The defense industrial mobilisation has a ceiling — and the market is on the wrong side of it
The thesis in one line. The Western defense industrial mobilisation is real, but it is structurally capped below the level required to outproduce the Russia + DPRK + Iran combined supply chain — and that ceiling is exactly what makes the trade work. Pricing in either a decisive rerating (NATO outproduces) or a collapse (NATO can't keep up) is the wrong frame. The right frame is Managed Friction at the production rate: enough output to sustain the war, not enough to win it, sustained for years. That equilibrium has a specific portfolio shape, and the consensus is on the wrong side of it.
If you have not been following the supply-chain side of the Ukraine war closely, here is the on-ramp before the equilibrium read: NATO's 155mm shell output has scaled from a wartime baseline well under one million rounds/year to over 2 million in 2025 (EU Defence Commissioner Kubilius via Kyiv Independent, Dec 2024). Russia uses the Soviet 152mm caliber, not NATO 155mm: Russian production scaled from approximately 0.4 million rounds in 2022 to roughly 4.2 million annually by 2025 (RUSI / Estonian intelligence estimates summarised by Euronews, Jul 2025), supplemented by North Korean shipments estimated at up to 5.8 million cumulative since 2023 per the Open Source Centre — accounting for up to half of all Russian shell use in late 2025 (Moscow Times citing Reuters, Apr 2025). The structural gap — NATO ≈ 2M vs Russia + DPRK ≥ 6M annual equivalent — is not closing on a five-to-seven-year horizon. Three players, three different bets on what happens next.
Three pillars
One. Putin is running the maximalist mobilisation play, and it is working. Russian artillery shell output scaled roughly tenfold from approximately 0.4 million rounds in 2022 to about 4.2 million annually by 2025 (RUSI / Estonian intelligence via Euronews) — partly through round-the-clock shifts at existing plants, partly through DPRK supply (an asymmetric subsidy the West cannot match), partly through Iran's drone production scaling. The cost is real — capital equipment quality is decaying, sanctioned dual-use components are degrading round reliability, the labor force is being burned. But the throughput is what matters at the front. Putin's trade is not "outproduce NATO forever." It is "outproduce NATO long enough that Western political will fragments." That is a credible-commitment play with a 5-7 year time horizon.
Two. Xi is holding the ceiling. China's commercial-defense capacity (uncommitted to any war, available if mobilised) is the largest unused defense industrial base in the world. China builds roughly half of all global commercial vessels (2024 share) and a leaked US Office of Naval Intelligence slide put Chinese shipbuilding capacity at 232× that of the US alone (reported by The Wire China, Oct 2024). Combined US + Japan + South Korea closes most of that gap on commercial tonnage, but the deterrent overhang is structural against any unilateral US naval-mobilisation play, which is the relevant comparison for the Pacific theatre. Xi's strategy is not to mobilise it — that would be expensive, would commit China to a side, and would burn Chinese diplomatic optionality. Xi's strategy is to signal that he could. The ceiling effect on Western mobilisation is precisely this overhang: NATO can scale, but not infinitely, because the ceiling is not "what Russia produces today" — it is "what Russia + China combined could produce if Xi flipped the switch." The deterrent runs both ways.
Three. Bessent and Netanyahu are running the secondary-sanctions and export-demand sides of the same equilibrium. Bessent's Treasury is using secondary sanctions to constrain Chinese dual-use exports to Russia — slowing the supply chain without forcing China into a binary commitment. Netanyahu, separately, is scaling Israeli defense exports (drones, air defense, precision munitions) into a global market that has discovered Western capacity is short. These two players are not coordinating, but their independent moves create the same structural effect: Western defense procurement gets routed through new primes (Israel, South Korea, mid-cap European) rather than scaling the legacy US primes proportionally. That is the marginal-buyer geography the consensus has not priced.
What the consensus is mispricing
The market is reading defense industrial as a US-large-cap rerating story (Lockheed, RTX, Northrop, General Dynamics — sustained multi-year procurement growth). That read has two structural problems. First, US large-cap primes are throughput-constrained at the facility level — adding capacity requires multi-year capex commitments that the procurement budget has not actually authorised at scale. Second, US procurement is heavily allocated to long-cycle platforms (F-35, Columbia-class submarines, B-21) which absorb dollars without scaling munitions output. The shell-production growth is happening on shorter cycles, in different geographies, by different companies.
The actual marginal-buyer story is European. EU's ASAP and EDIRPA instruments are funding mid-cap mobilisation directly. Rheinmetall is running shell-production lines that did not exist in 2022. BAE is scaling at facility level, not just margin. Saab and Leonardo are picking up air-defense and electronic-warfare contracts that legacy US primes are not bidding on (because of the long-cycle allocation problem above). South Korean exporters (Hanwha, KAI, LIG Nex1) are filling the same gap from the Asian side — Hanwha's K9 howitzer is the artillery-platform standard for several NATO procurements that legacy US primes were too slow to bid on.
Position implied
This is a mid-cap-European-and-Korean over US-large-cap-prime trade, with three concrete components.
One. Long European mid-cap defense primes — BAE, Rheinmetall, Saab, Leonardo, Thales — over a 12-month forward window. The repricing is happening but has not finished; current valuations still treat these as cyclical European industrials rather than structural-ceiling beneficiaries. As an asymmetric-exit-cost read: governments that have funded ASAP/EDIRPA cannot now defund without losing strategic autonomy, so the procurement floor is hard. Multi-year contract visibility is materially better than the equity prices imply.
Two. Long South Korean defense exporters (Hanwha, KAI, LIG Nex1). They are the Asian-side beneficiary of exactly the same marginal-buyer dynamic. Hanwha Aerospace alone posted record 2025 revenue and profit on strong ground-weapon sales — operating profit roughly +75% year-on-year, with the K9 howitzer now the world's most exported modern artillery system, in service with NATO members including Poland, Romania, Estonia and Norway (Korea Herald, Feb 2026). Order books across the Big Four Korean defense firms ran ahead of equity-multiple expansion through 2022-2025. The Western-procurement-is-slow gap is structural, not temporary, because US large-cap primes are not scaling the relevant short-cycle production lines.
Three. Skepticism on the US-large-cap-prime rerating trade. Lockheed, RTX, Northrop, GD will continue to grow revenue — but multiple expansion past current levels prices a mobilisation tempo that the procurement budget structure does not allow. Trade the volatility around earnings prints; do not trade the multi-year multi-expansion thesis.
I am staking this as a 12-month falsifiable prediction. The criterion is a relative total-return basket comparison through 2027-05-04: if the equal-weighted EU mid-cap defense basket (BAE / Rheinmetall / Saab / Leonardo / Thales) does not outperform the equal-weighted US large-cap prime basket (Lockheed / RTX / Northrop / GD) by ≥10 percentage points over the 12-month window, this thesis is wrong. That is a hard line, dated, with a defined comparison.
What I cut from the draft
- A long detour on the counter-sino-russian bloc-level coordination on industrial-base sharing — the China-Russia coordination on defense components is real but lighter than the bilateral DPRK-Russia version above; full treatment when the next BIS report drops.
- A read on Indian defense exporters (Bharat Dynamics, HAL) — they are the third leg of the marginal-buyer geography but the equity story has different dynamics (state-controlled, currency-exposed) that deserve a standalone issue.
- A position on US shipyard equities (Huntington Ingalls, General Dynamics shipbuilding) — the Columbia-class allocation is the long-cycle equivalent of this thesis but plays out on a 5-7 year horizon, not 12 months.
The defense industrial story is not "NATO is winning" or "NATO is losing." It is "NATO is producing enough to sustain attrition, not enough to break the equilibrium" — and that has a specific marginal-buyer geography the consensus has not priced. Position accordingly. The 12-month prediction above is the falsifiable line.