Anduril Industries Inc

Anduril is the defining "neo-prime" of the 2020s defence-industrial cycle: a vertically integrated, software-first weapons manufacturer that funds product development with venture capital and sells finished commercial items to governments, inverting the cost-plus, requirement-first model of the incumbent primes. Its strategic asset is Lattice, an open-architecture autonomy and command-and-control layer that acts as the connective tissue across an unusually broad hardware portfolio — counter-UAS interceptors, loitering munitions, cruise missiles, collaborative combat aircraft, extra-large autonomous submarines, solid rocket motors, surveillance towers and soldier-worn mixed reality. The March 2026 U.S. Army enterprise agreement, with a ceiling of up to USD 20 billion over ten years, converted Anduril from a project vendor into an enterprise-level supplier and is the single most consequential commercial event in its history. Revenue more than doubled to USD 2.2 billion in 2025 and is guided to roughly USD 4.3 billion in 2026, against a projected operating loss of about USD 1.2 billion.

Company's own characterisation

Anduril describes itself in its standard corporate boilerplate (reproduced consistently across job postings, press releases and its website) as "a defense technology company with a mission to transform U.S. and allied military capabilities with advanced technology," which by "bringing the expertise, technology, and business model of the 21st century's most innovative companies to the defense industry … is changing how military systems are designed, built and sold." It states that "Anduril's family of systems is powered by Lattice OS, an AI-powered operating system that turns thousands of data streams into a realtime, 3D command and control center," and commits to "bringing cutting-edge autonomy, AI, computer vision, sensor fusion, and networking technology to the military in months, not years."

Because Anduril publishes no annual report, this boilerplate — not an audited MD&A — is the closest available equivalent to a management description of the business.

Independent characterisation

Anduril is best understood as three interlocking businesses operating under one balance sheet:

(i) An autonomy software platform (Lattice). Lattice is a sensor-fusion, command-and-control and mission-autonomy layer that ingests feeds from radars, electro-optical/infrared sensors, towers, drones, satellites and third-party systems and renders a single machine-speed operating picture. It is deliberately hardware-agnostic and exposed through an SDK and a partner programme, so that third-party platforms can be commanded through it. Lattice is the strategic moat: once it is the command backbone for a mission set, Anduril hardware enjoys structural preference, and third-party hardware becomes a complement rather than a competitor. The March 2026 Army enterprise vehicle is explicitly built around Lattice as the counter-UAS command-and-control backbone, and Lattice was selected in November 2025 for the Army's IBCS-Manoeuvre (IBCS-M) programme, displacing Northrop Grumman's legacy Forward Area Air Defense Command and Control (FAAD C2) lineage in that mission area.

(ii) A high-rate, attritable hardware manufacturer. Anduril builds physical systems designed for volume and cost-per-effect rather than exquisite performance: Barracuda cruise missiles, Roadrunner reusable interceptors, Anvil kinetic interceptors, Bolt and Ghost quadcopters/VTOLs, Altius tube-launched loitering munitions, Fury/FQ-44 collaborative combat aircraft, Dive-LD/Dive-XL/Ghost Shark autonomous underwater vehicles, Sentry towers, Pulsar electronic warfare and Menace expeditionary C4. The Arsenal-1 factory in Pickaway County, Ohio and its "Arsenal OS" manufacturing software are the industrial expression of this thesis.

(iii) A vertically integrated component and energetics supplier. Through acquisitions Anduril has internalised solid rocket motors (Adranos → McHenry, Mississippi), cooled infrared cameras (American Infrared Solutions), ruggedised tactical edge compute and networking (Klas), radar and battle-management software (Numerica's Spyglass, Spark, Mimir), large autonomous aircraft aerostructures and composites (Blue Force Technologies), and space-domain-awareness sensing (ExoAnalytic Solutions, March 2026). This vertical integration is intended to remove the supply-chain bottlenecks that have historically constrained munitions surge capacity, and to capture margin that would otherwise sit with sub-tier suppliers. The Mississippi plant, brought online in August 2025, targets 6,000 tactical solid rocket motors per year by end-2026 and positions Anduril as the third domestic merchant SRM supplier alongside Northrop Grumman and L3Harris/Aerojet Rocketdyne (Sacra, 2026).

Revenue model

Anduril's defining commercial innovation is that it develops products on its own capital and sells them as commercial items at firm-fixed prices, rather than developing to government specification under cost-reimbursable contracts. CFO Babak Siavoshy has publicly described implementing "SaaS-like pricing" for Anduril products in explicit contrast to traditional cost-plus contracting (Ramp interview, April 2026). This has three consequences: Anduril retains intellectual property and can resell a product across many customers; it absorbs development risk on its own balance sheet (hence the projected ~USD 1.2bn 2026 operating loss); and it can be procured through commercial-item pathways such as the Army Enterprise Agreement, other transaction authorities (OTAs) and IDIQ vehicles, which compress award timelines dramatically.

Revenue mix (product vs. service vs. subscription vs. licensing) is not publicly disclosed. From the structure of publicly announced awards, the composition is dominated by (a) hardware production and delivery orders, (b) multi-year software licensing/subscription for Lattice, and (c) installation, integration and sustainment services (e.g. the Marine Corps I-CsUAS award explicitly covers "deliver, install and sustain").

Value-chain position, customers and end-markets

Anduril occupies a hybrid position: it is simultaneously a prime contractor (Marine Corps I-CsUAS, Army SBMC, Air Force CCA, CBP surveillance towers, Australian Ghost Shark), a major subcontractor / hardware systems architect (Army TITAN, where Palantir USG is programme team lead and Anduril is lead hardware systems architect), and a merchant component supplier (solid rocket motors).

End-markets served: counter-UAS and short-range air defence; integrated air and missile defence / homeland missile defence (Golden Dome); collaborative combat aircraft and autonomous air systems; undersea autonomy and seabed warfare; long-range precision fires; electronic warfare; border and installation security; space domain awareness; soldier-worn mixed reality and mission command; solid rocket propulsion.

Stated strategic themes

Anduril has no investor day and publishes no strategy document. Its stated strategy must be reconstructed from CEO and executive statements. The recurring themes are:

(1) Fund development with private capital; sell finished commercial products. CEO Brian Schimpf's Series H investor note (May 2026) stated Anduril would invest "aggressively" in manufacturing capacity, research and development, and infrastructure. Anduril's model, as characterised by defence analysts, is to front its own capital and offer the Department of War capability with reduced programme risk — the inverse of the incumbent posture of declining to invest without a guaranteed order.

(2) Hyperscale manufacturing as the core differentiator. Anduril's framing of Arsenal-1: "By harnessing a world-class workforce and a scalable, software-driven approach to manufacturing, Arsenal-1 will set the standard for how we respond to the challenges of the future fight" (Schimpf, January 2025). Arsenal-1 is designed for "maximum flexibility to reallocate the most critical manufacturing resources — people, capital, machines, and materials — to meet new requirements, launch new products, or scale production to meet surges in demand, indefinitely."

(3) Lattice as the horizontal layer. Per Fortune's May 2026 profile, "Lattice, the data and software product, is the spine, a platform that third parties (including militaries and other companies) can build on," and it "now has long-term budgetary commitments from the U.S. Army and is moving from experimentation into rollout."

(4) Affordable mass over exquisite platforms. Schimpf to Fortune: "If the Ukraine war were to start a year from now, the first thing we'd be sending is Barracuda."

(5) Become a full-spectrum prime. Luckey has stated the ambition to compete with Lockheed Martin, Boeing and Kratos at similar scale, asserting "You have to fight and win across multiple areas."

Announced strategic initiatives, last 24 months

Sustainability / ESG commitments

Anduril publishes no ESG report, no emissions inventory, no science-based target and no diversity disclosure. No sustainability-linked initiatives have been announced.

Management's medium-term financial targets

No medium-term guidance beyond FY2026 has been issued. There is no backlog disclosure, no book-to-bill and no multi-year revenue target.

Lattice (software platform)

Air Systems

Underwater Systems

Force Protection, Sensing and Electronic Warfare

Warfighter Systems

Propulsion and Manufacturing Platform

Framing: Anduril publishes no financial statements. The only company-confirmed figures in the public domain are 2025 revenue (USD 2.2bn, disclosed 13 May 2026 in the Series H announcement and confirmed to Fortune), 2026 revenue guidance (USD 4.3bn, confirmed by the company to Fortune, May 2026) and a projected 2026 operating loss of approximately USD 1.2bn (reported by Sacra and corroborated across secondary coverage; not directly confirmed by the company). Prior-year revenue figures rest on founder statements to the press (Luckey, January 2025: "We doubled last year to about a billion. That's the rumor") and on aggregator reconstruction. Everything else is unavailable.

Income statement (USD m unless stated)

Notes: FY2021 revenue of USD 150m is a press-reported figure (Value Add VC, Aug 2026, drawing on prior reporting). FY2022 is not publicly disclosed — no credible primary or press figure exists. FY2023 (~USD 500m) and FY2024 (~USD 1,000m) are press-reported/founder-indicated, not audited. FY2025 (USD 2,200m) is company-confirmed. Blank cells are not publicly disclosed. Dividends per share are zero because no dividend has ever been declared and none is expected while the company is private and loss-making. Revenue CAGR FY2021→FY2025 on the reported figures is approximately 95% per annum; this should be treated as indicative given the unaudited inputs.

Forward-year figures (FY2026, company-guided / reported)

Revenue guidance of USD 4.3bn was confirmed by Anduril to Fortune (May 2026). The ~USD 1.2bn projected operating loss is reported by Sacra and repeated widely; it is not a company-confirmed figure and should be treated as an estimate. Morgan Stanley's January 2026 defence-tech forecast of ~USD 4.5bn for 2026 is broadly consistent with company guidance.

Balance sheet

All cells: not publicly disclosed. What can be stated with confidence is that Anduril's capital structure is overwhelmingly equity-funded: approximately USD 11.3bn–11.9bn of cumulative primary equity across 8–9 rounds, with no publicly disclosed debt facilities, bond issuance or credit ratings. The company has never announced a leveraged financing. Cumulative announced capital-expenditure commitments — approximately USD 1bn at Arsenal-1, approximately USD 1bn at Long Beach/Lakewood, USD 75m at McHenry, plus the Sydney XL-AUV plant — total well over USD 2bn and are being funded from equity.

Cash flow

Operating cash flow, capex and FCF: not publicly disclosed. Anduril has conducted employee tender offers / secondary liquidity events in connection with primary rounds (widely reported but with undisclosed size); these are not buybacks in the public-company sense.

Ratios

All ratios: not computable — required inputs are not publicly disclosed. Any published Anduril ROE, ROIC or leverage ratio is modelled, not observed.

Commentary on trends, inflections and drivers

Growth inflection. Anduril has compounded revenue at roughly 95% per annum on the reported figures, with the sharpest inflection between 2024 and 2025 (USD 1.0bn → USD 2.2bn, +120%). Three drivers account for most of this: the maturation of the counter-UAS franchise into programmes of record (USMC I-CsUAS USD 642m IDIQ, MADIS CES USD 200m), the ramp of CBP tower sustainment and expansion, and the arrival of large allied programmes, above all Ghost Shark. The step to a guided USD 4.3bn in 2026 depends less on new customer acquisition than on conversion: Arsenal-1 throughput, Barracuda deliveries beginning in 2027 (i.e. bookings rather than 2026 revenue), CCA production, and task-order draw against the Army enterprise vehicle. The 31 August 2026 TITAN order (USD 65m) is the first visible large draw under that vehicle and is a useful leading indicator of the conversion rate.

Margin structure and the loss. The reported ~USD 1.2bn 2026 operating loss is the mechanical consequence of Anduril's business model: it capitalises very little and expenses heavily ahead of revenue recognition on multi-year production contracts, while simultaneously funding two ~USD 1bn campuses from equity. This is a deliberate, financed choice rather than a distress signal — but it means the company's economics cannot be assessed on conventional defence-prime metrics. The key unobservable is gross margin on hardware at rate. Anduril's entire thesis rests on the claim that software-defined manufacturing plus commercial-item fixed pricing yields structurally better gross margins than cost-plus at similar volumes. There is no public evidence either way.

Inflection to watch. The transition from "development-funded-by-equity" to "production-funded-by-orders" is the single most important financial inflection ahead. It is expected to become visible in 2027, when Barracuda-500M, Barracuda-500 (FAMM), FQ-44A and Ghost Shark all reach meaningful delivery volumes simultaneously. Anduril executives have indicated profitability is not expected until around 2030 (per prediction-market and analyst summaries, June 2026); this timeline has not been confirmed by the company.

Management guidance

No guidance beyond FY2026 has been issued, and no backlog figure has ever been disclosed.

Consensus growth expectations

There is no sell-side consensus. The available third-party estimate — Morgan Stanley's January 2026 defence-tech forecast of ~USD 4.5bn for 2026 — sits marginally above company guidance. Implied growth on guidance is ~95% in 2026 following ~120% in 2025 and ~100% in 2024. Any 2027+ estimate in circulation is modelled from contract ceilings rather than disclosed backlog and should be discounted heavily.

Bull case

1. The Army enterprise vehicle converts, and Anduril becomes structurally embedded rather than competitively re-tested. The March 2026 agreement is not just a large number — it is a change in how Anduril sells. Federal agencies can now draw on Anduril's stack without separate acquisition reviews. The USD 65m TITAN order of 31 August 2026 is the first visible large draw. If task-order velocity scales even modestly against the USD 20bn ceiling, revenue visibility improves in a way no other non-traditional defence company enjoys, and Anduril's cost of capture — historically the largest structural disadvantage a newcomer faces against incumbents — collapses.

2. The budget cycle is the most favourable in forty years for exactly Anduril's portfolio. The FY2027 request proposes tripling drone and related-technology spending to more than USD 74bn and over USD 30bn for critical munitions, within a USD 1.5 trillion defence budget, with USD 17.5bn earmarked for Golden Dome. The Iran war of February–April 2026 was a live proof of Anduril's central commercial argument: intercepting cheap mass with expensive interceptors is economically unsustainable, and stockpiles are now depleted. Roadrunner's reusability and Barracuda's ~70% commodity content are the direct answers.

3. The two largest programme risks have already retired ahead of schedule. The CCA down-select — the single greatest threat to Arsenal-1's economics — resolved in Anduril's favour on 17 June 2026, four months early, with both vendors moving to production. Three weeks later the FQ-44A executed the first live air-to-air missile shot by any U.S. CCA, retiring the "lethality unproven" objection ahead of its Q4 2026 target. Serial production started three months early, and the first production article shipped roughly five weeks after the award. On the two questions that mattered most — can it fly and fight, and can it be built at rate — the evidence in 2026 has been affirmative and early.

Bear case

1. The company is priced for flawless execution and has a documented record of imperfect execution. At USD 61bn on guided FY2026 revenue of USD 4.3bn, Anduril trades at roughly 14x forward sales — roughly eight times Lockheed's revenue multiple — while losing an estimated USD 1.2bn a year, with profitability reportedly not before ~2030. Against that, the November 2025 WSJ and Reuters investigations documented failures across four separate systems, and the only sustained combat deployment (Altius in Ukraine) reportedly ended when the SBU stopped fielding it in 2024 over electronic-warfare vulnerability. A reported ~USD 100bn mark would imply ~23x forward sales. There is no margin for a missed milestone.

2. The USD 20bn is a ceiling, and the largest visible draws total roughly USD 152m. The Army has been explicit that the figure is maximum potential value, not an obligated amount, and the vehicle runs to 2036. Anduril discloses no backlog, no funded backlog and no book-to-bill, so there is no way for an outside party to verify conversion. Meanwhile customer concentration is rising: the enterprise agreement makes Anduril more dependent on a single service, not less, and it has essentially no commercial revenue to cushion an appropriations reversal. The FY2027 USD 1.5 trillion budget is a request, not law.

3. EagleEye/SBMC is the tell, and it is slipping. Anduril inherited a programme that consumed ~USD 1.8bn and was cancelled. Approximately 100 prototypes were expected for soldier evaluation in April–June 2026; as of July 2026 no independent evaluation had been published, and in August 2026 Anduril clarified that the headsets made available to the Army were company-owned and not formal contractual deliveries — a materially weaker statement than a delivery. The downselect has moved to FY2028 against a better-funded rival (Rivet, USD 195m vs USD 159m). If Anduril cannot execute a soldier-worn programme against a competitor of comparable vintage, the thesis that its speed advantage generalises across all seven domains it has entered comes into question.

Catalysts and monitorables — next 12 months

Analyst verdict

Anduril has, in the space of eighteen months, completed the transition from promising challenger to structurally embedded supplier — and it did so faster than either its advocates or its critics expected. The March 2026 Army enterprise agreement is the fulcrum. It is not merely the largest contract vehicle ever granted to a non-traditional defence contractor; it changes the unit of competition from the individual programme to the enterprise relationship, and in doing so neutralises the capture-cost disadvantage that has historically killed defence entrants. Layered on top of that, the June and July 2026 CCA sequence — dual production award four months early, first CCA live air-to-air shot three weeks later, first production airframe five weeks after that — is the most convincing rebuttal available to the November 2025 reliability narrative. Anduril did not argue its way out of that coverage; it built its way out.

The bear case has not disappeared, however. It has simply migrated. The concern is no longer whether Anduril can build things; it is whether the valuation has outrun what is verifiable. At USD 61bn the company trades at roughly 14x forward revenue while losing an estimated USD 1.2bn a year, with no audited financials, no disclosed backlog, no book-to-bill and profitability reportedly seven fiscal years away. The USD 20bn is a ceiling against which roughly USD 152m of task orders are publicly identifiable. And EagleEye — the programme with the highest strategic visibility and the worst inherited history — is quietly slipping, with a company clarification in August 2026 that its prototypes were not contractual deliveries at all.

The honest position is this: the operational thesis is now substantially proven, the financial thesis is entirely unverifiable, and the gap between the two is precisely what an investor is being asked to underwrite. An S-1 would close that gap. Until one is filed, every valuation on this company is an act of faith in a management team that has, so far, earned it.

Prepared 5 September 2026. All figures stated in U.S. dollars unless otherwise noted. Fiscal years are calendar years. Data points marked "not publicly disclosed" reflect the absence of any primary or credible secondary source and have deliberately not been estimated. Where sources conflict — notably on employee headcount (7,000 per the company/State of California in January 2026 versus ~8,560 per Revelio Labs in March 2026), on total capital raised (USD 11.3bn per Tracxn versus USD 11.9bn per PitchBook), and on Series G/H valuations (USD 30.5bn/USD 61bn per company and primary reporting versus USD 28bn/USD 65bn per at least one secondary-market data provider) — both figures and the discrepancy have been noted in the relevant section.

--