Every element, sized by the economy it stands under rather than by its atomic weight. For each one: how it actually comes out of the ground, the form it is really traded in — sulfur as sulfuric acid, nitrogen as ammonia, titanium as white pigment — what a year of world extraction is worth, and the first-order markets that cannot exist without it. The ratio between those last two is the point.
| Z | Sym | Element | Traded form | Upstream | Downstream | Leverage | Conf. |
|---|
Upstream is world annual production times a representative price at the first marketable form. Production and price come from the USGS Mineral Commodity Summaries 2026 wherever that publication covers the commodity, which is about seventy of them; the rest are from industry bodies, named on each element. Quantity and price are always quoted on the same basis — element content, oxide content or gross ore — because mixing them is the fastest way to publish a wrong number.
That is the whole reason this table is not just a price list. Almost no sulfur is bought as sulfur: about 90% of it is consumed as sulfuric acid, and roughly half of that acid goes into digesting phosphate rock. Nitrogen is bought as ammonia and urea. Nineteen of every twenty tonnes of titanium becomes white pigment rather than an aerospace part. Sodium is really chlor-alkali. Each record names the form the element is actually traded in, and prices that form separately where a market for it exists.
Downstream is the annual revenue of the first-order product markets in which the element is a non-substitutable input — steel for iron, semiconductors and solar modules for silicon, lithium-ion cells for lithium. It deliberately stops at one hop. Following iron all the way to "buildings and vehicles" would reach most of world GDP and would double-count every other element in the same car. Where the wider claim is the interesting one — nitrogen underwriting roughly half the protein in living humans, for instance — it is stated in that element's note rather than smuggled into the number.
Upstream figures are good to the significant figures shown. Downstream figures are not:
published market sizes for the same industry routinely differ by a factor of two or three
between research houses, and the attribution rule above is a judgement, not a
measurement. Every element therefore carries a confidence tier, shown in its panel:
high means both ends are sourced, med means the downstream is a
reasoned estimate, low means read the exponent and ignore the mantissa, and
nil means no extraction economy exists at all. About a quarter of the table
is nil — everything above fermium is made one atom at a time in an
accelerator, and the honest economic answer for those elements is zero.
The full dataset, including the attribution rule for every figure, is in
finance/ptable/elements.js, and the method is written out in
finance/ptable/METHOD.md. Corrections are welcome and expected.