Defense Supply Chain Surge Capacity Atrophy: A Structural Problem Capital Cannot Fix

Between 1980 and the late 1980s, the Pentagon procured roughly 16,000 first-generation Stinger missiles, per the public production history of the program documented at FIM-92 Stinger. By the time Russia invaded Ukraine in February 2022, RTX/Raytheon had not built a Stinger for the U.S. Army in close to two decades. That single fact is the cleanest way to see what defense supply chain surge capacity atrophy actually looks like.

Key facts at a glance

• Raytheon could not restart Stinger production until 2023 because seeker-head components were obsolete, per Defense One reporting on the retiree recall.
• The U.S. Army awarded a fresh $580M Stinger contract in September 2025, signaling sustained multi-year demand, per Army Recognition.
• F-35A operating cost was roughly $34,000 to $36,000 per flight hour in FY24, per the Congressional Budget Office, well below the $100,000 figure often quoted in commentary.
• Lockheed Martin crossed 1,300 cumulative F-35 deliveries with a record 191 in 2025, per the company’s January 2026 press release.
• Anduril closed at a $61B post-money valuation in May 2026, per CNBC.

The structural picture

Program	Vintage scale	Recent reality	Implication
Stinger production line	Active 1980 to 2003	Restart ordered 2022, ramp into 2024	~20 years dormant
F-35 sustainment	~$34k per flight hour FY24 (F-35A)	Lifecycle cost $1.58T	Costs rising 44% since 2018
Boeing 787 program	$5B to $6B original budget	>$32B program cost	~5x to 6x overrun
Anduril valuation	$14B (2024)	$61B (May 2026)	~4.4x in under 18 months

Sources for the table: CBO, GAO-24-106703, Seattle Times on 787 program costs, CNBC.

The bull case: capital is finally rewiring the system

The bullish read is that $66B of estimated venture and private capital deployed into defense tech from 2020 to 2024 (a figure attributed to Baird’s Peter Arment on The Real Eisman Playbook Ep 61) is starting to produce contractual proof, not just decks. Anduril took a roughly $20B Army enterprise ceiling in March 2026 along with the IVAS mixed-reality program, per Responsible Statecraft and Anduril’s own announcement. PitchBook’s “Iron Bubble” analysis argues NATO 3 percent of GDP commitments and attritable-mass doctrine justify sustained capital flows, per PitchBook.

Policy is moving with it. The Pentagon replaced the Defense Acquisition System with a new Warfighting Acquisition System in late 2025, and the White House issued an executive order in January 2026 restricting buybacks and dividends at primes that miss schedules.

The bear case: cost-plus and ITAR are upstream of who wins

The bearish read is that cost-plus contracting under FAR 16.306 reimburses contractors for expenses plus a fee, so delay and complexity remain the business model regardless of which logo holds the contract. GAO-06-66 documented that the Department of Defense paid billions in award and incentive fees regardless of acquisition outcomes. Bill Greenwalt, former Deputy Under Secretary of Defense for Industrial Policy, argues in Breaking Defense that the same Stinger that cost about $25K in 1991 now exceeds $400K because the incentive architecture sits upstream of contractor selection.

The Project On Government Oversight calls the systemic bias toward small numbers of high-cost platforms the “exquisite defense” problem, with empirical performance failures catalogued in their analysis.

ITAR is the second binding constraint. Ukrainian drone manufacturers source 70 to 90 percent of components from China because U.S. and European export controls make Western parts slower and more expensive, per CSIS and DroneLife. The cheap drone solution that wins on the battlefield is not, on current rules, a solution the same Pentagon writing checks to Silicon Valley can deploy.

What the data shows

F-35 sustainment costs rose 44 percent since 2018, every 2024 jet was delivered an average 238 days late, and mission-capable rates declined for a fifth consecutive year, per GAO-24-106703 and the National Defense Magazine summary. Output is real (1,300 jets globally, 191 in 2025), but the depreciating cash-cow framing fits the GAO record.

Boeing illustrates the dilution arithmetic. The 787 program absorbed over $32B in cumulative deferred and non-recurring costs against an original budget of $5B to $6B, per Seattle Times and Leeham News. In October 2024 Boeing priced 112.5 million common shares at $143 plus $5B of mandatory convertible preferred, per the company’s 8-K. Bernstein lists Boeing as a top 2026 idea at a $298 target on roughly $14B of normalized 2029 free cash flow, per Investing.com. UBS warns that supply-chain content inflation could compress that target before per-share math even matters, per Morningstar synthesis.

Lockheed’s three-way breakup case rests on segment math: Honeywell is splitting into three with the aerospace spin completing in 2026, per Breaking Defense, and GE Aerospace is up roughly 140 percent since its 2024 separation. The counter is integration scope. The Space Force tasked twelve companies in April 2026 with space-based interceptor work inside a multi-domain Golden Dome architecture, per Breaking Defense. Lockheed owns space sensors, ground radars, and interceptors in one house. Q1 2026 backlog held at $186.4B with Missiles and Fire Control growing 8 percent, per Lockheed’s 8-K.

BWX Technologies sits in the corner of the market that venture capital cannot legally enter. The company has delivered more than 420 reactor cores to the U.S. Navy and added $1.4B in naval propulsion contracts in 2025, per the BWXT release, the FY2024 8-K, and Naval Technology. Bank of America values the equity at roughly 33x EV/EBITDA for that reason, per Yahoo Finance.

What would change our mind

• The Pentagon publishes audited evidence that fixed-price reform accelerates the median program, not only the headline ones.
• A cleared, second-source naval reactor manufacturer enters Q-clearance qualification, breaking the BWXT perimeter.
• Boeing reaches $10B of annual free cash flow without further equity issuance through 2028.
• Lockheed announces a strategic review led by an outside advisor rather than an internal portfolio study.

What this does not tell you

• The $66B venture capital figure is analyst framing dependent on definition. PitchBook’s narrower U.S. pure-play series puts cumulative 2021 to 2024 closer to $30B.
• Stinger production claims of 50,000 units between 1980 and 2000 conflate U.S. Army procurement with global licensed production, which exceeded 70,000 lifetime units of all variants.
• The often-quoted $100,000-per-flight-hour F-35 cost overstates the current CBO figure by roughly 3x.
• Predictions that neoprimes displace incumbents are forward-looking. The same dynamics could produce a new oligopoly (Anduril plus Palantir consolidating ~200 prior contracts) rather than a competitive market.

FAQ

What does “defense supply chain surge capacity atrophy” actually mean?

It is the gap between peacetime procurement rates and wartime replacement rates. When Stinger orders went from thousands a year in the 1980s to roughly zero for almost two decades, the tooling, suppliers, and skilled labor required to restart at scale dispersed. Rebuilding that base is measured in years, not quarters.

Will the new Warfighting Acquisition System fix cost-plus?

Possibly at the margin. Cost-plus exists because truly novel R&D risk is hard to fix-price without enormous padding. Reform will likely shift the boundary between cost-plus and fixed-price programs rather than eliminate either.

Is Anduril the new Lockheed?

Not on revenue. Lockheed booked over $70B in 2024 sales, while Anduril’s enterprise ceilings, large as they are, sit on much smaller realized revenue today. They compete in adjacent layers of the stack, not the same one.