Methodology — Scarcity Engine

The Thesis

Marshall Plan 2.0

We are entering an era in which the U.S. government is the dominant allocator of capital across defense, energy, critical minerals, and infrastructure. Reshoring incentives, strategic stockpiling, energy security mandates, and AI-driven grid demand are creating multi-decade tailwinds for companies positioned along these supply chains. The Scarcity Engine scores each holding on how well it aligns with this Marshall Plan 2.0 thesis and surfaces companies most exposed to the resulting scarcities.

Marshall Plan Score

0.00 – 1.00 alignment

0.75 – 1.00StrongStrong

0.50 – 0.74ModerateModerate

0.25 – 0.49WeakWeak

0.00 – 0.24Off-thesisOff-thesis

Score Components

Sector Alignment30%

How closely the company's sector matches Marshall Plan 2.0 priorities — defense, energy, critical minerals, and infrastructure.

Commodity Exposure40%

Strength of the company's exposure to ranked strategic commodities, multiplied by any active supply-shock severity.

SEC Filing Activity15%

Recency and density of 8-K, 10-K, and 10-Q filings — proxy for material events and disclosure cadence.

Revenue Quality15%

Government / institutional revenue mix, recurring contract value, and durability of cash flow streams.

Commodity Rankings

Strategic weight × shock multiplier

Commodity	Weight	Reason
Rare Earths	1.00	Defense magnets, EVs, wind turbines — concentrated supply chain risk.
Uranium	1.00	Reactor restarts, SMR buildout, and strategic stockpiling.
Lithium	0.90	Battery storage and EV demand backed by federal incentives.
Copper	0.90	Grid expansion, electrification, and data-center capex.
Natural Gas	0.85	LNG export expansion and baseload power for AI data centers.
Oil	0.85	Strategic reserve dynamics and refining capacity constraints.
Aluminum	0.75	Defense, aerospace, and reshored manufacturing demand.
Sulfur	0.70	Phosphate fertilizer feedstock and battery chemistry.

Active Shock Multipliers

HIGH×1.30MEDIUM×1.15

Active supply shocks boost the effective weight of an exposed commodity by the listed multiplier when computing the Commodity Exposure component.

How Shocks Are Detected

Five-stage agent pipeline

1
Spotter
Pulls data from FRED, EIA, Alpha Vantage, yfinance, NewsAPI, GDELT, and SEC EDGAR on each cycle.
2
Assessor
Sends collected signals to a local LLM that detects supply shocks and assigns severity + confidence.
3
Mapper
Evaluates each portfolio holding's exposure to detected shocks and updates risk narratives.
4
Scout
Searches for beneficiary companies that profit from the active scarcity environment.
5
Narrator
Decays confidence on bad company-specific news and refreshes per-holding recommendations.

Key Assumptions

What to keep in mind when reading scores

Exposure maps	Manually curated; not auto-generated.
Sector tags	LLM-classified, not audited — can be wrong.
AV fundamentals	Rotates 20 tickers per run due to API limits.
SAM.gov contracts	Signal is neutral until enough data accumulates.
OTC tickers	Less reliable scores; thin data and disclosure gaps.

Static reference · No live data