Zoom Communications ZM

Reference table

1. The vendor-data correction — read this first

This is the most consequential single fact in the dive, and it is a data-integrity finding rather than an analytical one.

Zoom's FY26 balance sheet (2026-01-31) shows:

• Cash and cash equivalents: $1,273M
• Short-term investments: $6,544M
• Cash and short-term investments: $7,817M
• Total debt: $58.5M
• Net cash: $7,758M

The vendor reports an enterprise value of $27,979M, which against a $28,810M market capitalisation implies net cash of only $831M — i.e. it nets the $1,273M cash line, subtracts the debt, and ignores $6,544M of short-term investments entirely.

Every EV-based figure in this dive uses the corrected number. The practical consequence is that any screen or model using the standard enterprise value will show Zoom as roughly 30% more expensive than it is, and since Zoom is precisely the kind of name investors find on screens, this is a plausible source of persistent mispricing. This is the single most concrete edge in this dive.

Short-term investments in a portfolio of this composition — held by a company with no debt, no acquisition programme of scale, and $1.9B of annual free cash flow — are cash for every practical purpose. Treating them otherwise is simply wrong.

2. What they actually sell — and the segment disclosure problem

Zoom's product disclosure has deteriorated. Here is the full seg_prod history:

In every year, all revenue sits in one line. There is no split between the core meetings product and the newer Phone, Contact Center, Rooms or AI offerings — and in the last two years the label has degraded to the uninformative "Reportable Segment." We cannot see whether the 4% growth is a declining meetings business offset by a fast-growing product portfolio, or a flat everything. That distinction is worth a great deal to a forecast and it is unavailable.

Geography is disclosed and is stable:

No region is growing above 5% and no region is declining. This is a mature global business with no geographic inflection anywhere. It is also, usefully, evidence against a hidden decline — if the core were eroding fast while new products grew, you would expect more regional dispersion than 3.4% to 4.7%.

3. The earnings-quality issue — why the 14.4x trailing P/E is not real

Zoom's reported GAAP net income is materially larger than its operating income, and we cannot identify why from this file. This must be handled before any valuation.

Non-operating income of $1.45B over the trailing twelve months exceeds operating income of $1.19B. Some of this is genuinely interest income — $7.8B of cash and investments at a plausible 4.5% yield would generate roughly $350M annually — but that accounts for less than a quarter of it. The remaining ~$1.1B is not identifiable from the fields available here. The most likely candidates are investment portfolio gains and tax-related items, but we will not assert what we cannot see.

Three consequences, and they matter:

1. The trailing GAAP P/E of 14.4x is flattered. On operating income alone, after tax at the observed 21.5% effective rate, the underlying operating earnings power is roughly $936M, or $3.12 per diluted share — a trailing operating P/E of 31.5x.

2. The vendor's ratio block encodes the same distortion. ebitdaMarginTTM is 56.5% and ebitMarginTTM is 53.5% while operatingProfitMarginTTM is 24.2% — the vendor's "EBIT" field is populated with pre-tax income, so its EBITDA figure absorbs all the non-operating income. The reported EV/EBITDA of 10.0x is therefore meaningless. Our corrected EV against genuine operating EBITDA (operating income $1,192.5M plus D&A of roughly $416M) is $21.05B / $1,609M = 13.1x.

3. Consensus EPS is the better anchor and is what this dive uses. FY27E EPS of $6.04 against FY26 revenue-based operating economics is a far more defensible basis than a trailing GAAP figure inflated by unidentifiable items. All forward multiples here use consensus EPS.

Free cash flow is the one number that cannot be distorted this way, and it is excellent: FY26 operating cash flow $1,989M less capital expenditure $65.0M = $1,924M, a 39.5% free-cash-flow margin. That is the real earnings power of this business.

4. The buyback — the actual engine of the return

Read the last three rows together — this is the whole investment case.

• Stock-based compensation has fallen every year, from $1,286M (FY23) to $761M (FY26) — a 41% reduction while revenue grew 11%. As a share of revenue it has gone from 29.3% to 15.6%. Still high in absolute terms, but the direction is unambiguously right.
• The FY26 buyback of $1,621M is 2.1x the stock compensation expense, which means the repurchase is not merely offsetting dilution — it is genuinely shrinking the company. Diluted shares fell from 316.7M (Q4'FY25) to 300.2M (Q1'FY27), a 5.2% reduction in five quarters.
• $1,621M is 5.6% of today's $28.81B market capitalisation, funded entirely from a $1,924M free cash flow with $7.82B of cash left untouched.

Capital expenditure of $65M — 1.3% of revenue and down 53% year over year — deserves a note. Zoom's cloud infrastructure is largely rented rather than owned, so the business converts nearly all operating cash flow into free cash flow (the ratio is 96.8%). That is structurally attractive and it is why the free-cash-flow margin exceeds the operating margin.

Zoom pays no dividend. Total shareholder yield is therefore the buyback alone: 5.6% at the FY26 pace.

5. Valuation — priced in or room?

At $98.25 (market cap $28.81B, corrected EV $21.05B), trailing revenue is $4.933B, trailing free cash flow is approximately $1.90B, and net cash per share is $26.46, giving an ex-cash price of $71.79.

The FY31E EPS of $5.72 sitting BELOW FY30E's $7.00 is not a coherent forecast — it reflects a changing 2-to-5 analyst panel. FY30E and FY31E are excluded from all valuation work. FY27E and FY28E carry 16 EPS contributors each and are the reliable base.

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

At $98.25 — 16.3x FY27E consensus EPS, or 11.9x ex-cash — the price embeds roughly:

• Revenue grows at ~4% annually and does not decline, reaching $5.29B by FY28. (Source: consensus, 22-23 analysts — the deepest coverage in this batch.) Falsifier: any quarter with revenue growth below 2% year over year. The last four quarters ran +4.7%, +4.4%, +5.3% and +5.4%.
• Gross margin holds near 77%. The trailing figure is 77.4% and the FY26 figure was 77.0%, up from 74.9% in FY23. (Reported.) Falsifier: gross margin below 75%, which would indicate price competition arriving.
• The buyback continues at roughly $1.5-1.6B annually, delivering ~5% annual share-count reduction. Without it, consensus EPS growth of +4.0% (FY27E to FY28E) is barely above the revenue growth rate — i.e. essentially all of the modelled per-share growth beyond revenue comes from the share count. (Our computation.) Falsifier: quarterly buybacks below $250M, or the diluted share count ceasing to fall from 300.2M.
• The market keeps paying only ~16x forward, or ~12x ex-cash. At an unchanged price the multiple rolls to 14.7x FY29E. No re-rating is priced.
• THE MOST FRAGILE ASSUMPTION: that 4% growth is a floor rather than a waypoint. Every bullet above rests on revenue not turning negative. A 4% grower and a -2% decliner are worth entirely different multiples, and there is no disclosure in this file — no product-level revenue, no customer count, no net revenue retention — that would give early warning of the transition. The single "Reportable Segment" line is a genuine impediment. Falsifier: two consecutive quarters below 2% growth.

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

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

Our base case of $115 is built ex-cash and deliberately so: 14.0x FY28E consensus EPS of $6.28 = $87.92 of operating-business value, plus $26.46 of net cash per share = $114.38. Round to $115 (+17%).

Decomposing the return: EPS grows from $6.04 (FY27E) to $6.28 (FY28E), just +4.0% — so earnings growth contributes very little over the horizon. The buyback contributes ~5% annually to share count and is already embedded in those consensus EPS figures. The bulk of the base-case return therefore comes from a modest ex-cash multiple expansion, from 11.9x to 14.0x — and that must be named as the fragile leg.

What would drive that expansion? Precisely the correction in section 1: as the market recognises that 27% of the market capitalisation is cash and that the ex-cash multiple is 12x rather than 16x, a 14x ex-cash multiple for a 4%-growing, 40%-free-cash-flow-margin, zero-debt business becomes obviously reasonable. That is a re-rating driven by an accounting recognition rather than an operational change, which is a legitimate but slower mechanism than an earnings beat. If it does not happen, the return is roughly 4-5% annually plus the buyback effect — call it 9-10% — which is still acceptable but far less interesting.

Shareholder yield is 5.6% and is doing more work here than in any other name in this batch. It is the reason a 4% grower can produce a double-digit return at all.

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

• Our variant perception is specific, mechanical, and it is the reason to write this dive at all: the market and the data infrastructure understate Zoom's net cash by $6.5B. Consensus rates the stock Hold (19 buy, 27 hold, 3 sell) with a $119.27 target — only 21% above spot — and the vendor's own enterprise value overstates the operating-business valuation by 33%. We think Zoom is roughly 25% cheaper than it appears on a standard screen, and that is not an opinion about the business, it is arithmetic about the balance sheet.
• Where we AGREE with the cautious consensus: the growth rate. We do not model reacceleration and we explicitly do not underwrite the AI-product optionality. On the operating business we have no variant perception, and per house rules that pulls the verdict toward the value end rather than the growth end. The Stage-In is for the balance sheet and the buyback, not for a thesis about Zoom winning anything.
• Positive surprise that would force a repricing: any quarter above 7% revenue growth, or a segment disclosure showing a fast-growing non-meetings product portfolio. The single "Reportable Segment" line is currently hiding whatever mix shift may exist, and its removal would be a material information event. Watchable at: 2026-08-25 and the FY27 annual disclosure.
• Negative surprise, and it is the one that matters: a large acquisition. Zoom holds $7.82B of cash, has no debt and pays no dividend. The entire base case values that cash at 100 cents and assumes it keeps funding buybacks. A multi-billion-dollar acquisition would simultaneously remove the cash from the valuation, halt the buyback engine, and introduce integration risk into a business whose sole attraction is predictability. This is the largest single-event risk in the name and it is entirely within management's control. Watchable at: any deal announcement.

Synthos fair values

• Bear ~$78 (-21%). AI-native communication tools and free bundled incumbents push growth from +4% to flat or negative; the market prices Zoom as a melting asset at 8.0x ex-cash FY28E EPS of $6.28 ($50.24) plus $26.46 of net cash = $76.70. Rounds to $78. Note this is still 12% above the 52-week low of $69.77, i.e. the market has already visited this scenario within the last year.
• Base ~$115 (+17%). 14.0x ex-cash FY28E EPS of $6.28, plus $26.46 net cash per share. Assumes 4% growth holds, the buyback continues, and the ex-cash multiple normalises toward what a debt-free 40%-free-cash-flow-margin business deserves.
• Bull ~$145 (+48%). The agentic product layer — the cognitive_revolution claim — produces a pricing tier and growth inflects above 6%; 18.0x ex-cash FY29E EPS of $6.69 ($120.42) plus $26.46 = $146.88. Rounds to $145. This is 9% above the street's high target of $133 and 30% above the 52-week high, so it requires genuinely new information, not just sentiment.

6. Knowledge base — heavy filtering required, and the result is cautionary

A raw search on ZM and Zoom returned 15 hits. Only 9 are usable claims about the company, and the filtering itself is instructive.

EXCLUDED — the "zoom out" collision (2 claims). Two hits use "zoom" as a verb and have nothing whatsoever to do with the company: raoul_pal (2025-11-27, conviction 80): "Zoom out past cycles; crypto is going in the right direction on fundamentals", and a quarantined entry (2023-07-18): "Zoom out and simplify a thesis to its most durable elements." This is exactly the name-collision failure mode that has previously polluted single-name lanes, and it would have inflated this one by 22%.

EXCLUDED — quarantined attributions (2 claims). Two hits carry the quarantine_misattributed thinker tag, including one from 2026-02-28 stating that "AI is pressuring jobs — even cheap-labor Fiverr, plus Zoom and Bosch, are cutting staff." Quarantined entries are excluded from all lanes on principle, regardless of how relevant the content appears. Note that this particular claim is directionally negative for Zoom, so excluding it makes the lane slightly more positive than the raw file.

INCLUDED — the 9 usable claims:

Constructive on the business:

• invest_like_the_best (2022-07-26, conviction 78, neutral): "Don't dismiss crowded markets with free incumbents — Zoom built a large independent public company despite WebEx/Google/Microsoft free products." The single most relevant claim in the lane, and it directly addresses the central bear case. Zoom has now compounded revenue and expanded margins for four more years since this was said.
• invest_like_the_best (2026-03-24, conviction 68, bullish): "Bought Zoom via a derivative/secondary structure since front-door was closed; made two-to-three times money." Historical and almost certainly pre-IPO; weighted low for current relevance.
• cognitive_revolution (2026-03-22, conviction 60, bullish): "Hyperscalers and tech-first firms (AWS, Zoom, Glean, Airtable) are building core agentic products on third-party tool-execution infrastructure, signaling real enterprise demand." The most current forward-looking claim and the sole source for the exponential-potential score of 2 rather than 1.
• lex_fridman (2021-04-10, conviction 70, bullish): "Zoom switched the entire higher-ed system to remote in ~two weeks with a remarkably high performance rate — competence under stress."
• nassim_taleb (2020-10-12, conviction 75, bullish): "Anti-fragile names like Zoom, Amazon, and warehousing real estate actually thrive — not just survive — when COVID conditions get worse."
• nassim_taleb (2025-08-21, conviction 80, bullish): "Healthy systems learn from errors and adapt (e.g. Zoom, remote work); whatever is fragile collapses."

Cautionary on the security:

• we_study_billionaires (2024-09-07, conviction 72, neutral): "Paying an infinite/nosebleed valuation ends in misery even when fundamentals keep growing; cites Zoom (1,600x to 15x EV/EBIT, 90% drawdown) and EDS (PE 500, stock $40 to $3 while earnings grew)." Zoom is used here as the textbook example of valuation destruction. This is the most-cited claim in the lane and it is about the stock rather than the business.
• raoul_pal (2026-04-23, conviction 62, bullish on the theme, bearish on Zoom): "The more life goes digital/virtualized, the more people crave real human experiences — tourism, wellness tourism, hotels and live events keep rising in price while digital (e.g. Zoom) costs go to zero." The cleanest statement of the commoditisation risk.
• arthur_hayes (2023-09-17, conviction 55, neutral): "Former 2021 bubble names like Peloton and Zoom lacked earnings to justify valuations, but are far better deals now than at their peaks." Notable as a claim that has aged well — Zoom subsequently grew GAAP operating income from $245M (FY23) to $1,124M (FY26).

Read. A breadth-7, net-cautionary lane. Four claims predate 2022 and describe pandemic-era operational competence rather than investment merit. The two most decision-relevant claims are the we_study_billionaires valuation warning and the raoul_pal commoditisation argument — both cautionary, and both about the stock rather than the product. The counterweight is the invest_like_the_best observation that Zoom has already survived free bundled competition for a decade, which the subsequent financial record supports.

This lane materially shapes the verdict. It is why the position is framed as a value and cash-return holding rather than a growth one, why exponential potential is 2, and why the bull case at $145 requires new information rather than sentiment.

7. Technicals — the best setup in this batch

• Price $98.25 (+2.27% today), -12.2% from the $111.88 52-week high, +40.8% above the $69.77 low.
• Above BOTH moving averages — 50-DMA $93.04 (+5.6%) and 200-DMA $88.13 (+11.5%) — with the 50-DMA ABOVE the 200-DMA. A fully intact, well-ordered uptrend, and the only such structure in this batch.
• RSI 64.3 — firm, with room before the overbought threshold. MACD +0.90, positive.
• Relative performance: +32.7% over twelve months against SPY +19.9% and QQQ +23.9%. Zoom is the only name in this batch that has beaten both indices over a year. Six-month +6.7%, three-month -5.0% — a recent consolidation within an uptrend.
• Maximum drawdown from peak: -82.7%. This is the 2021 bubble scar and it belongs in the record. Anyone underwriting Zoom should know the stock has already destroyed 80%+ of its peak value once, which is exactly the we_study_billionaires warning made concrete.
• Beta 1.04 — market-like.

The technical read is constructive and unusually clean. Above both averages, in a proper uptrend, with a mid-60s RSI and outperformance against the index. There is no reason to wait on technical grounds — which is why the staging here is two tranches rather than three.

8. Moat & competitive position

Zoom's moat is product quality plus workflow embedding in a category where the incumbents give the product away. That combination is unusual and it deserves careful handling because it is the crux of the whole debate.

The bear framing is straightforward: Microsoft Teams, Google Meet and Cisco Webex bundle equivalent video conferencing into suites enterprises already buy. A standalone product competing against free should erode. raoul_pal's formulation — "digital costs go to zero" — is the sharp version.

The evidence says otherwise, so far. Zoom has grown revenue from $4.10B (FY22) to $4.87B (FY26) through the entire period of Teams' aggressive bundling, and has simultaneously expanded gross margin from 74.3% to 77.0% and GAAP operating margin from 5.6% (FY23) to 23.1% (FY26). A company being commoditised does not expand gross margin by 270 basis points while raising operating margin fourfold. invest_like_the_best made this precise observation in 2022 and the subsequent four years of financials have vindicated it.

The mechanism is that Zoom is chosen by users and defended by IT rather than the reverse — a reliability and simplicity advantage in a category where failure is highly visible. Enterprise deployments of communications infrastructure, particularly Phone and Contact Center, also carry meaningful switching friction.

Three honest constraints:

1. We cannot verify the product mix. The single "Reportable Segment" line means we cannot tell whether the core meetings business is eroding while Phone and Contact Center grow, or whether everything is flat. This is a real gap in the moat assessment.

2. 4% growth is what a defended-but-not-expanding position looks like. The moat is holding the line, not advancing it.

3. AI-native communication tools are a new axis of competition that the historical record does not speak to. Surviving Teams says nothing about surviving a product category that does not yet exist.

Verdict on the moat: real, tested, defensive. Zoom has held a strong position against overwhelming distribution advantages for a decade. That is worth a fair multiple. It is not worth a growth multiple, and the price does not ask for one.

9. Data integrity — what we corrected and why

Vendor data corruption checking is standard for these dives. For ZM:

• CORRECTED — the largest single correction in this batch: enterprise value. The vendor reports $27,979M, netting only the $1,273M cash line against $58.5M of debt and ignoring $6,544M of short-term investments. The correct figure is $21,052M. This changes EV/Sales from 5.67x to 4.27x and EV/free-cash-flow from ~14.7x to ~11.1x — a 25% valuation difference. Every EV-based figure in this dive uses the corrected number and this is documented as the central finding in section 1.
• REJECTED: the vendor's EBIT and EBITDA fields, and every ratio derived from them. ebitMarginTTM (53.5%) equals pretaxProfitMarginTTM (53.5%) while operatingProfitMarginTTM is 24.2% — the vendor's "EBIT" is pre-tax income, so its EBITDA absorbs $1.45B of non-operating income. The reported evToEBITDATTM of 10.0x and enterpriseValueMultipleTTM of 10.0x are therefore meaningless. Our computed figure is $21.05B against genuine operating EBITDA of ~$1,609M = 13.1x. The same distortion appears in the quarterly income statements: Q4'FY26 reports EBITDA of $896.6M against operating income of $249.9M and roughly $104M of quarterly D&A — a figure that is impossible on an operating basis.
• REJECTED: FY30E and FY31E consensus EPS. The series runs $6.04, $6.28, $6.69, $7.00 (FY30E), then $5.72 (FY31E) — a non-monotonic decline driven by a collapsing analyst panel (16 contributors for FY28E falling to 2 for FY30E and 5 for FY31E). Excluded. FY27E and FY28E, with 16 contributors each, are the reliable base and are what the valuation uses.
• FLAGGED, unresolved: $1,445.6M of trailing non-operating income that we cannot attribute. Interest on $7.8B of cash and investments accounts for perhaps $350M; the remaining ~$1.1B is not identifiable from the available fields. We therefore do not use trailing GAAP EPS for valuation and present the 31.5x trailing operating P/E alongside the 14.4x headline figure throughout. This is disclosed rather than smoothed.
• GAP: product-level revenue disclosure. seg_prod reports all revenue in a single line labelled "Reportable Segment" for FY25 and FY26 (previously "Unified Communications and Collaboration Platform"). No split between Meetings, Phone, Contact Center, Rooms or AI products. For a company whose entire debate is whether a declining core is being offset by growing adjacencies, this is a material omission.
• GAP: no operating metrics. No customer count, no net revenue retention, no enterprise-customer cohort data, no seat counts. Standard software disclosures, all absent.
• NOTE: insider activity is routine and pre-planned. CEO Eric Yuan conducted conversions and sales on 2026-07-13/14 — approximately 31,400 shares sold at $89.49-$91.61 against a 20.7 million share position. Director H.R. McMaster sold 5 shares. The Yuan transactions are roughly 0.15% of his holding and are the signature of a scheduled trading plan. No signal. No open-market purchases in the file.
• NOTE: fiscal-year convention. Zoom's fiscal year ends 31 January. FY2026 ended 2026-01-31 and the most recent quarter (2026-04-30) is Q1'FY2027. All year labels in this dive follow the company's convention and are stated explicitly to avoid confusion.

10. Verdict, kill-criteria & flip conditions

Stage-In, as a value and cash-return holding. Zoom generates $1.92B of free cash flow on $4.87B of revenue — a 39.5% margin — with a 77.0% gross margin, $7.82B of cash and investments, $58.5M of debt, and a buyback that consumed 5.6% of the market capitalisation last year. After correcting the vendor's $6.5B understatement of net cash, the operating business trades at 4.27x sales, 11.1x free cash flow, and 11.9x forward consensus earnings ex-cash. Base fair value $115 (+17%).

The return does not require the business to improve. Roughly 4% revenue growth plus a ~5% annual share-count reduction plus interest on the cash produces high-single-digit per-share earnings growth, and you are buying that at twelve times ex-cash with a fortress balance sheet underneath. That is a 10-12% expected return with an unusually low probability of permanent capital loss — a rare profile in this batch.

The discipline this name demands is refusing to romanticise it. Consensus models 3.7-4.6% revenue growth for four consecutive years on 22-23 analysts. Our own knowledge base rates the category a commoditising utility and holds Zoom up as the textbook example of what paying a growth multiple for a decelerating business does to you. The maximum drawdown in this file is -82.7%. Growth Quality is 3 and Exponential Potential is 2, and both scores should be read literally.

Staged entry (build a 2-3% position in 2 tranches — two rather than three, because the technical setup gives no reason to wait):

1. Tranche 1 — now (~half of target), at ~$98. Above both moving averages, RSI 64, 11.9x ex-cash forward earnings, 27% of the market capitalisation in cash. The correction in section 1 is the reason to act rather than wait.

2. Tranche 2 — after the 2026-08-25 print, on confirmation that revenue growth held at or above 4% and the buyback continued near the $400M quarterly pace. Or on a pullback to the 200-DMA (~$88), which would take the ex-cash multiple to 10.2x.

Pre-registered KILL / avoid-adding criteria:

• A large acquisition. This is the single most likely way the thesis breaks and it is entirely within management's control. The cash is 27% of the market capitalisation and the buyback is the return engine; deploying the former and halting the latter would invalidate both. Any deal above ~$1.5B triggers a full re-underwrite.
• Two consecutive quarters below 2% revenue growth — the transition from a 4% grower to a declining asset, which is worth a completely different multiple.
• The buyback slowing below ~$250M a quarter, or diluted share count ceasing to fall from 300.2M.
• Gross margin below 75% — the signature of price competition finally biting after a decade of resistance.
• Stock-based compensation reversing higher from the $761M FY26 level, which would mean the buyback is once again merely offsetting dilution rather than shrinking the company.

Pre-registered FLIP TO HIGHER CONVICTION (upsize toward 3%):

• Any quarter above 7% revenue growth, which would break the four-year 3-5% band and force a genuine re-rating.
• A restored product-level segment disclosure showing a fast-growing non-meetings portfolio — this would remove the largest analytical gap in the name.
• A buyback acceleration above $2.0B annually, or the initiation of a dividend, either of which would signal management has concluded the cash has no better use.
• Evidence that the agentic product layer named by cognitive_revolution has produced a priced tier.

Where ZM fits in the Synthos Framework Portfolio. The value and cash-return sleeve, at a 2-3% core weight. Its portfolio function is specific: with a 1.04 beta, a 39.5% free-cash-flow margin, zero debt and 27% of its capitalisation in cash, it is a low-drawdown, cash-generative ballast that still participates in technology. It is emphatically not a substitute for growth technology exposure — pairing it against a high-beta name like RDDT (1.938) or GFS (1.796) is the intended use. Logged as a tracked Synthos call (Stage-In) as of 2026-08-04 at $98.25.

Single biggest risk: category commoditisation by bundled and AI-native competitors. Three larger companies bundle equivalent video conferencing free inside suites enterprises already buy, and AI-native communication products are a new axis Zoom's decade-long survival record says nothing about. Zoom has held its ground impressively — expanding gross margin 270 basis points and quadrupling operating margin while Teams bundled aggressively — but 4% growth is what a successfully defended position looks like, not an advancing one. The most fragile assumption in today's price is that 4% is a floor rather than a waypoint on the way to zero. And the file gives us no early-warning system: a single "Reportable Segment" revenue line, no customer count, and no net revenue retention. If the core erodes, we will learn about it from the consolidated revenue line, at the same time as everybody else.

Provenance & disclosures

• Traceability: 9 usable KB claims across 7 tracked voices (breadth 7, net conviction mixed-cautionary). Constructive: invest_like_the_best (2022-07-26, conv 78 — the most relevant claim in the lane; and 2026-03-24, conv 68), cognitive_revolution (2026-03-22, conv 60), lex_fridman (2021-04-10, conv 70), nassim_taleb (2020-10-12, conv 75; 2025-08-21, conv 80). Cautionary: we_study_billionaires (2024-09-07, conv 72 — Zoom as the textbook valuation-destruction example), raoul_pal (2026-04-23, conv 62 — "digital costs go to zero"), arthur_hayes (2023-09-17, conv 55). Six raw hits excluded: two "zoom out" verb collisions (raoul_pal 2025-11-27 and a 2023-07-18 entry) and two quarantine_misattributed entries — one of which was negative on Zoom, so the exclusion made the lane marginally more positive.
• Data as-of: fundamentals 2026-04-30 (Q1'FY27, filed 2026-05-22) · estimates 2026-08-04 · prices 2026-08-04 ($98.25, +2.27%; 50-DMA $93.04, 200-DMA $88.13, RSI 64.3, beta 1.04) · KB claims 2026-08-04. Sole data source: scripts/deepdive/vti_data/ZM_data.json. Fiscal year ends 31 January.
• Corrected data — the central finding: vendor enterprise value of $27,979M ignores $6,544M of short-term investments; the corrected figure is $21,052M, a 25% reduction that changes EV/Sales from 5.67x to 4.27x. All EV-based figures use the corrected number.
• Rejected data: the vendor's EBIT and EBITDA fields and all derived ratios (ebitMarginTTM equals pretaxProfitMarginTTM, so EBITDA absorbs $1.45B of non-operating income; the reported 10.0x EV/EBITDA is meaningless, our computed figure is 13.1x); FY30E and FY31E consensus EPS (non-monotonic, 2-5 analyst coverage).
• Flagged, unresolved: $1,445.6M of trailing non-operating income exceeding $1,192.5M of operating income, of which perhaps $350M is attributable to interest on the cash portfolio and the remainder is not identifiable from available fields. The trailing GAAP P/E of 14.4x is therefore not used for valuation; the 31.5x trailing operating P/E is presented alongside it.
• Disclosed gaps: all revenue reported in a single "Reportable Segment" line for FY25-FY26 with no product split; no customer count, net revenue retention, seat count or cohort data.
• Fair-value caveat: the $78 / $115 / $145 anchors are ex-cash scenario multiples on consensus EPS plus $26.46 of net cash per share, labelled at each step. The base case assumes ex-cash multiple expansion from 11.9x to 14.0x, driven by market recognition of the balance sheet rather than by operational improvement, and that expansion is named as the fragile leg. This is scenario arithmetic, not a discounted-cash-flow model.
• Not investment advice. Independent research, educational and informational only, never personalized. Nothing here is a recommendation to buy or sell any security.
• Version: 2026-08-04-full.