Tracking Who Owns What: 12,875 Ownership Events Across the Concession Universe

Last updated: August 5, 2026

Concession ownership is not always what the equity press release says it is. A single toll road can be held through three layers of SPVs, a consortium bid can split a stake five ways, and a stake that trades quietly inside a listed infrastructure vehicle may never generate a headline at all. This is a look at how Concessio tracks who actually holds what, and why the honest answer to "who's the most active buyer in infrastructure" is that the data isn't clean enough yet to rank.

What counts as an ownership event?

An ownership event in Concessio is a recorded acquisition or disposal of an equity stake in a state-granted concession — a change of controlling or minority interest in the private-sector counterparty to a concession contract, not a change in the underlying asset's operations. Concessio has recorded 12,875 ownership events across the concessions it tracks. Events are linked to a specific concession record and carry a date where the source supplies one.

An ownership event is distinct from a transaction record. A transaction carries the commercial terms of a deal — enterprise value, EV/EBITDA, EV/revenue where disclosed. Concessio separately tracks 6,596 transactions. 210 transactions carry a disclosed EV/EBITDA multiple (see the table below). Not every ownership event has a matching disclosed-terms transaction, and not every transaction is a change of control — some are refinancings or partial stake top-ups.

Why is concession ownership harder to track than ordinary corporate M&A?

Four structural features of the asset class make concession ownership unusually resistant to a clean, single-source answer:

• SPV layering. Almost every concession is held through a special-purpose vehicle created for that specific contract, and that SPV is frequently owned by a holding company, which is in turn owned by a fund vehicle. A stake sale at the fund level may not be visible at the SPV level, and vice versa.
• Consortium stakes. Concessions are routinely won by bidding consortia rather than single sponsors. A "sale" of a consortium's interest is frequently a sale by one member, not the whole group — collapsing that into a single blended record destroys the information a diligence team actually needs.
• Secondary transfers. Infrastructure equity trades in a secondary market that is much quieter than listed equities — stakes move between funds, pension co-investors, and strategic operators years after financial close, often disclosed only in a regulator filing or a competition-authority clearance, not a press release.
• Listed-vehicle disclosure. Where the concession sits inside a publicly listed infrastructure company or trust, ownership changes at the shareholder level may be disclosed to a stock exchange without ever naming the underlying concession, requiring the two records to be manually reconciled.

What sourcing standard applies to an ownership change?

Concessio's sourcing standard does not treat a press release, an aggregator summary, or a fund's own portfolio page as sufficient evidence of an ownership change on its own. The primary sources it accepts are:

• Competition-authority filings — merger-clearance decisions that name the acquirer, target, and stake, filed as a condition of closing.
• Stock-exchange disclosures — filings to SEC EDGAR, HKEX, and peer exchanges where the buyer, seller, or the concession-holding vehicle itself is listed.
• Regulator approvals — sector regulators (transport authorities, energy regulators, port authorities) frequently must approve a change of concessionaire control before it takes effect, and that approval is itself a primary record of the transfer.

Where a concessionaire's own official channel documents a transaction it was party to — a stock-exchange announcement, an annual report — that counts as a first-party primary source. A third party's marketing page describing an asset it merely holds does not.

That's the standard the pipeline is built to. It is not the same thing as a published compliance figure: unlike the clause library, where Concessio publishes exactly what share of visible records are source-cited and banned-source-free, no equivalent screened panel exists yet for ownership events or transactions. Concessio has previously had to remediate ownership records that an internal linking tool inferred rather than sourced — so an unqualified "every event is independently sourced" claim isn't one we'll make until that audit exists.

How deep is the transaction evidence behind these events?

Not every ownership event carries a disclosed price. The subset that does gives a sense of how much commercial detail is actually recoverable from primary sources, rather than left as a bare acquirer/target/date record.

Two individually-documented examples show what that paper trail looks like when the enterprise value is fully matched to a citable primary source: the Toulouse-Blagnac Airport concession in France, sourced to the French government legal publication service (legifrance.gouv.fr), enterprise value US$770 million (US$100.5 per passenger); and the Port of Brisbane 99-year lease in Australia, sourced to the Queensland government's official announcement record (statements.qld.gov.au), enterprise value US$2,096 million (US$2,141 per TEU). Treat both as single named transactions, not a sample — with this few matched examples, Concessio does not average or extrapolate a per-unit range from them.

How does Concessio handle consortium stakes?

The goal is one ownership event per equity holder in a bidding consortium — each tagged to that holder's actual stake — rather than a single blended entry for "Consortium X," because a blended entry can't answer a diligence question like "which of the three consortium members actually increased its position." That disaggregation is not complete across every historical consortium record in the database: some older entries are still recorded as a single blended holder string, and unwinding them into per-holder events is ongoing work, not a finished state.

Does Concessio publish a ranking of the most active acquirers?

No. Concessio can report the volume of ownership events (12,875) and the depth of disclosed transaction terms (6,596 transactions, 210 with a disclosed EV/EBITDA multiple) with confidence, because those are direct counts of records in the database. A ranked list of "most active buyers" is a different kind of claim: it requires the buyer-name field to be clean and de-duplicated across every SPV, holding-company, and fund-vehicle name variant a single sponsor might transact under, and that name-resolution work is not yet at a standard we'd put in front of an infrastructure investor doing diligence on a counterparty. Until it is, we report coverage and sourcing standards, not league tables.

Figures as of August 2026. Published August 5, 2026. Full sourcing methodology: /docs/methodology.

Request Demo