Methodology & Sourcing

How the work is built, graded, and sourced — and where its limits are.

AtomProphet is anonymous by design, so the work has to stand on its method, not a résumé. This page is the audit trail: how every causal link in the Cascade Graph is graded, where the numbers come from, and an honest account of what the analysis does and does not claim.

The Cascade Graph in one paragraph

The Cascade Graph is a directed map of the physical economy: 393 nodes connected by 562 sourced cause-and-effect links and 17 feedback loops. Nodes are drivers (structural demand forces), chokepoints (the constraints that bind), jurisdictions and geographies (who controls the supply), substitutes (what could cap a price), and terminals (the publicly-traded tickers that express exposure). Edges are not decorative — each one is a specific mechanism, written in plain language, assigned a relationship type, and graded by how strong the evidence for it is.

How every link is graded

Every edge carries a basis tag that tells you, without spin, how much to trust it. We would rather show you a defensible inference labelled as an inference than dress it up as a measurement.

Edges also carry a relationship type — drives demand for, strains, is the bottleneck for, constrains supply of, controls production of, is an input to, transmits price to, substitutes for, accrues value to, exposes risk to, and the loop-closing amplifies — each with a defined direction of effect. Open any node's evidence file in the graph and you can read each link, its grade, its magnitude, and its time horizon.

Where the data comes from

Claims are built on primary and institutional sources, not aggregators. In practice that means energy and minerals projections from the IEA, the World Nuclear Association, and S&P Global; climate physics from peer-reviewed literature (with DOIs where cited); and official records such as NOAA report cards and government supply-chain reviews. Market performance shown in articles is computed directly from real historical price data — not copied from a screenshot — and reported with its honest drawdown, not just its upside.

Where the data contradicts the narrative, the data wins. When a claim cannot be sourced, it is either labelled as reasoned inference or it does not go in.

How tickers are selected — and what that does not mean

This is the most important section to read carefully, because it is where most financial sites overclaim. The tickers in the Cascade Graph are mapped to a chokepoint by genuine business exposure — a company earns a place because its actual operations are levered to that node (a copper miner under the Copper node, an enrichment name under the SWU node), not because of any view that it is a good buy today.

• Thematic mapping, not stock-picking. Inclusion answers one question: does this security have real exposure to this constraint? It does not answer "should you buy it."
• Measured where charted. For tickers shown in a published article, performance and maximum drawdown are computed from real historical prices and stated honestly — including when a name underperformed or drew down roughly twice as hard as the S&P.
• Risk flags are explicit. Junior, pre-cash-flow, or single-asset names are tagged “junior · higher risk”, and the evidence panel lets you filter them out. The ETF-versus-single-stock tag is derived from each name's disclosed description.
• What we deliberately do not do. We do not screen tickers for current market cap, liquidity, balance-sheet solvency, or valuation, and we do not rank them as a buy list. The ETF/stock and risk-tier tags are derived from disclosed attributes — the site does not publish market-cap or liquidity figures it cannot source rather than invent them.

In short: the graph tells you where a structural pressure shows up in public markets and gives you an honest read on what owning it has historically felt like. It is a map of exposure, not a recommendation and not a portfolio. A formal financial screen (market cap, liquidity, solvency) is a separate effort we have deliberately deferred rather than fake.

The standing rule on honesty

Numbers shown to you are real and sourced, or they are explicitly labelled as illustrative or inferred. Drawdowns are shown next to returns. Counter-arguments and failure modes are kept in view, not buried. That discipline is the entire point of an anonymous signal — there is nothing to sell you except the reasoning itself.