Investment Memo: Anduril Industries
Category: Defense technology / autonomous systems Framing: Hypothetical Series A evaluation (as if evaluating the company at its early stage, informed by what the model has since proven)
---
Thesis
Defense is one of the last trillion-dollar markets untouched by Silicon Valley product velocity. The five legacy primes (Lockheed, Raytheon/RTX, Northrop, Boeing, General Dynamics) operate on cost-plus contracts that reward slow, bespoke hardware development and punish reusable software. Anduril's bet is structural: build products on its own capital, sell them off-the-shelf under fixed-price contracts, and let software (Lattice OS) compound across every hardware line. If the Pentagon's procurement culture shifts even 5–10% toward commercial-style acquisition — driven by drone warfare in Ukraine, China deterrence, and attritable-mass doctrine — the first credible software-native prime captures disproportionate share. That's a fund-returner profile: a category-defining company in a market where a single program of record can be worth billions annually, with venture-style margins layered on defense-scale contracts. The founder story matters too: Palmer Luckey (Oculus exit) plus a Palantir-alumni founding team (Brian Schimpf, Trae Stephens, Matt Grimm) is the rare combination of hardware credibility, DoD go-to-market scar tissue, and Founders Fund backing willing to underwrite long procurement cycles.
Product & Wedge
Lattice OS is the core asset: an AI-driven command-and-control platform that fuses sensor data across assets, autonomously detects/classifies threats, and lets one operator manage many systems. Hardware is deliberately the trojan horse for the software.
The initial wedge was border security sentry towers (CBP contracts) — politically accessible, fast sales cycle, and a live proving ground for Lattice. From there the hardware portfolio expanded: Ghost (autonomous helicopter drone), Anvil (kinetic counter-drone interceptor), Altius loitering munitions (via Area-I acquisition), Dive-LD autonomous undersea vehicles (via Dive Technologies), Roadrunner (reusable jet-powered interceptor), Barracuda cruise missiles, and Fury, its entry in the Air Force's Collaborative Combat Aircraft (CCA) program. The Arsenal-1 hyperscale factory (Ohio) is the manufacturing thesis: mass-producible, attritable systems rather than exquisite platforms.
The wedge logic is sound: each hardware sale deepens Lattice lock-in, and Lattice becomes the connective tissue the DoD can't easily rip out.
Market & Competition
- TAM: US defense budget ~$850B+; addressable autonomy/C2/counter-drone segments plausibly $50–100B+ over a decade, plus allied markets (Australia, UK — both already customers).
- Legacy primes: Lockheed, RTX, Northrop, GD, Boeing — dominant incumbency, deep program-of-record entrenchment, but structurally weak at software.
- New-generation rivals: Palantir (software layer, sometimes partner sometimes competitor), Shield AI (autonomy/Hivemind, direct CCA-adjacent competitor), Skydio (drones), Epirus (counter-drone), SpaceX (procurement precedent and cultural forerunner), General Atomics (Anduril's head-to-head competitor on CCA — both were downselected).
- Key validation: Anduril won the CCA Increment 1 downselect (with GA) over the legacy primes, and took over the Army's IVAS program from Microsoft — evidence a startup-era company can win prime contracts, not just subcontract crumbs.
Traction & Business Signal (public only)
- Contract wins: CBP border towers; SOCOM counter-UAS program of record (reported ~$1B ceiling); Air Force CCA downselect; IVAS takeover; Australia Ghost Shark undersea program; various Ukraine deliveries.
- Reported revenue: press reports suggest ~$500M (2023) growing to ~$1B (2024). Actual margins, backlog composition, and cash burn: unknown.
- Financing: valuation stepped from ~$8.5B (2022) to $14B (2024) to a reported ~$28–30.5B (2025). Signals extraordinary investor conviction; also signals heavy ongoing capital consumption (hardware + Arsenal-1 buildout).
- Profitability: unknown, presumed negative given factory capex.
Risks (the three that actually kill it)
- Procurement reversion. The entire thesis depends on DoD continuing to shift toward fixed-price, commercially developed systems. Procurement reform has failed repeatedly for decades; a change in administration priorities, a budget crunch, or prime-lobbying backlash could strand Anduril in "impressive demos, no programs of record" purgatory. This is the existential risk — Anduril doesn't control its only customer's buying behavior.
- Capital intensity vs. hardware margins. Anduril self-funds R&D and now factories. If fixed-price contracts on complex hardware (CCA, IVAS — a program Microsoft lost money on) produce cost overruns, Anduril eats them. Venture returns require software-like economics eventually emerging; if it becomes a lower-margin hardware manufacturer, the current valuation multiple collapses.
- Concentration & political exposure. One customer (USG), long sales cycles, and program cancellation risk (IVAS's history is troubled). Plus founder-driven political salience cuts both ways across administrations. A single lost flagship program (e.g., CCA Increment 2 going elsewhere) materially reprices the story.
Verdict
Invest at a hypothetical Series A. The team is exceptional, the wedge (border towers funding a Lattice land-grab) de-risks the cold-start problem, and the asymmetry is right: the downside is a decent defense subcontractor acquisition; the upside is the first new prime in 70 years. Ukraine-era tailwinds make the procurement-shift bet more credible than at any point since the Cold War.
The one question I'd need answered first: What does Lattice's software gross margin and stickiness look like independent of Anduril hardware — i.e., will the DoD pay for Lattice as a platform, or only bundled with drones? If Lattice is genuinely detachable, recurring, and sticky, this is a software company wearing a hardware costume — and a fund-returner. If not, it's a very good hardware company at a software price.