S18.1 · Government, Defense & Public Sector
Manufacturers of major military aircraft, ships and vehicles, facing new competition from software-defined, venture-backed entrants.
Defense primes design and build the major military aircraft, ships and vehicles at the top of the defense industrial base. Global military spending hit $2.72T in 2024 and is climbing toward roughly $2.9T in 2025 (SIPRI, 2026 fact sheet); the top five US primes — Lockheed Martin, RTX, Boeing Defense, Northrop Grumman and General Dynamics — generate combined revenue above $220B. The spending number is not the story, though. The story is that software-defined, venture-backed entrants are now bidding directly against the legacy primes for programs that were once closed to outsiders, and the moat they are attacking is not capital — it is engineering incumbency.
Real global defense spending grew 9.4% in 2024 (SIPRI) and is accelerating further on NATO's pledge to reach 5% of GDP by 2035 and on Indo-Pacific rearmament — genuinely new dollars, driven by sovereign threat perception rather than reallocation. The US topline, by contrast, sits closer to flat under successive continuing resolutions, so near-term growth accrues to whoever holds European and allied exposure. Production concentrates in the United States and Europe — BAE Systems, Airbus, Thales, Rheinmetall and Leonardo hold the major European share alongside the five US primes — co-located with government customers because ITAR export controls and offset obligations require it. Beneath the top ten platform integrators sit thousands of Tier 2 and Tier 3 suppliers. The primes' position in the chain is the valuable one: they integrate subsystems from component and sensor suppliers, answer directly to government program offices, and keep much of the downstream sustainment and maintenance revenue once a platform is fielded.
Contract phase sets the economics: research and development runs cost-plus, production fixed-price, and operating margins land at 8-12% on a capital-intensive base of shipyards and final-assembly lines. What the margin line understates is revenue quality. Lockheed Martin carries roughly $170B of backlog against about $71B of annual revenue — two to three times revenue, typical of major platform programs and protected by law under multiyear procurement authority. That is contracted, statute-backed visibility that annually appropriated, competitively rebid government work elsewhere in the sector does not have. Entry is barred by top-secret facility clearance, ITAR licensing, and qualified-supplier status that takes decades to establish.
Adjacencies run to defense electronics and subsystems (which primes increasingly integrate vertically), space systems, sustainment and logistics, and industrial manufacturing more broadly. The sustainment capture protects annuity-like revenue already attached to fielded platforms; the space push is the one that reaches for budget the primes do not currently hold.
The FTE-hour collapse working through labor-heavy services segments largely passes this segment by — the business sells manufactured platforms, not staffed hours. The exposure sits elsewhere, in the sustaining-engineering labor moat on legacy platforms. As software-defined and attritable systems substitute for exquisite, heavily engineered platforms, decades of proprietary sustaining-engineering knowledge loses value, and with it the incumbency it protected. Autonomy and attritable-systems programs have opened a distinct "non-traditional prime" category — Anduril and Shield AI are the visible examples — competing for programs once the exclusive preserve of the legacy primes. The Pentagon's Replicator initiative and the Collaborative Combat Aircraft competition show the shift is live now, even though full recapitalization of the platform base remains a multi-decade undertaking.