Latin America’s lithium story is usually told as a map that is left to do the arguing. Argentina, Bolivia and Chile sit on the Lithium Triangle, the U.S. Geological Survey puts an estimated 58% of the world’s identified lithium resources there, and the conclusion is supposed to follow on its own. It does not. The USGS Mineral Commodity Summaries 2026, published in February, are worth reading closely for a different reason: they show how much of the value in that endowment has not been built yet, and what building it would take. The opportunity is an industry above the brine, not a claim beneath it.

What the 2026 Vintage Puts on the Table

USGS keeps two categories apart, and that distinction is the most useful thing in the document. Identified resources are what the geology holds. Reserves are the portion that can be produced economically at today’s prices with today’s processing. World reserves in the 2026 summaries come to roughly 37 Mt. On identified resources, USGS puts Bolivia at 23 Mt, Argentina at 22 Mt and Chile at 11 Mt. On reserves, Chile dominates and Bolivia holds almost none.

Those tonnages reach us through a USGS-based compilation rather than from the USGS release itself, so we carry them as transcribed. The same compilation also puts the three Triangle countries at roughly 97% of world identified resources, a share the 58% cannot be reconciled with, and from the transcription we hold we cannot settle which of the two is right. The 58% is the figure our published record carries, so it is the one quoted here. Nothing below turns on a third digit. It turns on the ratio: Bolivia’s 23 Mt of identified resources is well over half of the entire world reserve base of roughly 37 Mt, and resource tonnages and reserve tonnages are not one quantity measured twice.

The 58% travels in both nouns, so be firm about which one you are quoting. The USGS figure is identified resources, including in the places where our own report’s foreword prints the word reserves. The value holds either way; the noun tells you whether you are looking at geology or at production capacity, and only one of the two can be sold this decade.

How a Resource Becomes a Reserve

This is the part that should make an investor sit up rather than sigh. Reserves are not a geological constant. They are a function of price, permitting and processing technology, and they are restated whenever those move. Bolivia’s 23 Mt of identified resources are not producible today. A decade of stronger prices, working processing capacity and reformed permitting could move part of that tonnage into the reserve column, and that movement is not a windfall. It is a construction project with a cost, a duration and an owner.

Buyers contract against reserves, because reserves are what a plant can ship. When a cell manufacturer signs offtake, the question is which permitted operation has tonnage available, and today the answer is set by a reserve table Chile leads. The work separates by capital. Chile holds a position that already trades, which makes permitting speed the live variable. Bolivia holds a resource position, so processing capacity and permitting terms are what have to be built before it can be spent. Argentina, at 22 Mt of identified resources, sits between them. Same geology, three work programmes, all of them above the brine.

The Value Chain Above the Brine Is Already Being Financed

Middle powers have worked this out and are buying the conversion rather than the endowment. In 2024 Japan and Brazil upgraded their strategic partnership to address energy transition needs, and their Joint Action Plan pledges to strengthen mineral value chains by adding value to Brazil’s mining output and sharing technical knowledge on recycling and processing, battery-grade lithium refinement among the named examples. Japan has already co-invested in LATAM lithium projects, Panasonic in Bolivia and Toyota in Argentina, and the Japan International Cooperation Agency invested US$1 billion into a private-sector fund, the JICA-TADAC Fund, managed by IDB Invest, a figure our report carries from the IDB.

South Korea is supporting Chile on battery minerals and the energy transition, including developing lithium processing capacity and modernising mining operations, with MoUs pursued on battery minerals and hydrogen. Chile and Korean partners have been collaborating on lithium joint ventures with the state firms Codelco and ENAMI, and Korean industry is investing in projects that include local processing to add value. A 2021 MoU between the two countries covers green hydrogen, leveraging Chile’s renewable resources. Chilean mines, meanwhile, are adopting Industry 4.0: autonomous trucks, AI-based ore-sorting and integrated operations centres.

The Gulf is in the same trade. In January 2025 Brazil signed a US$2.5 billion MoU with the UAE focused on strategic minerals including copper and lithium, concluded between Brazil’s Mines and Energy Ministry and the UAE’s Ministry of Investment, with explicit provisions for the transfer of cutting-edge technologies for exploration and processing. That figure reaches us through Net Zero Circle’s 2025 reporting rather than from the parties.

One input makes this more than diplomacy. Refining is power-hungry, and 65% of LATAM’s electricity comes from renewable sources against a global average of 41%, the Ember figure our report carries among the region’s endowments. Cheap clean power next to the resource is a processing advantage, and it is why a battery-grade refinery has a better argument for being built in Chile or Argentina than somewhere that has to import both the concentrate and the electricity.

Who Is Already Building It

The companies that will service this conversion are not lithium companies, which is why they get missed. Splight, out of Chile, runs grid-operations AI that tackles curtailment and congestion to maximise renewable integration, at TRL 9. We record US$26.1 million in total funding, including a US$12 million seed led by noa with EDP Ventures and the UC Berkeley Foundation, and we carry that total as extracted rather than independently checked. Grid headroom is the binding constraint on any refining build-out.

Tractian, with Brazilian roots, integrates hardware, software and AI to democratise predictive maintenance for industrial operations, at TRL 9, and our record carries a US$120 million Series C led by Sapphire Ventures with General Catalyst and Y Combinator among its backers. Continuous processing plants live on uptime. FabNS, a UFMG spinoff in Belo Horizonte, builds Tip-Enhanced Raman Spectroscopy for chemical imaging at nanometric resolution and has been recognised in the MIT Innovators Under 35, which is the instrument layer a materials industry needs at home rather than on a shipping manifest.

Resource-sector capital is moving into this class of company already. CMPC Venture Capital, the corporate venture arm of a Chilean multinational, led a EUR 4.8 million seed round in 2023 for Strong by Form, the Santiago team making bio-inspired structural wood components that can replace steel, concrete and aluminium, a figure we carry from CMPC. The country-level capital map rewards attention too: EMERGE’s Radar Deep Tech LATAM 2025, presented on 11 September 2025, records 2024 private deep tech investment of US$607 million in Chile and US$486 million in Argentina, which we cite from EMERGE rather than from our own count.

The Number That Would Price the Triangle

Here is what our record does not yet hold, stated plainly because it is the next thing we intend to build. We have no country-by-country conversion rate from identified resource to producible reserve, with the permitting, processing and offtake cost and the duration attached. That is the figure that would price the Triangle, and a fund raised against the 58% without it is selling geology rather than an asset. Two items sit behind it: the USGS tonnages need transcribing from the primary data release, and reserve tables are restated as prices move, so this belongs in a tracked series.

Our report proposes the vehicle for exactly this kind of gap, an open-access data commons of standardised statistics, deal and IP registries and dashboards that cut diligence friction. A conversion table for critical minerals is a strong candidate for an early release, and it is the sort of public good no single ministry or fund has an incentive to build alone.

What to Do With This

For founders, the addressable market above the brine is the larger one: processing and refining, water and brine management, power and grid, autonomy and ore-sorting, and the instrumentation that lets a plant certify its own output. These are deep tech businesses whose first customer already exists at home.

For allocators, the instrument to size is one that can underwrite permitting risk and plant capital, not exploration alone. The conversion has a duration a venture fund cannot carry by itself, which is why our recommendations point at sovereign co-investment vehicles, matching funds and upgraded special economic zones that turn long-term public R&D budgets into patient capital, and at pension money brought in through vetted specialist managers.

For ministries, the reserve table is the reason a regional minerals strategy has to start from shared geology and asymmetric commercial positions rather than from a single bargaining posture. The work that pays is domestic: permitting throughput, processing capacity, power, and technical capability of the kind Chile is assembling through its roadmap for a quantum technology sector with applications in mining, renewable energy and cybersecurity. Our report recommends a standing working group to map technological flows and broker corridors with superpowers and middle powers, through special economic zones, joint labs, procurement pilots and standards MoUs, publishing the heatmaps and deal playbooks openly.

We are building the critical-minerals conversion data into the next phase of LADP’s work, and we would rather build it with the people who hold the operating numbers. If you are financing a processing plant, running a permitting queue, or selling the instruments that make either one work, we want your figures in the series. Write to us, and read the report at ladp.io.