Latin America holds roughly US$1 trillion of pension assets across five countries, on OECD data, and runs a deep tech ecosystem with an unclosed Series A to C financing gap. That those two facts should meet is a proposal the region has heard before. What the meeting looks like in institutional practice, who signs what and in which order, has been the open question. As of 19 February 2026 there is a working answer to it, and it comes from Paris.

France’s Direction générale du Trésor reports that the Tibi initiative has reached EUR 14 billion invested, has homologated 160 funds, has raised its overall goal from EUR 13 billion to EUR 15 billion, and has launched a third phase with a European ambition. That is a validated instrument, not a proposal. This post walks through the design that produced those numbers, the assets Latin America already holds against them, and the single institution the region still has to name before it can copy the thing.

How Tibi Was Built, and in What Order

The policy was outlined in the 2018 Tibi report to the French Ministry of Economy and Finance, which framed the problem as a financing gridlock for technology companies rather than as a shortage of good science. The instrument that followed is unglamorous and exact. A treasury builds an approval process for funds. Institutional investors, pension and insurance money included, then commit capital against the approved list. Accreditation comes first and capital comes second, and that ordering is the whole trick: the state never picks a company, and the trustee never has to underwrite a category it has no staff to assess.

Under Tibi’s first phase, covering 2020 to 2022, institutional investors committed EUR 6.4 billion to approved funds, on DG Trésor’s own reporting. The February 2026 update is the part worth carrying into a policy memo: EUR 14 billion invested, 160 funds homologated, the overall goal lifted from EUR 13 billion to EUR 15 billion against a Phase 2 target of EUR 9 to 10 billion, and Phase 3 opened with a European ambition. One note for anyone quoting the programme’s size: use the EUR 14 billion invested, which is capital actually deployed, rather than the EUR 40 to 50 billion in technology funds under management that the French financial marketplace projected for 2026, because assets under management across approved funds and capital deployed answer different questions and should not be read as one ascending series.

Accreditation is the product. The capital follows the list.

The UK Is Running the Same Play, One Stage Earlier

The United Kingdom’s Mansion House Accord of May 2025 pledged to allocate 10 percent of pension provided portfolios to private markets by 2030, with at least 5 percent ring-fenced for UK assets, framed as voluntary, fiduciary-aligned and monitored. We carry those figures as we extracted them from the BVCA’s 2025 material and have not re-verified them since. Complementing the Accord, the BVCA’s 2025 expert panel proposes a UK programme named NOVA, modelled directly on Tibi, to accredit eligible venture and growth funds and stand up a fund-of-funds.

Read the two jurisdictions on their own clocks. France has a result a trustee can check today. The UK has a pledge signed in May 2025 with a 2030 horizon and an accreditation programme still at the proposal stage, which makes it early by construction rather than disappointing: its evidence arrives at the end of this decade. For a Latin American ministry building the case now, France is the evidence and the UK is the live design study, and both are more useful than a hypothetical.

What Latin America Already Has

Start with the pool. In Chile, Mexico, Brazil, Peru and Uruguay, pension assets under management total roughly US$1 trillion, on the OECD’s Pension Markets in Focus using preliminary 2024 data published in June 2025, a figure we carry as extracted from that edition and not yet re-verified against a newer one. Redirecting even 1 percent of it would mobilise US$10 billion for domestic deep tech. Set that against the region’s cheque sizes: median round size scales from US$0.6 million at seed to US$8.3 million at Series A and US$17 million at Series B on our own Tracxn-based mapping to a March 2025 cutoff, also carried as extracted rather than re-verified. Next to medians of that size, one percentage point of the pension pool is an enormous number, which is why this recommendation keeps coming back.

Then the manager supply, which is the part sceptics assume is missing. By 2023 there were 65 venture capital funds with at least one deep tech investment in Latin America, on the IDB’s count, again as extracted from that mapping. Some of them look exactly like the specialists a Tibi-style list would homologate. GRIDX has raised US$41.5 million across two funds and built a portfolio of 81 companies, 75 percent of them co-founded by women, spanning Argentina, Uruguay, Mexico, Colombia, Brazil and Chile and collectively employing 1,000 people including 700 scientists, on the figures our report carries for the firm from 2024. Its Fund I closed at US$10 million in 2018; Fund II added US$30 million in 2022 with IDB Lab joining as a limited partner, and its companies have since raised over US$100 million from international investors. In Brazil, Vesper Ventures evaluated approximately 4,500 scientific projects, chose to co-found 8 companies, and reports 16 patents, more than US$30 million raised and a team of more than 50 PhDs, figures the firm gave us directly.

That IDB Lab commitment matters more than its size. It is the anchor limited partner function that Tibi asks a public balance sheet to perform, already performed once in the region, on a Latin American deep tech fund, with a development bank’s balance sheet. The accreditable universe is not hypothetical. It exists, it is deploying, and it can be diligenced fund by fund.

The Accreditation Seat Is the Work Ahead

Here is where a regional version stops being a translation exercise. Fund approval works in France because one treasury holds a continuous supervisory relationship with one national financial marketplace and can accredit funds and set commitments without renegotiating anything with a second jurisdiction. Chile, Mexico, Brazil, Peru and Uruguay sit under five separate regulators with different fiduciary rules and different permitted-asset regimes, and no shared accreditation authority exists among them. Recommendation 12 in Accelerating Deep Tech in Latin America, published in September 2025, calls for an independently managed fund-of-funds mobilising a slice of pension assets through vetted specialist GPs, mirrored with development banks and multilaterals. Who holds the accrediting pen is the first design decision, and it is the one still unmade.

Be honest about the base rate the ask has to move. Across Europe and the United States, pension funds accounted for under 5 percent of the top-50 deep tech investors by deal count by 2024, on Atomico’s State of European Tech, extracted and not re-verified by us. Note the basis before using it: it counts deals inside a top-50 investor set, so it says nothing about capital volumes or about the share of pension assets allocated to the category. Even read that narrowly, it tells you pension participation in deep tech is something treasuries manufacture rather than something markets produce on their own. Tibi is what the manufacturing looks like when it works, and nothing on our record yet shows the same approval machinery running across five sovereign pension regimes at once. That coordination is the measurement LADP is building next: accreditation criteria that survive five fiduciary codes, and a running read on how Phase 3 actually disburses.

What To Do With This Now

For finance ministries and pension supervisors, the near-term deliverable is the accreditation architecture rather than the fund itself. Publish the criteria, name the body that applies them, and let capital follow the list, because that is the order in which France did it and the order in which the UK’s proposed NOVA would do it. A development bank or multilateral is the only candidate accreditor with standing across the five markets, and IDB Lab has already sat on the LP side of exactly this kind of vehicle.

For trustees and allocators being invited in, the comparison to run is EUR 6.4 billion committed by institutional investors in Tibi’s first phase, EUR 14 billion invested by February 2026, 160 funds homologated and a third phase live, alongside a UK pledge whose results land by 2030. For managers, the practical implication is to become accreditable before there is anything to be accredited by: audited track record, fiduciary-legible fund terms, reporting a supervisor can read. LADP is convening the trustees, supervisors and specialist managers this design requires, because the coordinated effort to do so did not exist when we went to press. If you sit on any of those three sides, write to us. We would rather build the list with you than describe it.