Defense Sector Reset: Why Silicon Valley Is Winning the War the Pentagon Can't Fight
Steve Eisman talks to Peter Arment, Aerospace and Defense Analyst from Baird about current state of Defense Sector & Investments.
- Defense stocks are selling off mid-war because a $350B budget reconciliation bill may not pass - not because the secular defense thesis is broken
- The five-company prime contractor monopoly built after the Cold War is structurally incapable of the fast, cheap production modern warfare demands; Silicon Valley is building the replacement
- Arment’s LMT breakup thesis (three-way spin, 50-100% upside) was rejected by management; the stock is $100 cheaper and he still believes it
- The cleanest ideas in the conversation are two nuclear-exposed defense names (Curtiss-Wright, BWX Technologies) where the secular tailwind compounds regardless of the reconciliation outcome
The Paradox: Wars Are Running and Defense Stocks Are Selling
Peter Arment is not calm. “I’m bewildered,” he told Eisman, “given that there’s a war going on, why people would be selling defense.” Peter @ 10:06
That bewilderment is earned. Multiple active conflicts. NATO rearmament. A documented munitions shortage. Defense stocks selling off anyway.
The explanation is one number: $350 billion. Arment came into 2026 expecting a $450B defense reconciliation package. The proposal landed at $350B. Then the Iran conflict fractured Republican votes - some in the party unwilling to fund what looks like a forever war - and the market started pricing in the bill failing entirely. Strip out the supplemental and the base defense budget grows only modestly from its $1.1T floor. Stocks that were pricing “pass” are now pricing “fail.”
The secular demand has not changed. Wars, depleted munitions, NATO rearmament, AI-driven nuclear power demand - those tailwinds run on a different clock than a congressional vote.
How It Got Here: The Last Supper
Understanding the selloff requires understanding how the defense industry got here.
The US had approximately 60 defense prime contractors (primes: the companies that assemble the final weapon system, as opposed to component suppliers) in the 1980s. Peter @ 1:47 After the Cold War, Defense Secretary Les Aspin and Undersecretary William Perry invited the industry CEOs to dinner - what became known in the industry as the “Last Supper” - and delivered a direct message: the peace dividend is real, budgets are being cut in half, consolidate or die.
Sixty primes became five: Lockheed Martin, Northrop Grumman, Boeing Defense, RTX/Raytheon, and General Dynamics/L3Harris. Those five now control roughly 50% of all US defense procurement spending. Rational in the 1990s context. Revealed as a structural problem by the 2020s.
When the industrial base over-consolidates under cost-plus contracting (the government pays actual cost plus a fixed margin), you optimize for exquisite, expensive systems - not fast or scalable ones.
Ukraine was the stress test. The demand came in and, as Arment put it, “we couldn’t supply that.” Peter @ 5:13 The industry was built for multi-year, cost-plus, one-at-a-time production. Not high-volume replenishment. It showed.
The Munitions Math
That structural failure has a price tag.
A $4 million Patriot interceptor missile used to shoot down a $20,000 Iranian Shahed drone. That is the cost-exchange ratio the US and its allies are currently operating at. Peter @ 14:00
The Patriot is not an aberration - it is the best tool available. And you cannot call Raytheon and order 4,000 more on short notice. Those are years-long ramp cycles. Peter @ 14:28 Stinger missiles, THAAD batteries, other interceptor stocks are all depleted through Ukraine and Middle East operations.
Arment characterizes the depletion level as consistent with his industry contacts, even if not publicly confirmed. The administration’s policy response was direct: force primes to reinvest capex in production capacity instead of buying back stock. That showed up in Q4 2025 earnings calls. Peter @ 21:00 For the first time in over a decade, not one prime CEO discussed buybacks. All announced elevated capital expenditure. The 7-day government payment cycle that made primes extraordinary cash machines for 15 years is now redirected toward rebuilding the industrial base.
The capex mandate matters for individual stock theses. It leads directly to the biggest single-company call in the conversation.
Lockheed Martin: A Catalyst Without a Buyer
Arment published a research report in October 2025 arguing that Lockheed Martin should break itself into three publicly traded entities. The logic is clean.
LMT trades at roughly 20x EBITDA. Peter @ 31:46 Defense tech companies - the Silicon Valley-backed names selling faster, cheaper systems - trade at 40x EBITDA and 70x earnings. The gap is structural, not noise.
Arment’s three-part split: F-35 and Sikorsky (a capital-light cash franchise that would trade at value-investor multiples), missiles ($15B revenue, growing with every ally rebuilding interceptor stocks), and space ($12B+, classified national security portfolio). Peter @ 32:23 Isolate each unit, price it at comp-appropriate multiples, and the result is 50-100% upside over an 18-month spin timeline. Peter @ 34:10
The report caused a brief stock reaction. Then management made clear a breakup was not under consideration. The stock has since fallen roughly $100 - from around $600 to $500 - from both the broader defense selloff and the absence of the catalyst. Arment calls himself “a voice in the wilderness.” Peter @ 35:14 He still believes the math.
The meta-signal worth noting: the thesis was rejected and the stock got cheaper. The sum-of-parts discount is wider today than when he published it. If management ever revisits - under activist pressure, continued valuation divergence with defense tech, or a capex environment that compresses free cash flow - the starting point is materially better.
Northrop Grumman: The Overlooked Prime
If the LMT breakup never materializes, Arment still has a preferred prime - and it is not Lockheed Martin.
When Eisman asks directly which prime has the best portfolio, the answer is immediate: Northrop Grumman. Four pillars: B-21 Raider (sole contractor for the next-generation stealth bomber replacing the B-2), Sentinel (sole contract to replace all 400 Minuteman III ICBMs - “another massive, massive program”), approximately one-third of all F-35 system components, and a large classified space portfolio supporting national reconnaissance. Peter @ 35:26 “Checking all the boxes,” in Arment’s framing.
The obstacle is identical to LMT’s: the same 20x EBITDA ceiling, the same capex mandate, the same absence of a near-term re-rating catalyst. Better portfolio. Same structural problem.
Silicon Valley’s Defense Playbook
The prime consolidation failure is not just an accounting problem. It created a market gap, and a new class of companies is filling it fast.
Anduril is the defining example. Founded by Palmer Luckey (who sold Oculus to Meta in 2014 and left after supporting Trump in 2016), Anduril is building an autonomous tactical fighter called Fury - private, valued at roughly $60 billion, backed by both defense capital and tech venture capital. Peter @ 5:58 When Luckey’s team first approached the Pentagon, they reportedly couldn’t get a first meeting for years. Now the DoD can’t respond fast enough.
The old exit route for PE and VC-backed defense companies - build, then sell to Lockheed or General Dynamics - is effectively blocked. The over-consolidated primes cannot make large acquisitions without antitrust concerns. So the public market has become the exit. Avex and ARXIS both recently IPO’d. Carmen Space and Defense, a PE roll-up of seven or eight niche missile and space launch businesses, went public and rose 4x in 2025 before pulling back 50% - yet still trades at 40x EBITDA. Peter @ 41:04
That 40x multiple after the selloff is Arment’s point. Defense tech is not being discounted to prime-comparable multiples. It is pricing on the right side of the asymmetry.
Drones: AeroVironment and the Directed Energy Bet
Among the investable defense tech names, Arment has one clear preference: AeroVironment (AVAV).
The company’s reputation was built on the Switchblade - a man-portable loitering munition (a drone with a warhead attached) carrying a Javelin-class warhead (approximately 7.5 lbs), with a range of 45 kilometers and a cost of roughly $150,000 per unit. Peter @ 37:23 Used extensively in Ukraine against armored vehicles. Eisman’s reaction unprompted: “$150,000 versus $80 million for an F-35.” That ratio is the loitering munition argument in one sentence.
The transformation story is the Blue Halo acquisition, which closed in May 2025 and doubled AVAV’s size. Blue Halo brought directed energy weapons, including a laser demonstrated in Albuquerque that neutralized a DJI-class drone at 2km in 2.5 seconds. Peter @ 38:47 Lasers are the structural answer to the $4M interceptor problem - near-zero marginal cost per shot. The current constraint is power: a 50kW laser requires 100kW of input power, limiting near-term deployment to ships and large platforms with excess power generation.
If directed energy becomes the dominant counter-drone solution, AeroVironment is positioned across the full spectrum: low-cost loitering offense and laser-based defense in one ticker.
Avex and ARXIS, both recently public names Baird supported in their IPOs, get brief treatment as supporting cast.
Ukraine: What the Wars Revealed
The demand signal from Ukraine is well-reported. The capability gaps are less so.
Russia’s electronic warfare - specifically GPS jamming - caused early US precision munitions to miss targets. Peter @ 45:41 The US had tested its precision systems for decades in non-peer adversary environments where GPS jamming was not a factor. Afghanistan and Iraq did not prepare the guidance systems for a near-peer contest. Ukraine was the first real test. The discovery was uncomfortable.
A secondary point: Ukrainian drones are innovative and cheap, but they run on Chinese-made electronics. The US cannot adopt them for security reasons. The gap between Ukrainian drone innovation pace (driven by existential necessity) and US procurement pace (driven by process) is not closing quickly.
Boeing: The Capped Recovery
The defense sector dominates the conversation, but Arment covers aerospace too - and the Boeing story is both compelling and permanently constrained.
The Boeing-Airbus commercial aerospace duopoly is genuinely exceptional. Global air traffic has grown in every year but four in four decades: 1991, 9/11, the 2008 financial crisis, and COVID. At roughly 5% compounding, the installed base of commercial aircraft doubles every 15 years. Peter @ 49:07 That structural franchise is intact.
What changed is the cash flow ceiling. The 737 MAX crisis traced to a 2012-2014 decision: Boeing chose to re-engine the existing 737 rather than start a clean-sheet design. Peter @ 52:02 A clean-sheet would have given Airbus eight years to take share - an unacceptable competitive risk. So they stretched the airframe for heavier engines, which required a flight control software workaround called MCAS (Maneuvering Characteristics Augmentation System, designed to compensate for altered aerodynamic behavior), which was flawed, and which caused two crashes.
The capital raises to rebuild the balance sheet permanently diluted the equity. Long-run free cash flow ceiling: from approximately $40/share to approximately $20/share. Peter @ 54:29 Production is recovering. Arment expects peak rates around 2028-2029. The franchise held.
The per-share ceiling did not.
Arment has a $300 price target and calls Boeing a good stock. But the path to $600 that investors thought they were on before 2018 is not coming back. Peter @ 55:30
Nuclear Defense: The Two Names Eisman Takes Home
Eisman enters this section not knowing much about Curtiss-Wright or BWX Technologies. He leaves calling them the most interesting companies of the episode. That reaction is itself a signal.
Curtiss-Wright (CW) runs two businesses under one ticker: defense electronics (~55% of revenue) and nuclear components (~45%). The defense electronics side is solid but unremarkable. The nuclear side is the thesis.
Two distinct nuclear demand streams are running simultaneously. First: CW makes reactor coolant pumps for US Navy submarines and aircraft carriers, along with other nuclear-qualified components. Naval nuclear spending is growing with accelerating shipbuilding. Second: of 94 commercial reactors operating in the US, more than 50% are applying to extend their operating lives from 40 to 60 years, or 60 to 80 years. Every extension requires NRC-mandated component replacements. CW is a qualified supplier for that work.
The AP1000 angle is the most specific number Arment offers: CW holds the exclusive contract for Westinghouse’s AP1000 reactor coolant pumps. With 14 AP1000 plants currently in development globally, at four pumps per plant and $30M per pump, that is $1.7B in revenue at 25-30% EBIT margins - approximately $11/share in earnings contribution, layering onto an already-profitable base. Peter @ 1:01:40 The 14-plant count is current, not a ceiling.
BWX Technologies (BWXT) is the simpler story and the harder franchise to replicate. Spun from Babcock and Wilcox, BWXT has built 430 naval nuclear reactors for the US Navy without a single accident. Peter @ 1:02:59 It is the de facto monopoly supplier for US Navy nuclear reactors - Arment’s exact phrasing: “they technically have a monopoly. We don’t use that word, right?” It also services 19 Canadian nuclear plants (acquired from GE at a time when, as Arment notes with dry irony, GE was not fully appreciating the value of its nuclear assets) and is positioned in the supply chain for every new small modular reactor (SMR, a compact factory-built reactor designed for distributed power generation) deployment.
The commercial nuclear angle for both companies is being pulled forward by AI data centers. Peter @ 1:04:22 Hyperscalers - the Microsofts and Amazons building AI infrastructure - need clean, reliable baseload power that utility companies historically refused to fund. Now the hyperscalers are writing the checks themselves, taking SMR power “behind the meter” and bypassing the utility’s risk aversion. GE Vernova’s BWR-300 (a 300MW small modular reactor) is the leading design. BWXT supplies the nuclear-qualified components that only a handful of North American companies can manufacture.
The European angle reinforces it. Germany shut its last nuclear plant just before Russia invaded Ukraine. Peter @ 57:29 The energy crisis that followed reset European attitudes toward nuclear in ways that years of advocacy had failed to accomplish. That is a second commercial demand signal running in parallel with the AI story.
Three forces converging: naval nuclear growth, commercial life extensions, and AI-driven SMR demand. CW and BWXT are the two public companies sitting at that intersection.
Alpha Takeaway
The defense selloff is driven by reconciliation politics, not secular demand. The fear that broke the prime multiples is creating the entry in the secular compounders. Those two things are not in conflict - they are the same moment.
Five conclusions from this conversation:
CW and BWXT are the cleanest positions. Revenue growth does not depend on the reconciliation vote. The nuclear tailwind - commercial life extensions, new AP1000 builds, AI-driven SMR demand - runs on a separate clock from defense budget politics. Eisman closed this episode saying he needs to do more work on both names. That is the host’s ranking signal.
The LMT breakup thesis is more compelling at $500 than at $600. Management rejection does not change the sum-of-parts math. The discount widened. Activist interest, continued valuation divergence with defense tech, or capex-driven free cash flow compression are the catalysts to watch. Arment is still long.
NOC has the best prime portfolio and the same structural ceiling. B-21, Sentinel, F-35 content, classified space - the program mix is superior. But the 20x EBITDA ceiling does not resolve without a structural catalyst that Arment did not apply to NOC in this conversation.
AVAV is the preferred defense tech name because it now owns both sides of the cost-exchange problem. The $4M interceptor vs. $20K drone ratio is unsustainable. Directed energy (near-zero cost per shot) is the structural answer. Power constraints limit near-term deployment to naval platforms, but the roadmap is clear and the company has the only publicly available offense-plus-defense portfolio.
Boeing is a recovery capped by dilution. The franchise is intact. The FCF ceiling moved from $40/share to $20/share permanently. $300 is the right target. The path to $600 requires assumptions about buybacks and multiple expansion that Arment does not endorse.
This newsletter is for informational purposes only and does not constitute investment advice. All views attributed to named speakers reflect their own opinions, not Podcast Alpha's. The author may hold positions in securities mentioned. Do your own research before making any investment decision.