AMZN (Amazon): The Everything Company's Moat in the Age of AI?

2026-07-10 · Read on Substack →

AMZN (Amazon): The Everything Company's Moat in the Age of AI?

By Moatery | July 10, 2026

Amazon at ~$245: Up only 12% in the last year. A company that generates $91B in trailing net income, dominates e-commerce (38% US market share), owns the largest cloud computing platform on Earth (AWS at ~$120B+ revenue run rate), and is spending $151B/year building out AI infrastructure. The market is skeptical.

Three years ago, AMZN traded at 60x+ earnings on the pandemic-fueled e-commerce surge. Today it trades at 29x — still a premium to the market, but the business is structurally more profitable than ever before. Operating margins have gone from 2-5% pre-2023 to 11.5% today. The turnaround in profitability is one of the most underappreciated stories in large-cap tech.

Let's run it through all six lenses.

Quick Background

| Metric | TTM (Mar '26) | FY2025 | FY2024 | |---|---|---|---| | Revenue | $742.8B | $716.9B | $638.0B | | Revenue Growth | 14.22% | 12.38% | 10.99% | | Net Income | $90.8B | $77.7B | $59.2B | | Operating Income | $85.4B | $80.0B | $68.6B | | Operating Cash Flow | $148.5B | $139.5B | $115.9B | | Free Cash Flow | -$2.5B | $7.7B | $32.9B | | Operating Margin | 11.50% | 11.16% | 10.75% | | EPS (Diluted) | $8.36 | $7.17 | $5.53 | | CapEx | $151.0B | $131.8B | $83.0B | | Market Cap | ~$2.63T | — | — | | P/E (TTM) | 29.14x | — | — | | EV/EBITDA | 17.32x | — | — | | ROIC | 13.49% | — | — | | ROE | 24.28% | — | — |

Amazon operates three major segments:

1. AWS (~16% of revenue, ~60%+ of operating profit) — Cloud infrastructure, AI/ML services, database

2. North America e-commerce (~60% of revenue) — Online retail, fulfillment, Prime subscriptions

3. International (~24% of revenue) — E-commerce outside North America, typically lower margin or loss-making

The critical recent development: Operating margins went from 2.4% (FY2022) to 11.5% (TTM) — a massive structural improvement driven by cost-cutting, fulfillment network optimization, and higher-margin revenue streams (advertising, AWS, 3P services).

Lens 1: Benjamin Graham (Deep Value / Margin of Safety)

Verdict: ❌ Not Graham's kind of stock. But he'd respect the business.

Graham would look at Amazon the same way he looked at high-growth industrials in the 1950s — interesting business, but not a Graham stock.

The Graham Calculation:

• Net Current Asset Value (NCAV): Total current assets of $255.2B minus total liabilities of $474.7B = negative $219.5B. Net-net fails spectacularly.

• Book Value: $441.9B total equity → ~$41.09/share. At $245, P/B = ~5.9x.

• Graham Number: √(22.5 × EPS × BVPS) = √(22.5 × $8.36 × $41.09) = ~$88/share.

The Graham Number suggests AMZN is worth $88/share, meaning it trades at ~2.8x Graham's estimate. But Graham's formula was designed for industrial and manufacturing businesses with tangible assets. Amazon's value is in its fulfillment infrastructure (PP&E of $486B), its AWS platform, its brand, and its data.

The Graham Perspective: Graham would see a company with $486B in physical assets (warehouses, data centers, logistics), $143B in cash, and $91B in earnings. He'd note that the P/E of 29x is not deep value by any measure. But he'd also acknowledge that $486B in hard assets backing a $2.63T market cap means there's some floor, even if the intangible premium is large.

Graham's Bottom Line: "This fails my formulas by a wide margin. The price-to-earnings ratio of 29x, the price-to-book of 5.9x, and the negative net current asset value make this a speculative investment by my standards. I would not buy. However, I acknowledge this is a business with enormous tangible assets — $486B in property and equipment is real. If you must own growth stocks, at least this one has hard assets behind it."

Grade: ❌ Not deep value — a business built on intangibles and growth expectations, not liquidation value.

Lens 2: Warren Buffett (Quality Moat / Durable Competitive Advantage)

Verdict: ✅ Yes — multiple reinforcing moats, finally at a sensible price.

Buffett's four-question framework:

1. Can we understand the business 10 years out? Yes — arguably better understood than any company in the world:

• Amazon will still be the default for e-commerce. The flywheel (selection → traffic → sellers → lower prices → more traffic) is self-reinforcing. No competitor has disrupted it in 30 years. Walmart has tried spending billions. Shopify enables D2C but doesn't capture the logistics moat.

• AWS will still dominate cloud. The switching costs are enormous — once your infrastructure runs on AWS services (S3, EC2, Lambda, RDS), migration costs are prohibitive. AWS has 31% cloud market share vs #2 Azure at 24%.

• Advertising will keep growing. Amazon Ads is now the #3 digital ad platform behind Google and Meta. $50B+ in annual ad revenue growing 20%+. It's an entirely new profit pool that didn't exist 5 years ago.

2. Does it have a durable competitive advantage (moat)? Yes — warehouse logistics is one of the hardest moats to replicate:

Moat Scorecard: | Moat Type | Score | Evidence | |---|---|---| | Cost Advantage | 3/3 | Lowest-cost logistics infrastructure; nobody can match 2,000+ fulfillment centers globally | | Switching Costs | 3/3 | AWS service lock-in + Seller ecosystem lock-in (FBA inventory, reviews, rankings) | | Network Effects | 3/3 | Marketplace flywheel: more buyers → more sellers → better selection/prices → more buyers | | Intangible Assets | 3/3 | The most trusted e-commerce brand, 200M+ Prime subscribers, 30 years of logistics IP | | Efficient Scale | 3/3 | Winner-take-most in e-commerce; natural oligopoly in cloud (AWS/Azure/GCP) | | Total | 15/15 | Perfect moat score — the only company I've scored this across all five categories |

3. Does the business generate high returns on tangible capital? ✅ ROIC of 13.49%. This is decent but not elite. The heavy CapEx ($151B/year) depresses ROIC. However, ROIC has been improving fast (was under 10% in FY2023-2024). As AWS margins grow and fulfillment density improves, ROIC should trend toward 20%+.

The key insight: Amazon's CapEx creates assets that last. Fulfillment centers have 20-30 year useful lives. Data centers last 15-20 years. The depreciation is accelerating, but the cash flows from these assets are compounding.

4. Can a fool run this business? Evidence for YES: The flywheel is institutionalized. AWS momentum is self-sustaining. Fulfillment is a machine. The culture of "Day 1" is embedded. Andy Jassy (CEO since 2021) has proven himself — margins are up 4x since he took over.

Evidence for UNCERTAINTY: Bezos is irreplaceable for long-term vision. Jassy is operational, not visionary. The AI strategy is fragmented — Amazon has its own chips (Trainium/Inferentia), an Anthropic investment, and Bedrock. No single coherent AI narrative vs Microsoft/OpenAI. But Amazon has historically won through execution, not narrative.

Buffett's Bottom Line: "For years I said Amazon was a wonderful business but the stock was too expensive for my taste. At 29x earnings — with $91B in real profits, a 15/15 moat score, and margins quadrupling in 3 years — it's finally in the conversation. The operating business is printing $150B+ in cash annually. The CapEx is staggering but creates assets that generate real returns. This is the kind of business I wish I could have bought at 15x. At 29x, it's a good business at a fair price — not a bargain, but not crazy either."

Grade: ✅ BUY — Widest moat in retail history, margins finally reflecting the structural advantage.

Lens 3: Peter Lynch (Growth at a Reasonable Price / Story Stock)

Verdict: ✅ Interesting GARP candidate. Not cheap, but growing into the multiple.

Lynch's categories for Amazon:

• Fast Grower: 14% revenue growth, 38% earnings growth. At $743B revenue, it's a stalwart pretending to be a fast grower.

• Turnaround: The 2022-2023 margin explosion is a turnaround within a growth company. Operating margins went from 2.4% to 11.5% — that's turnaround territory.

Lynch's Checklist:

• PEG Ratio: PE of 29.14 ÷ EPS growth of ~37% = PEG of ~0.79x. This is well within Lynch's <1.0x buy zone for fast growers. Even using 3-year forward earnings growth of 20%: PEG = 1.46 — still reasonable.

• P/E to Growth: 29.14 ÷ 20 (3yr consensus) = 1.46x. Right at the boundary.

• Inventory: Turnover of 10.14x is excellent. Amazon manages inventory better than almost any retailer.

• Debt/Equity: 0.53. Very manageable for a company with $148.5B in operating cash flow.

• Free Cash Flow: -$2.5B FCF looks scary, but this is all CapEx. Operating cash flow of $148.5B covers all needs. The FCF drag is the AI/fulfillment buildout — a temporary phenomenon.

• Is there a "story"? Yes — the AI infrastructure buildout, the margin transformation, advertising becoming a $50B+ business, international profitability turnaround. Multiple stories, all positive.

Growth Drivers:

1. AWS AI: The AI workload shift to the cloud is AWS's biggest growth opportunity since its founding. Amazon Bedrock, SageMaker, and custom Trainium chips are capturing AI inference workloads.

2. Advertising: $50B+ run rate, growing 20%+ annually. This is high-margin (nearly pure profit) revenue that barely existed 5 years ago.

3. International margins: Amazon International has historically lost money. In FY2025, it's approaching breakeven. As developing markets mature, this becomes a multi-billion dollar profit stream.

4. Fulfillment optimization: Regionalization (8x8 network vs national) dramatically lowers shipping costs. This margin expansion has room to run.

Lynch's Bottom Line: "29x earnings on a business growing 14% top-line and 38% bottom-line. The PEG is 0.79 — well within my buy zone. The margin story is real — they cut $10B in costs and it stuck. Advertising is a hidden profit pool. International is a future profit pool. This is a GARP opportunity disguised as a mature company. I'd buy."

Grade: ✅ GARP BUY — PEG 0.79, accelerating margins, hidden profit pools emerging.

Lens 4: Li Lu (Long-Arc Quality / Enduring Value Creation)

Verdict: ✅ Yes — a long-arc compounder with huge optionality.

1. Enduring competitive advantage? Yes, possibly the most enduring in all of business. Amazon's flywheel is not just a moat — it compounds on itself:

• More selection → more customers → more sellers → lower prices → more customers

• More AWS customers → more services → higher switching costs → more customers

• More Prime members → more engagement → more data → better recommendations → more sales

Crucially: The moat is widening, not shrinking. 20 years ago, Walmart could compete with Amazon on price. Today, Amazon's fulfillment network (same-day delivery in major metros) is a moat Walmart can't match despite spending billions.

2. Superior capital allocation? Amazon has one of the best capital allocation records in corporate history:

• AWS investment (2006-present): Arguably the best organic investment in tech history. From a $0 startup to a $120B+ revenue business with 30%+ margins.

• Fulfillment network (2010-present): $150B+ invested in logistics. Created a moat that no competitor has replicated.

• Advertising pivot (2016-present): Recognized an adjacent profit pool and captured it. Now a $50B+ business.

• Anthropic investment ($4B): Backing a leading AI lab to compete with OpenAI/Google. Early, but strategically sound.

• Kiva Robotics acquisition ($775M, 2012): One of the best acquisitions in tech — transformed warehouse efficiency.

The only criticism: Amazon has never returned capital to shareholders (no dividends, no buybacks). This is a feature, not a bug. Bezos/Jassy believe they can generate >15% returns on retained capital. The historical record supports this.

3. Management with aligned incentives? ✅ Andy Jassy owns significant Amazon stock. Bezos is still Executive Chairman and owns ~9% of the company. The "Day 1" culture is deeply embedded. Capital allocation decisions (invest everything) are aligned with long-term shareholder value, even if they frustrate short-term investors.

Li Lu's Bottom Line: "This is the quintessential long-arc compounder. The competitive advantage is structural, self-reinforcing, and global. Capital allocation has been near-perfect for 25+ years. Management is deeply aligned with shareholders. The moat is widening. At 29x earnings, this is not a discount, but for a business of this quality and durability, 29x is a reasonable entry point for a 10+ year hold. If they ever return capital to shareholders (dividends, buybacks), the re-rating could be significant."

Grade: ✅ LONG-ARC BUY — Elite business quality, reasonable (not cheap) entry, 10-year compounding trajectory.

Lens 5: Howard Marks (Cycle / Psychology / Risk)

Verdict: ⚠️ Neutral to positive — sentiment is mixed, no extreme pessimism.

Cycle Position:

• Sentiment: Mixed. Up 12% in a year — not euphoric, not despairing. The stock has been range-bound between $210-265 for 18 months. Neither bulls nor bears are dominant. This is a "show me" market — investors want to see the AI CapEx translate into revenue.

• Earnings Cycle: Strong and improving. Margins are at all-time highs. Revenue growth is accelerating (14% vs 11% last year). The business has never been stronger.

• Capital Cycle: Massive investment phase — $151B in CapEx. Investors are right to ask: will these investments earn a decent return? Amazon's track record says yes, but the magnitude is unprecedented.

Risk/Reward Asymmetry:

• Upside: Multiple expansion to 35x + 20% EPS growth = ~$350-380 in 12-18 months. That's 43-55% upside. Analyst target is $313.

• Base case: 29x + 15% EPS growth = ~$275 in 12 months. 12% upside.

• Downside: Multiple compression to 22x + CapEx disappointment + competition = ~$185. 24% downside.

Asymmetry ratio: ~2:1 (upside 43% / downside 24%)

This is less asymmetric than MSFT (3.5:1) or the best Marks plays. The market isn't pricing in excessive pessimism — it's pricing in reasonable caution.

Key risks:

• Regulatory risk: The FTC has been aggressive under the current administration. An antitrust breakup is unlikely but possible. Separation of AWS from retail would destroy value. European regulations (Digital Markets Act) are a growing compliance burden.

• CapEx overshoot: $151B in annual CapEx is more than all of Amazon's competitors combined. If AI demand slows, Amazon could have excess capacity. But data centers have long lives and can serve general cloud demand.

• Competition in cloud: Microsoft Azure is gaining share. Google Cloud is a credible #3. The cloud market is still growing 20%+ annually — enough room for multiple winners.

• International losses: International segments are still losing money. Emerging market e-commerce (India, Brazil) is capital-intensive and low-margin. These bets could take a decade to pay off.

• Labor risk: 1.58M employees. Unionization efforts are growing. Wage inflation in fulfillment centers is real.

Marks' Bottom Line: "The cycle is not screaming 'buy' with the same conviction as some other names. Sentiment is cautious but not pessimistic. The asymmetry is ~2:1 — good, but not exceptional. The real opportunity in Amazon is structural, not cyclical. The business quality is elite. The stock valuation is fair. I'd buy more on any meaningful pullback — say, below $220."

Grade: ⚠️ ACCUMULATE — Fair price for an elite business. Wait for a better entry or buy for the long term.

Lens 6: Carl Icahn (Activist / Catalyst / Asset Value)

Verdict: ⚠️ Sum-of-parts reveals hidden value, but Amazon doesn't listen to activists.

Sum-of-Parts Valuation:

| Segment | Est. Revenue | Multiple | Value | Basis | |---|---|---|---|---| | AWS | ~$120B | 10x sales | $1.20T | Cloud leader, 30%+ margins, 20% growth | | North America Retail | ~$445B | 1.5x sales | $668B | Low margin (~4%), high volume, advertising embedded | | International Retail | ~$178B | 1.0x sales | $178B | Approaching breakeven, turnaround optionality | | Advertising | ~$50B | 15x sales | $750B | High-margin, 20%+ growth, #3 ad platform | | Prime Subscriptions | ~$35B | 5x sales | $175B | Recurring revenue, 200M+ members | | Sum of Parts (excluding advertising overlap) | | | ~$2.50-2.80T | |

Current market cap: ~$2.63T.

Sum of parts is roughly inline with the stock price. But the sum-of-parts analysis is conservative:

• AWS alone might be worth $1.5-2.0T if you apply a software multiple (20-25x earnings) rather than a revenue multiple

• Advertising is probably worth more than 15x sales given it's the #3 ad platform globally

• The sum-of-parts doesn't capture the flywheel synergy — these businesses reinforce each other

What an activist might push for:

• Break up AWS: This is the perennial activist dream. An independent AWS would trade at 25-30x earnings vs 29x consolidated. A spin could unlock $500B-1T in value. But Jeff Bezos sits on the board and has consistently said no.

• Return capital: Amazon pays no dividend and does no buybacks. An activist would argue: $148.5B in operating cash flow means you can invest AND return capital. Initiate a $20-30B buyback program.

• Stop the International bleeding: Close or restructure unprofitable international operations. Focus on NA + AWS + advertising. This would add $5-10B in annual profit.

• Slow the CapEx growth: $151B is a lot. An activist would ask: are you sure all of this is earning 10%+ returns? Slow down, wait for demand visibility.

The reality: Amazon is essentially activist-proof. Bezos controls ~9% of shares with outsized voting power. Jassy is aligned with the long-term vision. Short of a massive share price decline and internal revolt, Amazon will continue on its current path.

Icahn's Bottom Line: "The sum of parts is a rounding error away from market cap — but the components are undervalued. AWS alone could be worth $1.5T in a bull case. Advertising is a free option. A break-up would unlock billions, but this management team has no interest. Without a catalyst, it's a hold. If they ever announce a spin-off or buybacks, it's an immediate buy. Until then, the hidden value stays hidden. "

Grade: ⚠️ HOLD (passive) / BUY (if catalyst emerges) — Sum of parts validates current price; break-up or buybacks would unlock significant value.

The Moatery Synthesis

| Lens | Verdict | Key Reason | |---|---|---| | Graham | ❌ Not Value | 29x earnings, 5.9x book — fails every formula | | Buffett | ✅ Buy | 15/15 moat score, margins expanding, fair price | | Lynch | ✅ GARP Buy | PEG 0.79, accelerating earnings, hidden profit pools | | Li Lu | ✅ Long-Arc Buy | Elite business quality, 10-year compounder at reasonable price | | Marks | ⚠️ Accumulate | Fair price for elite business; wait for pullback | | Icahn | ⚠️ Hold | Sum of parts = fair value; catalytic event needed for unlock |

The Bottom Line

Amazon at $245 is a fair price for one of the best businesses in the world.

The bull case: margins are structurally higher than ever (11.5% operating margin vs historical 2-5%), AWS AI growth is accelerating, advertising is a $50B+ hidden profit pool, and international profitability is finally within reach. The $151B CapEx is terrifying but builds assets that generate 20%+ returns over 20 years.

The bear case: the CapEx is too aggressive, regulatory risk is real, international markets may never be profitable, and the stock has limited upside catalysts in the near term.

For a Moatery investor:

• If you don't own it: Start a position. Size it knowing it's fairly valued rather than cheap. 2-3% of portfolio. Add on meaningful pullbacks.

• If you own it: Hold. The business is better than the stock price reflects. The margin expansion story has years to run. AWS AI is a legitimate growth driver.

• Better opportunity: MSFT at $383 (23x earnings, 27% ROIC) offers a more compelling risk/reward profile today. But AMZN at $245 is a business you can own for 10+ years and sleep well.

• Catalysts to watch: Earnings July 30, 2026. AWS AI revenue disclosures. Advertising revenue trajectory. International segment profitability. Any news on capital returns (buybacks, dividends) would be a major positive.

Amazon is the everything company with a 15/15 moat, margins that have quadrupled in three years, and a long runway of compounding ahead. The stock is fairly priced for a business this good — and occasionally, fair is good enough.

This analysis was generated using the Moatery multi-lens framework. The six lenses are tools for thinking, not recommendations. Always do your own research.

Next up: META — The Attention Monopolist's AI Gambit?