rating from S&P (A1 from Moody’s).

Goodwill

When a company buys another company, it usually pays more than the target’s tangible assets are worth on paper. The excess price paid is recorded as goodwill and sits on the balance sheet indefinitely. Under US GAAP, goodwill is not amortised, but it is tested for impairment annually and written down if the acquired business under-performs.

Amazon’s balance sheet carries roughly $23 billion of goodwill, largely from the MGM (2022) and Whole Foods (2017) acquisitions, on total assets of approximately $650B. That is about 3.5% of total assets, a low share for a business this large. No impairment indicators exist in the FY25 10-K.

Verdict

The balance sheet is a fortress relative to the scale of the business. $30B of net financial debt on $2.5T of equity value gives Amazon the flexibility to fund the AI cycle without dilution or external financing. Credit ratings support the position.

PILLAR 5: EARNINGS QUALITY

Not every dollar of reported profit is a real dollar of cash. Companies with weak earnings quality use accounting choices (accruals, capitalisation, deferred revenue) to make reported income look better than the underlying cash generation. This pillar tests whether Amazon’s earnings hold up to a cash-based inspection.

The Gap Between Accounting Profit and Real Cash

In FY25 Amazon reported net income of approximately $77.7B. Over that same period the business generated $139.5B of operating cash flow (OCF). Operating cash flow ran roughly 1.8 times net income, a high-quality signal: real cash generation runs above reported profit, not below.

The gap exists for benign accounting reasons. Depreciation and amortisation of approximately $66B is a non-cash accounting charge: it reduces net income even though no cash actually leaves the business. Stock-based compensation of roughly $19B (2.7% of revenue) is deducted from net income but is a non-cash expense at the OCF level (though it is a real cost to shareholders through dilution, see below).

How Operating Cash Converts into Free Cash

The next layer is free cash flow (FCF): operating cash flow minus capital expenditure. FCF is what a valuation actually discounts, because it is the cash truly available to owners after the business reinvests in itself.

In FY25 Amazon generated $139.5B of OCF against $131.8B of CapEx, giving free cash flow of only $7.7B. This is the central paradox of Amazon’s current earnings profile. A business earning $80 billion of operating income and $77.7B of net income produced only $7.7B of free cash flow, because the AI capex cycle absorbed nearly all of the operating cash.

This is not a signal to sell. It is a signal that the business is investing at maximum capacity for future returns. By FY28 as capex normalises toward 15% of revenue, and by FY30 toward 9%, free cash flow rises sharply. The DCF discounts this forward inflection, not today’s depressed figure.

Stock-based Compensation

One accounting item deserves special attention because it is increasingly the largest hidden cost in technology-company earnings: stock-based compensation (SBC). SBC is pay issued to employees as shares rather than cash. Wall Street traditionally adds SBC back to earnings when computing "adjusted" or "non-GAAP" figures. This is a mistake, and a serious one.

Every share issued to an employee dilutes the ownership stake of every existing shareholder by a tiny fraction. Cumulatively, if a company issues 3% of its shares as SBC every year, existing shareholders own 3% less of the business each year. That is a real cost, and it must be charged.

Amazon spends ~2.7% of revenue on SBC annually, in line with Microsoft (3%) and lower than Meta (5%) or AMD (4.7%). The Compoundex methodology charges SBC in full in the DCF and never adds it back. This is more conservative than the sell-side normalisation.

Verdict

Amazon’s earnings are clean. Real cash generation runs above reported profit through the D&A add-back and working-capital source-of-funds. The FY25 FCF of $7.7B is a temporarily depressed figure driven by the AI capex cycle, and the DCF discounts the forward inflection. SBC is charged in full.

PILLAR 6: VALUATION

This pillar values Amazon three different ways: a full discounted cash flow (DCF) model, a reverse-engineered version of the same model, and an owner-earnings-yield shortcut. When all three methods agree, the analytical case is stronger than any one method alone.

The Headline

My Case intrinsic value stands at $285.31 per share against a market price of $232.69 on 28 June 2026. That is a fair-value margin of safety of positive 18.4%, a 22.6% upside to the market. Base case at $276.80 gives a positive 15.94% MOS. Bear case at $194.48 sits below spot. Bull case at $374.86 gives a positive 37.9% MOS.

The Four Scenarios

Each row is the DCF-implied share price under its own revenue, margin, and discount-rate switches. Fair-value margin of safety is measured against intrinsic value at a market price of $232.69 on 28 June 2026.

The Discount Rate (WACC)

The DCF discounts future cash back to today using a rate that represents what Amazon’s investors demand at similar risk elsewhere. This rate is the Weighted Average Cost of Capital (WACC).

Amazon’s WACC is 9.05%, built from four inputs: a 10-year US Treasury risk-free rate of 4.49%, an equity risk premium of 4.24% (per Damodaran), a levered beta of 1.15 (per Yahoo Finance 5-year regression), and a pre-tax cost of debt of 5.0%. The CAPM formula gives a cost of equity of 9.37%. Blended with an after-tax cost of debt of 3.9% at a 94/6 equity/debt weighting, the WACC lands at 9.05%. Enterprise value under My Case is approximately $3.12T; after adding cash of $123.0B and subtracting debt of $153.0B, equity value is $3.09T, which divided by 10,847 million diluted shares (per the Q1 FY26 10-Q) gives the implied share price of $285.31.

Reverse DCF

The reverse DCF takes today’s price and solves for the growth and margins the market must already be assuming. For Amazon at $232.69, the reverse DCF says today’s price embeds AWS revenue growth around 20% in 2026 declining to 12% by 2030, and consolidated EBIT margin holding near 12% (below the Base case 14% trajectory).

In plain words: the market is not pricing in the Anthropic-driven AWS acceleration or the operating margin lift from advertising scaling. The gap between market-implied and My Case assumptions is the source of the positive margin of safety.

Owner-earnings Yield versus the Graham Hurdle

A shortcut check: divide free cash flow by market capitalisation. Amazon generated $7.7B of free cash flow in FY25 against a market cap of $2,500B. That is an owner-earnings yield of 0.31%. The Graham hurdle, the risk-free rate plus the equity risk premium, is 8.73%. Amazon’s trailing yield falls short by 842bp.

This gap is not a signal to sell. It is the mechanical consequence of the AI capex cycle depressing trailing FCF. Amazon’s forward owner-earnings yield in FY28-FY30, as capex normalises, clears the Graham hurdle in the Base case. The DCF discounts this forward inflection, not today’s trailing figure.

Triangulation

The three methods agree in direction. The DCF My Case says the price is below intrinsic value. The reverse DCF says the price embeds an assumption set well below the conviction view. The forward owner-earnings yield, once capital spending normalises, clears the Graham hurdle. All three converge on a positive margin of safety for the disciplined long-term owner.