Cloudflare NET

Reference table

1. What they actually sell — and the segment disclosure is essentially nil

Cloudflare provides an integrated cloud platform spanning security (cloud firewalls, bot mitigation, DDoS protection, IoT security, SSL/TLS, rate limiting), performance (content delivery, intelligent routing, content/mobile/image optimisation), reliability (load balancing, its Anycast network, a virtual backbone, DNS and resolvers, waiting rooms), infrastructure on-ramps and filters (connecting users, devices and locations to the network, with data protection, inspection and access management), and developer services (serverless compute, programmable networks, site creation, domain registration). Founded and led by CEO Matthew Prince. 5,156 full-time employees — worth pausing on, because it means the company generates roughly $420,000 of revenue per employee, a striking figure for an infrastructure business.

The segment disclosure is the weakest of any name in this batch, and it must be stated plainly. The seg_prod block contains exactly one line:

That is the entire product-level disclosure: one segment equal to total revenue. There is no breakout of security versus performance versus developer platform versus edge compute, and critically no separate disclosure of any AI, inference or agent-related revenue line. This is the single largest analytical hole in the name, and it matters more here than almost anywhere, because the entire premium multiple rests on an AI narrative that cannot be verified or sized from the company's own segment reporting. Any figure you see quoted for "Cloudflare's AI revenue" is not sourced from these filings.

Geographic revenue (FY25, $2,167.9M, sums exactly):

Cloudflare is meaningfully more international than Snowflake (US 49.5% versus 77.9%), and Asia Pacific is growing 47.7% — the fastest line in the disclosure and a genuine, under-discussed growth vector.

2. The growth story — accelerating revenue, decelerating gross margin

This table contains the two most important facts in the dive, and they point in opposite directions.

Revenue growth accelerated from +26.5% to +33.5% over five quarters — genuinely excellent. And in exactly the same five quarters, gross margin fell every single quarter, from 75.9% to 71.2% — a 4.7-point compression with no reversal. That is not noise; it is a monotonic five-quarter trend, and it is the clearest evidence available that Cloudflare is buying its acceleration with capacity.

Annual context:

The honest reading of that table is genuinely mixed, and both halves are true:

• The good and it is very good: free-cash-flow margin has improved three years running (9.2% → 11.7% → 15.0%), the accumulated deficit is only $1.205B (versus Snowflake's $9.494B — Cloudflare has burned far less capital to reach scale), stock-based compensation runs 20.8% of revenue (versus Snowflake's 34.1%), and the GAAP loss in Q1'26 was just $22.9M, on track to break even.
• The bad: capex jumped from 10.4% to 15.8% of revenue in one year, gross margin fell 2.8 points annually and 4.7 points across five quarters, and only 52.3% of trailing operating cash flow converts to free cash flow (versus Snowflake's 94.6%). This is a network business becoming a compute business, and compute businesses have different economics.

3. The balance sheet — the leverage step-up nobody talks about

Balance sheet (2025-12-31):

• Cash and short-term investments $4.101B; total debt $3.700B — of which $1.362B is classified as SHORT-TERM and $2.156B long-term.
• A reconciliation note that matters: on a cash-and-short-term-investments basis the company is in a net cash position of roughly $0.401B. The data file's netDebt field reads $2.757B because it nets only pure cash equivalents against total debt. The enterprise value used throughout this dive ($102.997B) reflects the latter, more conservative convention. Both figures are disclosed; neither is smoothed away.
• Total equity $1.459B; debt-to-equity 2.31x; financial leverage 4.04x; accumulated deficit -$1.205B.
• Deferred revenue $684.2M; current ratio 1.96x.
• Property, plant and equipment net $856.3M, up from $635.8M — the physical buildout is visible on the balance sheet.

The step-up is recent and large. Total debt went $1.435B (FY23) → $1.463B (FY24) → $3.700B (FY25) on $1.933B of net debt issuance in FY25 alone. For a company with $1.459B of equity and a $102.997B enterprise value, the absolute leverage is not dangerous — but the direction is, and it coincides exactly with the capex acceleration and the gross-margin compression. Those three facts are one fact.

Dilution, by contrast, is well controlled. Diluted shares: 333.7M (FY23) → 341.4M (FY24) → 348.4M (FY25) → 352.6M (Q1'26), roughly 2% a year, with no buyback programme at all (commonStockRepurchased is zero in FY25 and FY24). That is materially better discipline than Snowflake's, which needed $2.8B of repurchases to still increase its share count.

4. Technicals — a healthy consolidation at the high

• Price $282.91 (the technicals block's last is $282.74), -0.23% from the $283.39 52-week high. Effectively at the top of the range.
• RSI(14) 50.8. This is the most important technical fact and it is a positive: the stock is at its high with neutral momentum, meaning it has gone sideways at the top rather than spiking into it. Contrast Snowflake in this same batch, at its high with RSI 70.0. MACD +8.18 — positive but not extreme.
• 50-DMA $250.61 (price +12.8%), 200-DMA $213.62 (price +32.4%). The 50 is above the 200. Intact uptrend, moderately extended.
• Returns: 3-month +30.0%, 6-month +59.4%, 12-month +36.1%, versus SPY at +5.1% / +9.5% / +19.9%. Note the shape: unlike Snowflake (where +118% of a +37.6% twelve-month return came in three months), Cloudflare's gains are spread across the year — six-month +59.4% versus twelve-month +36.1% means the stock actually gave ground in the first half of the window and has compounded steadily since. That is a healthier structure.
• Maximum drawdown from peak over the window: -0.23% — i.e. the stock is at its peak.
• 52-week low $160.19; the stock is +76.5% off it.

Honest technical read: this is the best chart in the batch. Nothing about it argues for selling. The problem is entirely that a healthy chart at 44.2x trailing sales, two days before earnings, with the sell-side target 9.3% below spot, is not an entry — it is a hold.

5. Knowledge base — one tagged claim, and it is a caution

State the position precisely. Cloudflare has exactly ONE claim tagged to the NET ticker, and it is neutral. A name-level search returns eight further Cloudflare-referencing claims, of which four are substantive and bullish (all from a single 2025-08-07 source), two are tangential, and two carry a source slug that our attribution audit flags as unreliable and are therefore EXCLUDED from the conviction reading rather than counted as support. The front matter carries kb_claim_count: 1.

The one tagged claim — and it is the single most useful sentence in the file:

• Gavin Baker, 2026-06-15, stance NEUTRAL, conviction 45: "CDNs like Cloudflare have a path toward the token path and command latency premiums, but currently deliver under 1% — maybe under 10bps — of tokens."

Read that carefully, because it is both halves of the argument in one line. The path exists. Latency premiums are real. And the current delivery is under 1%, possibly under 10 basis points. This is the most recent, most specific, most quantified statement any tracked voice has made about the exact thesis that justifies a 44x sales multiple, and it is a caution rather than an endorsement.

The bullish name-level lane — all four claims from a single source, dated 2025-08-07:

• No Priors, conviction 80: "Whatever the platform shift—crypto, AI, agents—it still needs a network, so Cloudflare stays at the center of transactions."
• No Priors, conviction 78: "Cloudflare, sitting in front of much of the web, can enforce crawler permissions and become the payments layer for content licensing."
• No Priors, conviction 76: "Agent-to-agent traffic (MCP) must pass through the network, so Cloudflare becomes central to agent security, rails, and payments."
• No Priors, conviction 75: "Some models are too big for device inference; the edge network is the next-best place, and that market has finally arrived for Cloudflare."

Two honesty notes on that lane. First, all four are from the same source on the same date, which means the "breadth" of 4 claims is really breadth of 1 — a single conversation, distilled into four propositions. Second, they are now roughly a year old, and the more recent quantified check (Baker, 2026-06-15) says the revenue has not yet followed.

Excluded and disclosed: two 2026-04-17 claims referencing Cloudflare's crawler-gating and publisher-compensation position carry a source slug that the Synthos attribution audit identifies as unreliable at the speaker level. They are not counted toward conviction and not quoted as evidence, though the reader should know they exist and are directionally bullish. Two further claims (a 2026-04-06 post-quantum-migration comparison and a 2026-04-07 processor-adoption note) mention Cloudflare only incidentally and carry no investment view on the equity.

Read. This is a Low-Moderate, thin lane whose most current entry is a neutral caution. It supports the structural thesis (the network is architecturally advantaged) and explicitly does not support the timing thesis (the AI revenue is here now). That is exactly how this dive is weighted.

6. Valuation — priced in or room?

At $282.91 (market cap $100.4B, EV ~$103.0B, ~354.9M shares):

Other current metrics: EV/free cash flow 289.7x, price/free cash flow 282.4x, EV/EBITDA 747.0x, price/book 65.3x, free-cash-flow yield 0.35%, no dividend.

Estimate-quality warning, stated up front. FY26 and FY27 are properly covered (25 revenue analysts, 18-19 on EPS). Beyond that the EPS series collapses: FY28 EPS rests on 4 analysts with a range of -$6.27 to +$8.18 — a range so wide it carries no information — and FY29 on 2 analysts. The FY30 estimate has netIncomeAvg of 0 and epsAvg of 0, i.e. the earnings estimate is simply missing. This dive therefore anchors valuation to FY27 revenue and treats FY28-30 EPS as unusable. That is why every scenario below is built on a revenue multiple rather than a forward P/E.

6a. What today's price assumes (the inversion)

At $282.91 — 44.2x trailing EV/sales, 36.7x FY26E sales, 236x FY26E non-GAAP EPS — the price makes these falsifiable claims. Sources labeled.

• That revenue compounds at ~28% through FY28. Consensus: $2.810B → $3.595B → $4.581B, i.e. +29.6%, +27.9%, +27.4%. Arithmetic: if growth settles at 22% instead, FY28 revenue is ~$4.05B and the implied FY28 EV/sales rises from 22.5x to 25.4x — roughly a 13% valuation headwind from the growth assumption alone.
• That gross margin stabilises. Arithmetic, and this is the fragile one: gross margin has fallen from 75.9% to 71.2% across five consecutive quarters. The price implies it holds near 71-73%. If the trend continues at its recent pace for another four quarters, gross margin approaches 67-68%, which on FY27E revenue of $3.595B is roughly $140-180M less gross profit than a 71% assumption — meaningful against a consensus FY27 net income of $609M. This is the single most fragile assumption in the price.
• That the AI token path becomes revenue. Panel, labeled as such: Gavin Baker (2026-06-15) puts current delivery at "under 1% — maybe under 10bps — of tokens." Arithmetic: at 44.2x trailing sales, Cloudflare is priced roughly 2.1x Snowflake's 21.3x trailing multiple despite growing at a similar rate (+33.5% versus +33.5%) with a lower gross margin (73.3% versus 67.1% — Cloudflare higher) and worse free-cash-flow conversion (52.3% versus 94.6%). The premium is the AI option, and there is no disclosed revenue line supporting it — recall seg_prod contains one undifferentiated segment.
• That the market keeps paying ~29x forward sales two years out. Arithmetic: even hitting FY27 consensus exactly leaves the stock at 28.7x EV/sales, the highest forward multiple among the peers this data file supplies.
• That the sell side is 9% too cautious. From the data file: consensus target $256.58 (high $330, low $136) on ratings of 29 Buy / 9 Hold / 2 Sell. The stock trades 10.3% above the average target of a heavily bullish analyst group.

6b. The return bridge (why the multiple moves)

Expected return ≈ EPS growth + multiple drift + shareholder yield.

Shareholder yield is slightly negative — no dividend, no buyback, and ~2% annual dilution from stock compensation (352.6M diluted shares at Q1'26 versus 333.7M at FY23).

Revenue and earnings growth is the entire return, and it is strong: consensus revenue compounds ~28% and non-GAAP EPS goes from $1.198 to $1.587 (+32%) between FY26E and FY27E. Our base case assumes multiple COMPRESSION — from 36.7x FY26E EV/sales today to roughly 25.4x FY27E at our $250 base. We assume compression rather than a hold for three specific reasons, each of which we can name: (i) the gross-margin trend is going the wrong way, which lowers the quality of each revenue dollar being multiplied; (ii) the AI premium in the multiple is unsupported by any disclosed revenue line and is contradicted by the panel's most current quantification; (iii) the entire covering analyst group already has a target below the price, which historically bounds how much further a multiple runs without an estimate revision.

Note carefully what is NOT re-rating. The fall from 44.2x trailing to 28.7x FY27E at a constant price is mechanical rolldown — it happens automatically as revenue grows and is not evidence the stock gets cheaper. Only movement relative to the multiple the market is paying at each date counts, and on that basis our base case is a genuine de-rating of roughly 30%.

If your thesis requires the multiple to HOLD near 36x forward sales, that is your fragile leg — and it means you are underwriting the token-path revenue arriving inside the next 12-18 months.

Scenario arithmetic:

• Bear ~$150 (-47%). The AI premium unwinds and the market re-prices Cloudflare as an excellent-but-conventional infrastructure compounder. Market cap $53.2B, EV $55.8B = 15.5x FY27E revenue ($3.595B) — still a premium multiple, and roughly where high-quality 25%+ software compounders have historically cleared. Cross-check: $150 is 6.4% below the low sell-side target of $136 plus a margin, and 6.4% below the $160.19 52-week low.
• Base ~$250 (-12%). Market cap $88.7B, EV $91.3B = 25.4x FY27E revenue, or 32.5x FY26E. This assumes consensus revenue is met and the multiple compresses roughly 30% from today's 36.7x FY26E. Cross-check: it sits within $7 of the street's own $256.58 consensus target — again, two independent methods converging.
• Bull ~$375 (+33%). The token path converts: an AI or agent revenue line becomes separately disclosed and material, gross margin stabilises, and the market pays for proof rather than possibility. Market cap $133.1B, EV $135.7B = 37.7x FY27E revenue and 29.6x FY28E. Cross-check: $375 is above the highest sell-side target of $330, so this case requires the street to raise, not merely to be met.

6c. Variant perception (where we differ, and what would surprise)

Where we differ from the market:

• We think the five-quarter gross-margin decline is the most under-discussed fact in this equity. The market is focused on the +33.5% revenue line. Nobody is discounting a monotonic 4.7-point margin compression that coincides exactly with a 50% jump in capex intensity and a $1.933B debt raise. Those three data points are the same story — Cloudflare is becoming more capital-intensive as it serves inference — and a 44x sales multiple is not the multiple for a business whose gross margin is trending toward the high sixties.
• We think the panel's "breadth" here is illusory and should lower, not raise, sizing. Four bullish claims from one source on one date in August 2025 is one opinion, not four. The only current quantified check on the thesis is neutral. We treat this as a one-voice lane.
• We are MORE positive than consensus on the international line. Asia Pacific grew 47.7% off a 15.2% revenue share and the US is only 49.5% of the business. That geographic runway is a genuine, unpriced positive, and it is the most likely source of an upside revenue surprise.
• We are more positive than the multiple implies on capital discipline. ~2% annual dilution with no buyback and only a $1.205B accumulated deficit is materially better stewardship than most of this cohort. It does not justify 44x sales, but it is a reason the bear case is $150 rather than $100.

Positive surprises that would force a repricing upward:

• Gross margin stabilising or rising in the 2026-08-06 print. After five consecutive declines, one flat quarter would be a genuine surprise and would remove the single most fragile assumption in section 6a. Watch this number before you watch revenue.
• Any separately disclosed AI, inference or agent revenue line. seg_prod currently contains one undifferentiated segment. Segmentation that sizes the AI business would convert an unverifiable narrative into a modellable one — the highest-information event available on this name.
• Revenue growth above ~34% on 2026-08-06 (consensus ~$664.7M implies ~+29.7%), which would mark the five-quarter acceleration as continuing rather than plateauing.
• Capex intensity falling back toward 11-12% of revenue while growth holds, restoring free-cash-flow conversion above 60%.

Negative surprises that would confirm the bear case:

• Gross margin below 70%. A sixth consecutive decline through a psychologically round number, at 44x sales, is the most likely trigger for a violent de-rating.
• Revenue growth below ~28%, ending the acceleration narrative.
• Further debt issuance beyond the $3.700B already outstanding, particularly with $1.362B already classified short-term.
• Continued insider selling. The CFO sold into $266-278 on 2026-07-17 (see section 8); more of the same near the high would compound the signal from a consensus target already below spot.

7. Moat and competitive position

A real architectural moat, of an unusual kind. Cloudflare's advantage is not a product; it is placement. A globally distributed Anycast network sitting in front of a large share of internet traffic is extraordinarily expensive to replicate, and its value compounds with every additional point of presence and every additional customer whose traffic improves the threat-intelligence dataset. The security business is genuinely sticky (nobody rips out DDoS protection casually), and the developer platform creates switching costs at the code level.

The margin evidence is strong: 73.3% TTM gross margin on a business that operates physical infrastructure in hundreds of locations, and $420,000 of revenue per employee with just 5,156 people. Those are not the numbers of a commoditised CDN.

Where the moat is contested:

• The edge-inference economics are unproven. Serving models at the edge requires expensive accelerators in many locations, which is precisely what the five-quarter gross-margin decline and the 15.8% capex ratio look like. A moat that costs you 4.7 points of gross margin to defend is a moat with a toll.
• The hyperscalers can price against it. Every major cloud provider offers CDN, security and edge compute, and can bundle them below cost to defend a larger relationship.
• The token path is aspirational. Per Gavin Baker (2026-06-15), current delivery is "under 1% — maybe under 10bps." A moat around a market you do not yet meaningfully serve is a moat around a plan.

Peer context from this data file (prices as supplied): Fortinet $163.21 / $119.7B, Marvell $193.78 / $169.7B, Snowflake $307.53 / $106.6B, Cadence $336.15 / $92.6B, Zscaler $154.46 / $25.0B, CoreWeave $85.76 / $46.8B. Cloudflare's $100.4B market cap on $2.17B of FY25 revenue is a 46x price-to-sales ratio; Fortinet, its nearest security comparator by size, is a materially larger revenue business at a similar market cap.

8. Management, capital allocation and insider activity

Capital allocation is growth-first and debt-funded: no dividend, no buyback in any year of the available cash-flow data, $342.6M of capex (15.8% of revenue, up from 11.1%), $451.5M of stock-based compensation (20.8% of revenue), and $1.933B of net debt raised in FY25. R&D is $512.5M (23.5% of revenue) and SG&A $1.310B (59.1% of revenue) — a heavy go-to-market load, which is where the GAAP loss comes from. Founder-CEO Matthew Prince.

The dilution discipline deserves credit and should be said clearly: ~2% annual share growth with zero buybacks is better capital stewardship than a peer spending billions on repurchases just to stand still. Cloudflare has reached $2.17B of revenue with a cumulative accumulated deficit of only $1.205B — it has burned remarkably little to get here.

Insider activity is mildly negative and small in scale. All eight insider filings in the data file are dated 2026-07-17 and belong to a single person, CFO Thomas J. Seifert:

Total: 7,699 shares sold for roughly $2.1M, following a 10,000-share conversion on the same day, leaving 116,091 shares held. This is a small, laddered, single-officer sale that looks like a planned disposition rather than a conviction signal — materially less concerning than the ~$23M of multi-insider selling seen at Snowflake in the same period. There is no insider buying. We flag it and do not over-read it.

9. Verdict, kill-criteria and flip conditions

Watch (do not initiate). Hold existing.

Cloudflare is an excellent company with a genuinely unusual structural position, and this dive should not be read as bearish on the business. Revenue growth accelerated across five quarters to +33.5%; free-cash-flow margin has improved three years running to 15.0%; the GAAP loss has narrowed to $22.9M in Q1'26; capital discipline is better than the cohort; Asia Pacific is growing 47.7%; and the structural case — that every platform shift still needs a network — is compelling.

But you are asked to pay 44.2x trailing sales, 236x this year's non-GAAP earnings and 282x trailing free cash flow, for a 0.35% free-cash-flow yield, at 0.2% from the all-time high, two days before an earnings print, with a consensus target 9.3% below the stock, in a business whose gross margin has fallen five quarters in a row and whose capex intensity has jumped 50% in a year. And the panel's single most current, most specific check on the thesis that justifies that premium says the token path today is "under 1% — maybe under 10bps."

Base fair value $250, -12%, converging within $7 of the street's own target. Watch.

Pre-registered entry triggers (what would move this to Stage-In or Buy):

• A pullback to ~$250 or the 50-DMA (~$251) with revenue growth and gross margin both intact. At $250 the base case is fair value and the FY27E multiple falls to 25.4x — defensible for a 28% compounder.
• Gross margin stabilising above 71% in the 2026-08-06 print, ending the five-quarter slide. This single number is the most important thing in the next report.
• Any separately disclosed AI, inference or agent revenue line that sizes the option. That converts narrative into model and would justify raising the base case materially.
• Revenue growth above ~34%, marking the acceleration as continuing.

Pre-registered KILL / avoid criteria:

• Gross margin below 70% — a sixth straight decline through a round number, at a 44x sales multiple.
• Revenue growth below ~28%, ending the acceleration story with no valuation cushion beneath it.
• Further debt issuance beyond the $3.700B outstanding, especially with $1.362B already short-term, or capex intensity rising above 18% of revenue.
• A decisive break of the 50-DMA (~$251) on a disappointing print, which at these multiples opens the gap toward the 200-DMA (~$214) with nothing structural in between.

Where NET fits in the Synthos Framework Portfolio. The AI-software / network-infrastructure sleeve, as the network-layer leg. Hold existing at 1-3%; new capital waits for the print and then for ~$250 or better. On overlap: NET is complementary, not redundant to Snowflake in this same batch — Cloudflare sits at the network layer where traffic is routed and secured, Snowflake at the data layer where queries are answered. But they share the same dominant risk factor (a very high multiple on consumption-linked revenue, both at or near 52-week highs, both with consensus targets below spot), so the two should be sized as a combined sleeve rather than independently, and the combined weight should not exceed what you would give a single high-multiple position. Logged as a tracked Synthos call (Watch) as of 2026-08-04 at $282.91.

Single biggest risk: 44.2x trailing sales with a gross margin that has fallen five consecutive quarters. There is no valuation floor — the free-cash-flow yield is 0.35% and the company is GAAP loss-making — so a single quarter that disappoints on either revenue growth or margin has nothing beneath it, and the entire covering analyst group already has a target below the price.

Most fragile assumption in the price: that gross margin stabilises near 71-73%. It has declined monotonically from 75.9% over five quarters while capex intensity rose from 10.4% to 15.8% of revenue — and nothing in the disclosure indicates that trend has turned.

Provenance and disclosures

• Traceability: 1 knowledge-base claim tagged to NET — Gavin Baker, 2026-06-15, stance NEUTRAL, conviction 45. Four further bullish name-level claims (conviction 75-80) all originate from a single source on a single date (2025-08-07) and are counted as one voice, not four. Two additional Cloudflare-referencing claims dated 2026-04-17 are EXCLUDED because their source slug is flagged as unreliable at the speaker level by the Synthos attribution audit; they are disclosed here but not counted toward conviction or quoted as evidence. Two further mentions (2026-04-06, 2026-04-07) are incidental and carry no view on the equity. Net conviction: thin, positive on structure, cautionary on timing. All quotations verbatim with source and date.
• Data as-of: fundamentals 2026-03-31 (Q1'26, filed 2026-05-08; annual FY25 ended 2025-12-31, filed 2026-02-26) · estimates 2026-08-04 · prices 2026-08-04 (quote $282.91, technicals last $282.74, 50-DMA $250.61, 200-DMA $213.62, RSI 50.8, MACD +8.18, 52-wk range $160.19-$283.39) · knowledge-base claims to 2026-07-25. All market data comes from the pre-pulled Synthos data file for NET; no figure has been sourced from memory or any other retrieval.
• Named data gaps: seg_prod contains a single undifferentiated "Reportable Segment" line equal to total revenue — the company provides NO product-level segmentation in this dataset, and therefore no AI, inference, security, performance or developer-platform revenue split exists. This is the largest analytical hole in the name and no such split is asserted anywhere in this dive. Analyst EPS coverage collapses beyond FY27: 4 analysts on FY28 with a -$6.27 to +$8.18 range, 2 on FY29, and the FY30 EPS estimate is literally 0 (missing) — all valuation scenarios are therefore built on revenue multiples, not forward P/E.
• Reconciliation note: the data file's netDebt of $2.757B nets total debt of $3.700B against cash equivalents only. On a cash-and-short-term-investments basis ($4.101B), Cloudflare holds roughly $0.401B of net cash. The enterprise value used throughout ($102.997B, from the vendor's key-metrics block) reflects the more conservative convention. Both figures are disclosed.
• Non-GAAP note: all forward P/E figures use the analyst (non-GAAP) EPS series. Cloudflare is GAAP loss-making (FY25 net loss $102.3M; Q1'26 net loss $22.9M) and the trailing P/E is not meaningful. Stock-based compensation runs 20.8% of revenue.
• Fair-value method: scenario multiples on consensus revenue (not EPS, for the coverage reasons above), applied to ~354.9M shares (market cap $100.4B ÷ $282.91) and the vendor enterprise value of $103.0B. Base $250 = 25.4x FY27E revenue. Bear $150 = 15.5x FY27E revenue. Bull $375 = 37.7x FY27E revenue, cross-checked at 29.6x FY28E. No discounted cash flow is used and none is implied.
• Sell-side note: consensus target $256.58 (high $330, low $136), 9.3% below the current price, on ratings of 29 Buy / 9 Hold / 2 Sell. FMP composite rating D+, overall score 1 of 5. This dive's base case sits within $7 of the street's target.
• Timing caveat: this dive is written two days before the Q2'26 print (2026-08-06), deliberately, to frame the decision around it. The result may move the picture materially — particularly the gross-margin line — and the verdict is sized accordingly.
• Not investment advice. Independent research, educational and informational only, never personalized.
• Version: 2026-08-04-full.