Preparing your space
Your library and workspace are being arranged now.
Preparing your space
Your library and workspace are being arranged now.
Loading the business overview…
Explore the research to see how business evidence informs assumptions and valuation.
Start with the business evidence, then follow the assumptions through to scenarios and valuation.
Valuation and assumptionsWritten at USD 182.5600; now USD 167.23000000 (-8%). Upside to the weighted case has moved from -47% to -42%.
Palantir is an exceptionally high-quality, capital-light enterprise software compounder whose economics are primarily driven by the duration of its current hypergrowth, its terminal operating margin, and the discount rate applied to a very long-duration cash flow stream. Its principal competitive advantage is a combination of genuine per-account switching costs created by the ontology its customers build their workflows on, and two decades of classified and sovereign deployment accreditation that protects roughly half of revenue - evidenced by 93% revenue growth, a 46.8% GAAP operating margin achieved after expensing stock compensation, 34% ROIC, and capital expenditure of 0.6% of revenue. Its most important vulnerabilities are that over 81% of revenue is United States revenue exposed to a single political and appropriations cycle, that total contract value is growing at 49% against revenue growth of 93%, that reported earnings are flattered by a fading tax benefit, and that the competitor set comprises the best-capitalised technology companies in the world, any of which could bundle a competing orchestration layer. Under the base case, six more years of 30%-plus growth taking revenue to roughly $52bn by 2035 at a terminal 60% GAAP operating margin imply intrinsic value of approximately $85 per share versus a market price of $182.56. The investment therefore depends primarily on whether the current growth rate persists for roughly ten more years rather than six - a question for which no historical base rate exists, because reaccelerating growth at this scale is something enterprise software essentially has not done before. The most important evidence to monitor is total contract value and remaining deal value relative to revenue growth, US commercial revenue growth, GAAP rather than adjusted operating margin, and whether international revenue reaccelerates.
| Field | Value |
|---|---|
| Company | Palantir Technologies Inc. |
| Ticker / exchange | PLTR / NASDAQ |
| Current share price | $182.56 (2026-09-01) [S6]; sources range $180–$186 |
| Market capitalization | ~$469bn on 2,569.8M diluted shares [V]; ~$438–448bn on ~2.40bn basic shares [S6][S7] |
| Enterprise value | ~$460bn [Est., net cash ~$8.5bn Unverified] |
| Industry | Enterprise software / data and AI platforms |
| Primary geographies | United States (over 81% of revenue) [S5]; UK, Europe, Asia-Pacific |
| Investment classification | Compounder |
| Business quality | High |
| Competitive advantage | Strong |
| Financial quality | Strong |
| Growth outlook | Strong |
| Balance-sheet risk | Low |
| Valuation attractiveness | Unattractive |
| Overall risk | High |
| Base-case intrinsic value | $85 |
| Bull-case intrinsic value | $190 |
| Bear-case intrinsic value | $35 |
| Probability-weighted value | $96.25 [Est.] |
| Upside/downside vs price | −47.3% |
| Expected annualized return, base case (3–5yr) | −8% to −2% [Est.] |
| Investment stance | Avoid |
The stance is about price, not about the company. On the operating evidence, Palantir is the highest-quality business I have assessed in this series: 93% revenue growth [S1], a 47% GAAP operating margin in the first half [V, derived], near-zero capital intensity, and no debt. Everything below that says "Avoid" is a statement about $182.56, not about the business. Read §28 before acting on it.
Why it is attractive. The operating results are close to unprecedented at this scale. Revenue grew 93% year-over-year to $1.94bn in Q2 with US commercial up 149% and US government up 90% [S1][S2]. Management raised full-year revenue guidance from $7.65–7.66bn to $8.15–8.16bn [S3] — a ten-point increase in the growth rate mid-year. Verified GAAP operating income of $1.67bn on $3.57bn of half-year revenue [V] is a 47% GAAP margin, achieved after expensing stock compensation. Capital expenditure was $22.0M against that revenue [V] — a capital intensity of 0.6%.
What the market may be misunderstanding — or what I may be. Nothing about the near term. My revenue and margin forecasts for 2026 and 2027 are broadly at or above consensus. The disagreement is entirely about duration and terminal value. My reverse-DCF (§22) says the current price requires Palantir to generate roughly $60bn of annual free cash flow at maturity — comparable to what Meta generates today — on a business that will do $8.15bn of revenue this year.
The most important value driver. Not the growth rate this year, which is known. How many more years the 50%+ growth lasts. Each additional two years of hypergrowth is worth roughly $45 per share [Est., §17]; the CEO's own framing is "at least another 18 months" [S2], which is a shorter horizon than the price implies.
The largest risk. That the current acceleration is a budget-cycle phenomenon rather than a permanent share shift. US commercial revenue has compounded 380% since 2024 [S2]; enterprises are buying AI platforms in a once-in-a-decade procurement wave. Wave revenue and run-rate revenue look identical for about three years.
What must be true. That Palantir's ontology creates switching costs strong enough to hold the installed base once hyperscalers and foundation labs offer competing enterprise AI orchestration; that GAAP operating margin reaches and holds near 60%; and that US commercial growth stays above 50% into 2029. All three, not one.
I want to justify not saying "Wide" for a company growing 93% with 47% GAAP margins.
Two decades of clearances, accreditations and classified-environment engineering. Apollo's cloud-agnostic deployment [S10] is the technical embodiment. This is the most conventional moat asset Palantir owns and it protects roughly half of revenue.
A genuine organisational capability that competitors have explicitly struggled to copy, because it requires hiring engineers to sit in customer sites rather than shipping software. But it is a services advantage, and services advantages scale sub-linearly.
Palantir has extraordinary mindshare. It also carries political baggage from defence and immigration work that closes some doors, particularly in Europe.
This matters. My deployment of Foundry does not improve yours. There is no data network effect across customers — indeed Palantir's core promise is the opposite, that customer data stays isolated [S1]. Palantir has deliberately given up the mechanism that produces the widest software moats, in exchange for trust. That is probably the right trade, but it caps the moat.
R&D at 9.9% of revenue, down from 18.2% two years ago [V], is scale economics working. But the sales and deployment cost of each new customer does not fall the same way.
| Test | Result |
|---|---|
| ROIC above WACC | Yes, enormously — see §11 |
| Rising market share | Yes — 149% US commercial growth [S1] |
| Pricing power | Probable but not directly disclosed. Revenue per customer rising sharply [Est.] is consistent with pricing power or with volume expansion; Palantir does not separate them. |
| Gross margin | ~85% [S7] |
| Incremental margin | Exceptional — FY2024→FY2025 revenue +$1.61bn produced operating income +$1.10bn [V], a 68% incremental margin |
| Retention / churn | Not disclosed. No net revenue retention figure in my sources. Insufficient evidence to conclude. |
| Customer concentration | Not disclosed [Unverified] |
The missing row matters. For a company whose entire moat thesis rests on switching costs, the absence of a disclosed net revenue retention figure is a real gap. Most enterprise software companies report it. Its absence means the switching-cost claim rests on inference from expansion growth rather than on a stated metric.
| Risk | Type | Prob. | Financial impact | Horizon | Leading indicator | Mitigating factors | Priced in? |
|---|---|---|---|---|---|---|---|
| Growth decelerates faster than the price implies | Business | 60% | −$60 to −$100/share | 1–4yr | TCV vs revenue growth gap; RDV | Backlog; expansion motion | No — this is the core disagreement |
| Multiple compression at a ~104× forward P/E | Valuation | 65% | −$60/share | 1–3yr | Peer software multiples | Growth delivery | Partially |
| Hyperscalers bundle competing orchestration | Business | 40% | −$25 to −$40/share | 2–5yr | Competitive win rates; deal cycles | Ontology switching costs; sovereign deployment | No |
| US government appropriations disruption | Business | 30% | −$15 to −$25/share | 1–3yr | Budget process; contract awards | Multi-year contracts; accreditation moat | Partially |
| Enterprise AI budget normalisation | Business | 45% | −$30 to −$50/share | 2–4yr | US commercial growth rate | Mission-critical deployments | No |
| Political exposure damages access | Business | 25% | −$10 to −$20/share | 1–3yr | Contract awards; press cycle | Bipartisan outreach [S6] | No |
| Foundation labs move up-stack into deployment | Business | 30% | −$25 to −$45/share | 3–6yr | Product announcements from labs | Model-agnostic positioning; governance | No |
| Effective tax rate normalises upward | Financial | 70% | −$5 to −$8/share | 1–2yr | Tax note in filings | None — arithmetic | No |
| Founder voting control limits shareholder recourse | Governance | 100% (structural) | Embedded | Permanent | — | None | Yes |
| Key-person dependence on Karp | Business | 10% | Hard to quantify | Any | Public statements | Deep bench (Sankar, Taylor) | No |
| Balance-sheet distress | Financial | <1% | Immaterial | — | — | No debt, $8.5bn liquidity | Yes |
The three types matter unusually much here. Business risk is moderate and mostly about growth duration rather than about failure. Financial risk is close to zero — there is no debt, no covenant, no funding need. Valuation risk is overwhelming and is essentially the entire risk of owning the stock at $182.56.
| Assumption | Value | Basis |
|---|---|---|
| Risk-free rate | 4.3% | [Unverified — could not confirm the current 10yr Treasury yield] |
| Beta used | 1.35 | [Est.] — reported beta is 1.62 [S7] or 2.49 [S11]. Both are measured over a period of extreme multiple volatility and overstate business risk; Palantir's operating results are among the most stable in the market. I use 1.35 and sensitize widely. |
| Equity risk premium | 5.0% | [Est.] |
| Cost of equity | 11.05% | Derived |
| Cost of debt / debt weight | n/a | Zero debt [V] |
| WACC | 11.0% | Equals cost of equity |
| Terminal growth | 3.5% | [Est.] — above nominal GDP, reflecting software's continued share gain |
| Tax rate | 21% | Statutory. Palantir currently pays less (§10), but NOLs fade. |
| Diluted share count | 2,569.8M | [V], as of 2026-06-30 |
| Share count assumption | Flat — SBC dilution has fallen to 0.2% per half-year [V] and is offset in the margin | Deliberate |
| Stock-based compensation | Fully expensed. The verified operating income of $1.67bn [V] is GAAP and already deducts SBC (~$470M in H1 [Est.]). I never use adjusted operating income. | Deliberate |
| Capex | ~0.5% of revenue | $22.0M on $3.57bn [V] |
The SBC treatment is the single most important comparability note. Management's headline profitability measure is adjusted income from operations, which excludes stock compensation and related payroll taxes [S1]. A DCF built on that measure would carry roughly 13 points more margin than mine and would produce a value materially higher. If another valuation of PLTR differs from this one, check this first.
Growth vs reinvestment vs incremental ROIC — consistent, and unusually so. Palantir's growth genuinely does not consume capital: capex was $22.0M against $3.57bn of half-year revenue [V], and incremental operating margin was 68% from FY2024 to FY2025 [V]. I am not assuming growth arrives while reinvestment falls; reinvestment was never the constraint. The reinvestment here is R&D and sales headcount, both expensed, both already inside my margin path. This is the rare case where high growth and low capital intensity are genuinely compatible rather than an analytical sleight of hand.
Pricing vs volume vs share — consistent but under-evidenced. I assume revenue per customer keeps rising while customer count keeps growing. That is what the data shows [S3], and unlike a consumer business there is no obvious volume/price tension in enterprise software where each deployment expands. The strain is that I cannot verify it: Palantir discloses neither pricing nor net revenue retention, so my expansion assumption rests on inference.
Margins vs competitive intensity vs required investment — this is where my forecast is strained, and I want to name it. I hold terminal GAAP operating margin at 60%, materially above today's 46.8%, while simultaneously assuming that hyperscalers and foundation labs intensify competition. Those two are in tension. A 60% operating margin implies R&D and sales fall further as a share of revenue; heavier competition implies the opposite. My defence is that the margin expansion comes from scale on a fixed platform investment rather than from underinvesting, and that R&D dollars can grow while the ratio falls. But if competition forces sales and marketing spend up, my terminal margin is too high, and that is a more likely direction of error than my growth path.
Capex vs depreciation vs free cash flow — consistent and trivially so. Capex of $22.0M and D&A of $15.0M in the half [V] are both immaterial. There is no deferred cost landing in a future P&L. This is the cleanest capex-to-FCF relationship in this research series and it is a genuine mark of business quality.
Terminal growth vs terminal reinvestment vs terminal ROIC — consistent. 3.5% terminal growth on a business with near-zero capital requirements implies a very high terminal ROIC, which is exactly what Palantir has (§11).
All figures $bn. GAAP operating margin throughout — SBC expensed.
| Year | Revenue | Growth | Op. margin | Op. income | NOPAT (21%) | FCF | PV @11% |
|---|---|---|---|---|---|---|---|
| 2026E | 8.15 | +82% | 50% | 4.08 | 3.22 | 3.22 | 2.90 |
| 2027E | 12.6 | +55% | 53% | 6.68 | 5.28 | 5.28 | 4.29 |
| 2028E | 17.7 | +40% | 55% | 9.74 | 7.69 | 7.69 | 5.62 |
| 2029E | 23.0 | +30% | 57% | 13.11 | 10.36 | 10.36 | 6.82 |
| 2030E | 28.5 | +24% | 58% | 16.53 | 13.06 | 13.06 | 7.75 |
| 2031E | 33.9 | +19% | 59% | 20.00 | 15.80 | 15.80 | 8.45 |
| 2032E | 39.0 | +15% | 60% | 23.40 | 18.49 | 18.49 | 8.91 |
| 2033E | 43.7 | +12% | 60% | 26.22 | 20.71 | 20.71 | 8.98 |
| 2034E | 48.1 | +10% | 60% | 28.86 | 22.80 | 22.80 | 8.91 |
| 2035E | 51.9 | +8% | 60% | 31.14 | 24.60 | 24.60 | 8.66 |
I round the base case to $85, adding roughly $7 for the international/sovereign optionality (§15, item 4) which I did not model explicitly and which is real but unevidenced.
Terminal value is 63% of the total — high, but lower than it would be for most growth companies because the explicit period generates so much cash.
Note what this forecast contains: revenue compounding to $51.9bn by 2035, a 23% CAGR over nine years following an 82% year, and a 60% terminal operating margin higher than almost any software company has sustained. This is not a conservative forecast. It is a generous one, and it produces $78.
Detailed in §22. The current price requires roughly 2.4× my present value — approximately $110bn of 2035 revenue, or $60bn of terminal free cash flow, or a WACC near 8% with substantially higher revenue and margins.
| Multiple | PLTR @ $182.56 | Comment |
|---|---|---|
| P/E (TTM) | 153× [S7] | On earnings flattered by below-statutory tax |
| Forward P/E (NTM) | 104.5× [S7] | The cleanest headline figure |
| P/E on normalized owner earnings (~$1.25) | 146× [Est.] | The most economically meaningful |
| EV / FY2026E revenue | ~56× [Est.] | Among the highest in large-cap software history |
| EV / EBITDA (TTM $2.663bn) | ~173× [S7, Est.] | |
| Price / FCF (FY2026E ~$5bn) | ~94× [Est.] | |
| FCF yield | ~1.1% [Est.] | Against a 4.3% risk-free rate |
| Gross margin (TTM) | 84.8% [S7] | Justifies a premium |
| ROE (TTM) | 38.4% [S7] | Justifies a premium |
Should Palantir trade at a premium? Emphatically yes. 93% growth, 47% GAAP margins, zero capital intensity, no debt, and 34% ROIC deserve a multiple far above the software median. A 40–50× forward earnings multiple would be defensible and historically unusual. 104× is a different proposition, and at a 1.1% free cash flow yield against a 4.3% risk-free rate, the equity offers no current yield and requires roughly a decade of delivery to earn its cost of capital.
Earnings power value (normalized owner earnings of $3.2bn capitalized at 11%, no growth): $29bn = $11.30 per share. So approximately 94% of the current price is growth. For context, that figure was 65% for SpaceX and roughly two-thirds for Netflix in the earlier reports in this series. Palantir is the purest growth-dependent valuation I have assessed.
Replacement value: not meaningful — the asset is code, an installed base and clearances, none of which has a replacement cost.
Sum-of-the-parts: not meaningful — one product family, one architecture.
Each scenario is stated as a causal chain: trigger → operating mechanism → financial consequence → valuation consequence.
Trigger: The enterprise AI procurement wave normalises during 2027–28. Corporate AI budgets stop growing at triple-digit rates and settle into ordinary IT budget cycles.
Operating mechanism: The gap already visible between bookings and revenue widens — total contract value grew 49% while revenue grew 93% [S3], and revenue converges down toward bookings rather than bookings up toward revenue. New logo acquisition slows first; expansion within existing accounts slows a year later as the initial use cases are exhausted and incremental ones have lower ROI. Simultaneously, hyperscalers bundle adequate AI orchestration into existing cloud agreements, so Palantir's incremental deal must now be justified against a nearly-free alternative rather than against nothing. Sales and marketing spend rises to defend win rates.
Financial consequence: revenue growth falls from 82% (2026) to ~35% (2027) to ~20% (2028) and into the low teens by 2030, reaching roughly $18bn. GAAP operating margin plateaus near 45% rather than expanding, because S&M rises as R&D falls. Free cash flow reaches roughly $6bn and grows slowly.
Valuation consequence: terminal FCF around $8bn, WACC 12.5% on a de-rated growth story, terminal growth 3.0%. Intrinsic value: $35 — still a very good company, at a multiple appropriate to a 15%-growth software business.
Trigger: Palantir's product advantage is real and durable, but the extraordinary growth rate is partly a category-formation phenomenon that decays on a normal curve rather than a cliff.
Operating mechanism: The installed base keeps expanding — the ontology creates genuine switching costs, and customers who have production workflows running add use cases each year. New logos continue at a healthy but decelerating rate. US commercial growth falls from 149% to roughly 80%, 50%, 35% and 25% over four years as the base grows. Government grows steadily at 25–40%, protected by accreditation. International finally becomes material but does not become a second US. R&D stays near 10% of revenue and S&M grows in line with revenue, so operating margin expands to roughly 60% by 2032 and holds.
Financial consequence: revenue of $12.6bn (2027), $23.0bn (2029), $33.9bn (2031), reaching $51.9bn by 2035 — a 23% CAGR over nine years. GAAP operating margin 50% rising to 60%. Free cash flow $3.2bn (2026) rising to $24.6bn (2035). No capital consumed; cash accumulates toward $50bn+ unless returned.
Valuation consequence: WACC 11.0%, terminal growth 3.5%, terminal FCF $25.5bn. Intrinsic value: $85.
Trigger: "AI sovereignty" [S1] is a genuine global structural shift, not a US corporate budget cycle — and Palantir's model-agnostic, deployable-anywhere architecture makes it the default standard for governed enterprise and government AI worldwide.
Operating mechanism: International and sovereign demand, currently under 19% of revenue [S5], reaccelerates to match US growth as governments and regulated industries outside the US conclude they cannot route proprietary data through US frontier labs. The NVIDIA collaboration for sovereign Nemotron deployment [S10] becomes a repeatable channel. Simultaneously, agentic workflows arrive: because AIP turns processes into tools for LLM-driven agents [S10], each deployment's value scales with agent adoption rather than seat count, and revenue per customer compounds faster than headcount at the customer. Competitors fail to replicate the classified and air-gapped deployment capability, so the government moat holds and extends internationally. Operating margin reaches 62% because R&D stays flat in dollars while revenue compounds.
Financial consequence: revenue growth stays above 40% through 2029 and above 25% through 2033, reaching roughly $95bn by 2035. Terminal FCF approaches $48bn. The equity re-rates as duration risk falls — the market becomes confident growth persists, and the discount rate compresses toward 9.5%.
Valuation consequence: WACC 9.5%, terminal growth 4.0%. Intrinsic value: $190 — modestly above the current price and near the 52-week high of $207.52 [S6].
| Scenario | Value | Probability | Weighted |
|---|---|---|---|
| Bear | $35 | 0.30 | $10.50 |
| Base | $85 | 0.45 | $38.25 |
| Bull | $190 | 0.25 | $47.50 |
| Total | 1.00 | $96.25 |
These sum to 1.00 and I am content that they should. Unlike the SpaceX report in this series, where a genuine unvaluable tail justified leaving probability unassigned, Palantir's outcome space is well spanned by these three: the business either keeps compounding fast, decelerates on a normal curve, or normalises early. There is no fourth outcome I can identify that the three do not cover.
Versus $182.56, the weighted value implies −47.3%.
Note the shape. The bull case contributes 49% of the weighted value from 25% probability. If you believe the bull case is 45% likely rather than 25%, the weighted value rises to roughly $124; at 60% it reaches $145. Even at a 60% bull probability the weighted value stays below the current price — which is a meaningfully different situation from the SpaceX and Micron reports, where a single probability shift flipped the stance. Here, the price is above even a heavily bull-weighted valuation, and that is the most robust finding in this report.
85% gross margin, 46.8% GAAP operating margin [V], 0.6% capital intensity, cash conversion above 100% of net income, no debt. Growth that requires no capital. The highest business-quality score in this research series, and the only point withheld is for the absence of network effects and undisclosed retention.
Genuine switching costs via the ontology, and a durable classified-deployment moat protecting roughly half of revenue. Marked down from 8 because there are no network effects — deliberately, since customer data isolation is the product promise — and because the competitor set is the best-capitalised group of companies in the world. Moderate moat.
An outstanding execution record: margin from −107% to +46.8% [V], guidance raised twice in a year and beaten substantially [S3]. Not repurchasing stock at 104× forward earnings is the right call. Marked down for promotional communication ("otherworldly", "Forget consensus"), founder voting control, and no record through a downturn.
Zero debt, ~$8.5bn liquidity, $4bn+ of annual free cash flow, and no scenario in this analysis where financing is a question. Marked down one point only because the marketable-securities balance is unverified.
93% growth, accelerating, with guidance raised ten points mid-year. The bookings/revenue gap is the only blemish.
104× forward earnings, 146× normalized owner earnings, 56× revenue, 1.1% free cash flow yield. 94% of the price is growth. The price requires roughly ten more years of 30%+ growth.
Bear −81%, base −53%, bull +4%. A bull case that barely exceeds the current price is the defining feature, and it is why the score is low despite the low probability of permanent capital loss.
1. Is this a good business? It is an exceptional one — the best in this research series on returns, capital intensity, margin trajectory and growth. Nothing in this report should be read as doubting that.
2. Stronger or weaker? Stronger, on every operating measure, in every period. Growth reaccelerating, margin tripling, dilution arrested, cash accumulating. The only metric moving the wrong way is bookings growth relative to revenue growth.
3. Durable competitive advantage? Yes — moderate. Real switching costs per account and a durable government accreditation moat. Not wide, because there are no network effects and the advantage protects the installed base rather than the next sale.
4. Three most important value drivers? (i) The duration of 50%+ revenue growth; (ii) terminal GAAP operating margin; (iii) the discount rate.
5. Three largest risks? (i) Growth decelerating faster than the price implies; (ii) multiple compression from 104× forward earnings; (iii) enterprise AI budget normalisation being mistaken for competitive loss, or vice versa.
6. Most important dependency? US government budgets and political alignment — $809M of quarterly revenue growing 90% [S1], inside a business where over 81% of revenue is US [S5].
7. What is the market pricing in? Roughly ten more years of 30%+ growth at a 60% operating margin, reaching approximately $110bn of revenue by 2035, or terminal free cash flow near $60bn — comparable to the largest cash generators in the world.
8. Where could the market be wrong? On duration, principally. Secondarily on the bookings/revenue gap (TCV +49% vs revenue +93%), which is the clearest available leading indicator and is currently being discounted as noise. And on reported EPS, which is flattered by a fading tax benefit.
9. Fair intrinsic value? $85 base case; $96.25 probability-weighted; an honest range of $35 to $190.
10. Margin of safety today? Deeply negative. The price is 115% above the base case and 90% above the weighted value.
11. Reasonable annualized return over 3–5 years? −5.3% weighted. Bear −19%, base −5%, bull +10%. Note that fundamentals alone contribute +37% a year in the base case; the negative return is entirely multiple compression.
12. Single development that would most strengthen the thesis? Total contract value growth falling below 30% for two consecutive quarters while revenue growth stays high — the signature of a backlog converting faster than it replenishes.
13. Single development that would most invalidate it? International revenue growth reaccelerating to 80%+ for two quarters, demonstrating that sovereign AI demand is a global structural shift rather than a US budget cycle. That is the bull case's mechanism and it would extend the growth duration my base case denies.
14. At what price does it become clearly attractive? $80 — below my base case and roughly 17% below the weighted value, which leaves room for my growth-duration assumption to be wrong by two full years and still not lose money. Note this implies a 56% decline; the 52-week low was $106.37 [S6], so it is outside the recent range.
15. What price implies excessive optimism? $190 — the bull case, requiring sovereign AI to be a decade-long global re-architecture with Palantir as the default standard.
Palantir is an exceptionally high-quality, capital-light enterprise software compounder whose economics are primarily driven by the duration of its current hypergrowth, its terminal operating margin, and the discount rate applied to a very long-duration cash flow stream. Its principal competitive advantage is a combination of genuine per-account switching costs created by the ontology its customers build their workflows on, and two decades of classified and sovereign deployment accreditation that protects roughly half of revenue — evidenced by 93% revenue growth, a 46.8% GAAP operating margin achieved after expensing stock compensation, 34% ROIC, and capital expenditure of 0.6% of revenue. Its most important vulnerabilities are that over 81% of revenue is United States revenue exposed to a single political and appropriations cycle, that total contract value is growing at 49% against revenue growth of 93%, that reported earnings are flattered by a fading tax benefit, and that the competitor set comprises the best-capitalised technology companies in the world, any of which could bundle a competing orchestration layer. Under the base case, six more years of 30%-plus growth taking revenue to roughly $52bn by 2035 at a terminal 60% GAAP operating margin imply intrinsic value of approximately $85 per share versus a market price of $182.56. The investment therefore depends primarily on whether the current growth rate persists for roughly ten more years rather than six — a question for which no historical base rate exists, because reaccelerating growth at this scale is something enterprise software essentially has not done before. The most important evidence to monitor is total contract value and remaining deal value relative to revenue growth, US commercial revenue growth, GAAP rather than adjusted operating margin, and whether international revenue reaccelerates.
Most of the above are readable in the Q2 2026 Form 10-Q, the FY2025 Form 10-K and the 2026 proxy. This report should be re-run against those documents before any capital is committed, and re-run again after Q3 2026 results in early November [S7], which will provide the next reading on the bookings-versus-revenue gap that §28 identifies as the key leading indicator.
The duration of hypergrowth is the weakest link and carries the conclusion; it is unsupported by any base rate because reaccelerating growth at this scale has essentially no precedent in enterprise software, and shifting from six to ten more years of 30%-plus growth would take the base case from $78 to $189 and flip the stance. Net revenue retention is not disclosed in any available source, which is conspicuous for a company whose moat thesis rests on switching costs and means that claim rests on inference rather than a stated metric. Stock-based compensation is explicitly not stored by the platform; the approximately $470M half-year figure used here is derived from the gap between GAAP and adjusted operating income and should be replaced with the filed number. Also unresolved: marketable securities and total liquidity, since only cash and equivalents of $2.06bn is verified; deferred revenue, receivables and remaining performance obligations; customer concentration, likely material given $809M of quarterly US government revenue; the full geographic revenue split beyond the over-81%-US figure from the earnings call; the effective tax rate and remaining loss carryforwards, which materially flatter reported EPS since half-year net income of $1.93bn exceeds operating income of $1.67bn; executive compensation and Class F share mechanics; employee headcount, where a third-party figure of about 4,430 implies $1.84M of revenue per employee; and competitor financials on a comparable basis. Note also that platform-verified figures dated 2026-06-30 are six-month year-to-date rather than quarterly, confirmed by summing the two reported quarters.
These model scenarios depend on their inputs. Review the assumptions and supporting evidence alongside the implied values.
earnings multiple v2 · as of 2026-09-12
Sign in for the price chart, where every input traces back to a filing, an adjustable scenario sandbox, and companies that move with PLTR.
Create a free account →