Silicon Valley vs the Defense Primes: The Stinger Story, the LMT Breakup Thesis, and Anduril at $61B

On a recent episode of The Real Eisman Playbook, Baird’s aerospace and defense analyst Peter Arment made the disruption case for the defense sector with concrete mechanics, not abstractions. The framing has three parts that lock together: Patriot interceptors fire at drones costing a fraction as much, supply chains for legacy munitions take multi-year ramps that can’t keep up with current demand, and a generation of VC-backed defense startups is now competing for dollars that used to flow exclusively through Lockheed, General Dynamics, and Northrop Grumman. Anduril Industries’ May 2026 Series H put a public number on the capital flow. This article anchors Arment’s framing to the SEC filings and lists the parts that are still narrative.

Key facts at a glance

• Anduril raised $5B in May 2026 at a $61B post-money valuation, led by Thrive Capital and Andreessen Horowitz, per TechCrunch’s reporting on the Series H.
• The US Army’s Lattice enterprise contract awarded to Anduril in March 2026 has a 10-year ceiling of up to $20B, consolidating more than 120 prior procurement actions.
• Lockheed Martin’s total backlog hit $193.6B at year-end 2025, with 191 F-35 aircraft delivered in 2025 and roughly 416 jets still to be delivered, per the Lockheed 8-K Q4 2025 filing.
• AeroVironment, a public US drone maker, trades at a market capitalization in the $10B range with trailing-twelve-month revenue of $1.61B per public market data.
• Arment said on the podcast that he has published a report arguing Lockheed Martin should be broken up.
• Arment cited a “$66B Silicon Valley bet on defense” figure on the podcast. That specific figure cannot be tied to a single primary source and should be read as his framing rather than a confirmed statistic.

Who is Peter Arment

Peter Arment is Senior Research Analyst and Managing Director for Aerospace and Defense at Robert W. Baird, per his Baird bio page. He joined Baird in 2016 after 23 years inside the aerospace and defense industry. His coverage list includes Lockheed Martin, Boeing, RTX, General Dynamics, BWX Technologies, AeroVironment, Curtiss-Wright, and selected supply-chain names.

The mechanical asymmetry: expensive interceptor versus cheap drone

The core operational argument is the simplest one and the hardest to disagree with. Modern integrated air-defense systems built around Patriot missiles fire interceptors that cost in the high six figures per round. The drones they’re shooting down cost in the low thousands. The cost-exchange ratio is several orders of magnitude in the wrong direction for whoever is defending.

Arment on the podcast: “You can’t just turn to Raytheon, which is the defense unit inside of RTX, the big defense company, or Lockheed Martin who also makes Patriot, and say, hey, we need more 4,000 more of those. It takes time for these to ramp up these big systems.”

That sentence has two embedded claims. The cost asymmetry forces consumption of expensive interceptors against cheap attackers. And the legacy industrial base cannot scale production in time to match the consumption rate. Both are observable; neither is a thesis that requires speculation.

The Stinger case study: why supply chain is the binding constraint

The cleanest single illustration in Arment’s framing is the Stinger missile. Stingers became famous when the US supplied them to Afghan fighters against Soviet helicopters in the 1980s. When Ukraine became the new high-consumption theater after 2022, the Pentagon went to RTX’s Raytheon operation and asked for more.

Arment’s framing of the response: “They said, you know, that’s that’s awesome. You bought 20, you bought 40”, and then the conversation turned to multi-year ramp time. The supplier base for Stinger components had been sized for peacetime annual replacement orders. Surging to wartime consumption rates required new lines, new vendors, requalified components, and years of certification.

The lesson generalizes. The same is true for Patriot interceptors, for guided artillery rounds, for almost any precision munition. The Pentagon learned during 2022-2024 that the US industrial base is not currently sized for sustained munitions consumption at modern combat rates. That realization is what creates the opening for new entrants who can manufacture different categories of munitions at different price points.

The old playbook: build, then sell to LMT or GD

Until roughly 2020, the defense-tech startup playbook was clear, per Arment’s framing on the episode: “Build up a business in the defense world and then sell to Lockheed Martin or General Dynamics or Northrop Grumman.” The primes were the consolidating buyers; the startups were the supply chain.

That has changed. The current generation of defense-tech entrants, Anduril, Palantir Technologies, Shield AI, and others, is not selling. They are scaling independently and competing directly for Pentagon contracts that historically would have flowed through the primes. Anduril’s $20B Lattice ceiling is a single data point in support of the shift, but it is a large one: an enterprise contract that consolidates more than 120 separate procurement actions previously distributed across the prime ecosystem.

The F-35 dimension: both can be true

A common misreading of the disruption thesis is that it implies the F-35 program is going away. It isn’t. Arment was explicit on this point. The defense sector is going to spend more money on F-35s while also spending more money on VC-backed autonomy plays. Lockheed delivered 191 F-35s in 2025 and has roughly 416 in backlog as of January 2026. Production has accelerated, not decelerated.

The thesis is about the marginal defense dollar. The F-35 program continues at scale; the next dollar of incremental defense spending is more likely to flow to drone-and-autonomy programs than to a tenth-generation manned fighter. That doesn’t make the F-35 a bad investment for its primes. It does mean the share of total defense spend that flows to the three primes is gradually shrinking.

The LMT breakup thesis

The sharpest specific recommendation in Arment’s coverage is his published research arguing that Lockheed Martin should be broken up. He referenced it directly on the podcast, “you wrote a report on Lockheed Martin where you argued that the company should be broken up. Why?”, and tied it to the F-35 division.

The argument, in the form it tends to take in published prime-defense research, is that the conglomerate discount on Lockheed reflects the difficulty of valuing aeronautics (F-35), missiles and fire control, rotary and mission systems, and space as a single integrated business. Splitting them would unlock value because each piece could be valued on its own multiple. Public-equity history suggests this kind of argument tends to surface late in a multi-year run rather than at the start of one; whether the current period is right for that activism is the question.

The full case requires Arment’s actual research report. It is mentioned here as a directional signal, not as a definitive read.

How the public players stack up

The spread matters more than the ranking. The defense-disruption thesis is not a single trade. It is a story about how a set of distinct businesses with different risk profiles compete for the same Pentagon dollar.

What this disruption thesis does not tell you

The framing is internally consistent. It also assumes several things that are not yet proven.

• Public-versus-private valuation is not comparable. Anduril at $61B is a private valuation set by a single round. Lockheed at its public market cap is set continuously by every share traded. The two numbers are not the same kind of object.
• Multi-year contract ceilings are not obligated spend. A $20B Lattice ceiling is what the Army could spend over 10 years. The actual annual run-rate is much lower. Treating the ceiling as a revenue forecast is a category error.
• The $66B Silicon Valley defense figure cannot be verified to a primary source. It came from Arment’s framing on the podcast. Closest published figures: $130B total VC into defense tech since 2021 (cumulative, multi-source) and $28-38B VC into US defense tech in 2025 alone (per various industry trackers). Treat the $66B as Arment’s framing, not a confirmed datum.
• Primes are not standing still. Lockheed’s Skunk Works and RTX’s Raytheon Advanced Technology have working autonomous-systems programs. The “Silicon Valley vs primes” framing describes pace, not absence.
• DoD procurement is slow. Multi-year program-of-record changes still require Congressional authorization. A genuine reshuffling of the defense-spending mix takes years to settle through the budget cycle.
• Geopolitics changes faster than budget cycles. The thesis assumes a multi-year shift in DoD acquisition priorities. A single national-security event can speed that up or slow it down by years.

FAQ

What is Anduril’s most recent valuation?

$61B post-money on a $5B Series H closed in May 2026, per TechCrunch’s coverage. The prior round was $30.5B, so the valuation roughly doubled in ~11 months.

What is the Lattice contract?

A US Army enterprise contract for Anduril’s Lattice command-and-control software with a 10-year ceiling of up to $20B. The ceiling is the maximum addressable spend over the life of the contract, not an obligated amount.

Why is the Stinger missile relevant?

It’s the cleanest case study of US munitions industrial base ramp time. The Pentagon asked RTX for more Stingers after Ukraine demand spiked; the supply chain required multi-year ramp because peacetime production rates were not designed for wartime consumption.

Is the F-35 program ending?

No. Lockheed delivered 191 F-35s in 2025 and has a backlog of roughly 416. Production has accelerated. The disruption thesis is about the marginal defense dollar, not about cancelling existing programs.

What did Arment publish about Lockheed Martin?

A research report arguing Lockheed should be broken up. The argument ties to the F-35 division and the conglomerate-discount question. Baird’s published research is the canonical source; the podcast reference is a directional pointer.

Where can I read Arment directly?

Baird Equity Research is the canonical source for his published notes. Public commentary surfaces in podcast appearances such as The Real Eisman Playbook.