The Rearmament Supply Chain

Executive Summary

Global defense spending hit $2.72 trillion in 2024 -- a 9.4% real increase, the steepest since the Cold War. Europe alone added $100 billion. NATO committed to 3.5% of GDP by 2035. The demand signal is legislated and bipartisan.

The consensus response -- buying defense primes through ETFs -- misses the real opportunity. The bottleneck is not at the top of the supply chain. It is deep inside it: in munitions production running at a fraction of wartime demand, in critical minerals where China controls 60-90% of processing, in shipyards that have halved over fifty years, and most critically, in the precision electronics embedded inside every modern weapons system.

But scarcity alone does not guarantee profit. Shipyards are scarce yet earn mediocre returns under cost-plus contracts. The real edge lies where scarcity meets pricing power -- sole-source component suppliers with multi-year certification moats, small enough to avoid government margin scrutiny, too critical to replace.

The cycle will unfold in waves: munitions (2024-27), electronics and sensors (2026-32), shipbuilding (2028-40), and software/autonomy (ongoing). Each wave creates different winners. The later waves are less crowded.

This paper provides the analytical framework. ClarityX's MARY system -- a multi-agent AI platform -- executes it: scoring candidates, monitoring crowding, managing wave rotation, and enforcing exit discipline across themes simultaneously.

The scarcity is never where everyone is already looking. Finding it is the first step. Capturing it systematically is the real edge.

---

The Demand Signal

Global defense spending hit $2.72 trillion in 2024 -- 9.4% real growth year-over-year. Europe added $100 billion in a single year. Germany legislated EUR 95 billion rising to EUR 162 billion by 2029. Poland is at 4% of GDP. NATO agreed to 3.5% by 2035.

These are legislated commitments, not projections. The demand signal is as close to guaranteed as government spending gets. And yet, the way most investors play this trade reveals a fundamental misunderstanding of where the constraint sits.

---

The Consensus Trade and Its Problem

The default response is to buy the primes -- Lockheed, RTX, Northrop, Boeing. This is the trade $15 billion of passive money made in 2024-2025, flowing into 25 defense ETFs, 11 launched in a single year.

The problem is structural. These companies are systems integrators. They assemble. They do not produce the scarce inputs. The bottleneck is not at the top of the supply chain. It is deep inside it.

---

Where the Scarcity Lives

Munitions. The U.S. produces ~40,000 155mm shells per month -- a 178% increase from pre-war levels, yet Ukraine consumes that in weeks. The Army's 100,000/month target is not expected until mid-2026. Patriot interceptors: ~740/year. Iran's assault on Israel consumed $800 million of interceptors in 11 days.

Critical minerals. China controls 60-90% of processing for defense-critical materials. Near-zero domestic U.S. tungsten production. Titanium disrupted by Russia sanctions. The GAO found DoD lacks data to project shortfalls for nearly half its critical materials.

Shipbuilding labor. Seventeen shipyards closed over fifty years. Virginia-class submarines: 1.1-1.2 boats/year against 2/year procurement. One yard builds carriers. You cannot train specialized welders in six months.

But these visible bottlenecks obscure a deeper one.

---

The Constraint Beneath the Constraint

Modern weapons are electronics systems wrapped in steel. A Patriot interceptor is a guidance seeker and radar processor that happens to fly. A drone swarm is sensor fusion and RF components. A submarine is sonar arrays and combat management electronics.

The deepest scarcity sits inside the weapons: guidance electronics, high-temperature semiconductors, radiation-hardened chips, solid rocket motor propellants. These come from extraordinarily thin supplier bases -- sometimes a single qualified source -- with certification cycles measured in years.

Ammunition shortages are a factory capacity problem that capital can solve. Electronics constraints depend on specialized talent, proprietary fabrication, and certification regimes that money alone cannot accelerate.

---

A Supply Chain Framework

Map the chain from most visible to most obscured:

Layer 0: Primes. Lockheed, RTX, Northrop, Boeing, BAE. Every defense ETF. Maximum analyst coverage. The consensus already owns them.

Layer 1: Subsystems and components. Flight controls, EW modules, missile seekers, guidance electronics. McKinsey found ~75% of value is created here, not at final assembly. Deep certification moats, real switching costs. Electronics scarcity meets pricing power.

Layer 2: Munitions. The most acute near-term bottleneck. Demand far exceeds capacity -- but capacity expansion is the most straightforward fix.

Layer 3: Critical materials. Rare earths, tungsten, titanium, energetics. China dependency driving onshoring investment, but Western alternatives are early-stage.

Layer 4: Defense tech and software. Autonomy, AI C2, space sensing. Ukraine proved cheap drones neutralize expensive platforms. Most leaders are private; one public exception trades at ~80x revenue.

Layer 5: Shipbuilding. Longest-cycle bottleneck. Five to ten years to build new capacity. More about patience than price discovery.

Value creation concentrates below the primes, where crowding is lower and bottlenecks are physical.

---

Scarcity Is Not Enough

Shipyards are among the scarcest defense assets. They also earn mediocre returns -- governments squeeze margins through cost-plus contracts and regulatory oversight. Scarcity without pricing power is a trap.

The best economics accrue to nodes that are difficult to replace and face limited buyer leverage. Component suppliers with multi-year certification requirements and no qualified second source have both. Re-qualification takes two to five years and costs tens of millions. The component cost is small relative to the platform -- too small to negotiate aggressively, too critical to risk disrupting.

The filter: what is scarce, hard to replace, and small enough to avoid margin pressure? That points to Layer 1 subsystems and specialized electronics -- deep enough to carry certification moats, below the threshold of political visibility.

---

Bottlenecks Unfold in Waves

Rearmament is not one trade. It is a sequence.

Wave 1 (2024-27): Munitions. Most visible, most immediate. Already partially priced -- some names up 200-400% from 2022 lows.

Wave 2 (2026-32): Electronics, sensors, precision components. Thinner supplier bases, longer resolution timeline. Cannot be accelerated with money alone.

Wave 3 (2028-40): Shipbuilding and heavy platforms. Decade-long structural constraint. New yards take five to ten years.

Wave 4 (ongoing): Software, autonomy, AI. Does not resolve -- it accelerates. Spending compounds rather than peaks.

Each wave creates different winners. Later waves are less crowded.

---

What History Tells Us

1930s. Vickers surged 5x (1933-36). But Anglo-Iranian Oil and US Steel -- supply chain inputs -- were among the strongest performers by 1938-39.

Reagan buildup. Defense spending rose from 4.9% to 6.8% of GDP. The fifty largest suppliers consolidated into today's top five. Those who held through the 1990s "peace dividend" saw a decade of deterioration. Exit timing mattered more than entry conviction.

Post-9/11. Defense stocks returned 14% annualized vs 1% for the S&P 500 (2001-07). But stock prices peaked ~2007 -- two to three years before spending peaked ~2010. The market prices the rate of change, not the level.

---

What Makes This Cycle Different

Multi-polar threat. No single peace agreement ends the spending pressure.

Industrial base starts from weakness. Decades of consolidation. Munitions factories from the 1920s. Starting weakness amplifies scarcity.

Europe is a new demand source. EUR 300 billion projected by 2030. ReArm Europe and SAFE facility (EUR 150B) are new financing mechanisms.

Great-power competition has no off switch. The Cold War lasted four decades. Today's threat landscape -- China, Russia, Iran, North Korea -- is multi-polar. There is no single treaty, summit, or regime change that resolves it. This is not a cycle that peaks with a peace agreement. The spending pressure is structural and generational.

Crowding forms faster. $15 billion into defense ETFs, 11 new ETFs in one year. Layer 0 crowding is already advanced.

---

The Risk That Matters Most

The risk is not that spending declines. It is that spending growth decelerates.

Defense stocks peak before spending peaks. If European budget growth slows from 13% to 5% -- still growing -- the market may re-rate downward. The trade is in the delta, not the level.

Fiscal constraints amplify this. Reagan built up at ~30% debt-to-GDP. Today: ~120%. Europe is carving out defense from fiscal rules, but fiscal limits may cause deceleration earlier than commitments suggest. The market prices deceleration before it appears in budgets.

The secondary risk is crowding. Twenty-five defense ETFs suggest Layer 0 is approaching saturation. The question is whether deeper layers follow the same trajectory.

---

Quantifying the Opportunity

Within each layer, how do you separate the names that capture economic profit from those that merely participate? We use a composite model -- the Supply Chain Alpha Score (SCAS) -- across five dimensions.

Scarcity. How fast is demand growing relative to the capacity to supply it? A munitions line running at wartime demand with two-year expansion timelines scores high. A commodity input with elastic supply scores low.

Pricing Power. Does the supplier have the leverage to expand margins? Certification barriers, sole-source status, and low component cost relative to platform cost all contribute. A guidance seeker that costs 2% of a missile but takes four years to re-qualify has structural pricing power. A cost-plus shipyard does not.

Crowding (inverted). How much institutional money has already found the name? ETF inclusion, analyst coverage, and ownership velocity all erode future asymmetry. The best scores go to names the consensus has not yet discovered.

Valuation Gap. Is the current valuation below the historical norm, adjusted for backlog visibility? The framework rewards names where the market has not yet priced the scarcity.

Margin Trajectory. Are operating margins expanding over the trailing eight quarters? In a capacity-constrained environment, expanding margins are the clearest signal that pricing power is being exercised.

Each dimension scores 0-10. Weights shift by wave -- Wave 1 emphasizes Scarcity; Wave 2, Pricing Power; Wave 4, TAM expansion over traditional valuation. Every input is observable. Every weight is adjustable. The framework prevents conviction from outrunning evidence. The full scoring methodology and formulas are in the Appendix.

---

The Execution Problem

Knowing where to look is the first problem. It is not the hardest.

The hardest problems are: what to buy within the right layer, when to enter, how much to size, and when to exit before crowding destroys the asymmetry.

Wave 1 names have moved 200-400% from 2022 lows. Correct thesis, wrong sizing meant missed returns. Correct thesis, missed exit signals meant giving back gains. This paper tells you where the scarcity sits. It does not tell you how to scale in, hedge drawdown risk, rotate from Wave 1 to Wave 2, or set exit triggers that fire before consensus catches up.

These are solvable problems. They require systematic monitoring of SCAS inputs across dozens of candidates, real-time crowding tracking, continuous conviction recalibration, and the discipline to exit positions that no longer meet criteria -- even when the narrative still sounds compelling.

No human analyst monitors all signals across all layers, waves, and geographies simultaneously.

---

From Framework to System

The execution problems described above -- scoring, monitoring, rotating, exiting -- are not unique to defense. Every thematic investment cycle creates the same challenges: too many candidates, too many signals, too little discipline.

This framework was developed alongside MARY, ClarityX's analytical infrastructure for systematic thematic investing. MARY is a network of specialized agents, each responsible for a different dimension of the problem:

• Macro Agent -- monitors regime shifts, fiscal dynamics, and spending growth trajectory. The single most important signal for timing thematic exposure.

• Fundamental Agent -- scores companies against frameworks like SCAS using real financial data. Tracks margins, maps certification moats, flags deteriorating scarcity-to-crowding ratios.

• Portfolio Agent -- translates conviction into position sizes, manages wave rotation, enforces risk limits against thematic concentration.

• Alpha Lab -- scans for emerging scarcity nodes before they appear in ETFs and sell-side initiations. Identifies Wave 2 and 3 names early.

The system runs this process across multiple macro themes simultaneously -- AI infrastructure, reshoring, energy transition, defense -- managing cross-theme correlations at the total portfolio level.

The white paper gives you the framework. The analytical infrastructure executes it continuously -- monitoring the signals a human analyst cannot track across all layers, waves, and geographies at once.

---

A Framework, Not a Recommendation

This paper offers a framework, not specific positions -- and a glimpse of how systematic analytical infrastructure turns frameworks into actionable intelligence.

The real scarcity sits below the primes: precision electronics with certification moats, munitions capacity, critical materials, and the software layer redefining warfare. The nodes that earn the best economics are those where scarcity meets pricing power -- hard to replace, small enough to avoid margin pressure, and embedded too deep for political visibility.

The cycle unfolds in waves. The investors who benefit most will be positioned where scarcity, pricing power, and patience intersect -- with the analytical infrastructure to monitor, size, and exit as the cycle evolves.

The scarcity is never where everyone is already looking. Finding it is the first step. Capturing it systematically is the real edge.

---

Parson Tang is the founder of ClarityX Research Institute. Prior to ClarityX, he held investment and technology roles at Goldman Sachs, J.P. Morgan, and Credit Suisse over a 20-year career.

For inquiries about ClarityX's analytical capabilities and the MARY system, contact clarityxresearch@gmail.com.

This paper is for informational purposes only and does not constitute investment advice. The views expressed are those of the author and do not represent recommendations to buy, sell, or hold any securities.

---