Thirty-six names, four owners: Chinese state-enterprise concentration in African public procurement

AqlData — procurement insights · 52M public contracts (OCDS) · 5 African countries · 80 award records · FX-approximate values · aggregate, entity-level only

Five African countries believe they are contracting with thirty-six different Chinese construction firms. Entity resolution against public HKEX filings shows they are, in most cases, contracting with the same four state-owned holdings — all under direct Chinese state oversight through SASAC. The fragmentation is real; the common state ownership behind it is documented in public HKEX filings.

That is the structural gap this data surfaces: open contracting records are public, but the parent–subsidiary linkages that reveal concentration are not encoded in any OCDS field. The join requires a unified multi-country corpus and a corporate hierarchy layer. We have both.

The signal

AqlData cross-references 80 Chinese state-enterprise award records across five African OCDS corpora — Kenya, Uganda, Rwanda, Zambia, and Nigeria — against public corporate disclosure filings for CRCC (HKEX: 1186), CCCC (HKEX: 1800), PowerChina (HKEX: 2727), and CSCEC (HKEX: 3311). All four are majority-owned by SASAC, China's State-owned Assets Supervision and Administration Commission.

Aggregate finding: ~$1.56B in awarded contract value (FX-approximate), 80 awards, 5 countries, 36 distinct supplier names in procurement registries — resolves to 4 parent holdings.

Source: OCDS public procurement registries (Kenya PPIP, Uganda GPPB, Rwanda PPA, Zambia ZPPA, Nigeria BPP). Parent-entity mapping from HKEX Annual Reports and SASAC public registry. Value in USD is FX-approximate (single current exchange rate applied historically; directional, not precise). All figures represent procurement contract values, not investment or loan volumes.

What the fragmentation looks like in practice

CRCC — China Railway Construction Corporation, a $30B+ market-cap state enterprise — appears in Uganda's procurement registry under at least ten distinct names: "China railway seventh group co. ltd," "China Railway N03 Engineering Group," "China Railway 18th Bureau Group," "China Railway No.10 Engineering Group," "M/s China Railway Construction Engineering Group (CRCEG)," and others. In Kenya, the same group appears as "China Railways Seventh Group Company Ltd (CRSG)."

A standard string-match across the two countries' OCDS records returns zero hits. These are treated as separate suppliers in every single-country procurement analysis. They are not.

CCCC — China Communications Construction Company, HKEX: 1800 — presents a similar picture. In African procurement records it appears as China Road and Bridge Corporation, China Civil Engineering Construction Corp, CCECC, China Harbour Engineering, China Wu Yi, and several joint-venture styled names. Across Kenya, Uganda, Zambia, and Nigeria, we count approximately fifteen distinct name variants for a single Hong Kong-listed entity.

The entity resolution is not speculative: CRBC, CCECC, China Harbour, and China Wu Yi appear as subsidiaries in CCCC's audited HKEX annual reports. The linkage is public-record based, not inferred.

Three illustrative patterns

Uganda — road authority concentration

The Uganda National Roads Authority (UNRA) awarded 721 contracts (value > 0) totalling ~$2.16B in our corpus period. Chinese state-enterprise affiliates received 16 of those contracts — ~$547M, or 25.3% of UNRA's documented award value. Of that, $253M (11.7% of total UNRA) went to what the registry lists as ten separate firms, all of which resolve to the CRCC group.

A procurement auditor examining Uganda's roads portfolio in isolation would see a reasonably distributed supplier base. A cross-country, parent-resolved view sees a single state-owned group holding more than one-tenth of one agency's contract book.

Zambia — energy infrastructure stack, 2025

Between 2019 and 2023, Chinese state-enterprise awards in our Zambia corpus totalled approximately $5M. In 2025, that figure reached ~$378M — a roughly 70-fold increase in a single year, concentrated in energy and roads infrastructure.

The core of the 2025 signal: PowerChina International won a ~$173M EPC+Finance contract with Zambia's national electricity utility (ZESCO) for a 330kV transmission line. The same year, Huawei Technologies won a ~$5M contract with ZESCO for a central control and monitoring system across the same transmission infrastructure. Two separate contracts, two separate procurement registries, same underlying infrastructure. The physical build and the digital monitoring layer were procured from the same national ecosystem within twelve months. The monitoring system contract covers operational network management of the transmission infrastructure; this procurement record does not establish any intelligence-collection or surveillance function.

EPC+Finance structure means the contractor likely also intermediates Chinese policy-bank financing (China Exim Bank / CDB). The OCDS record does not name the financing source; that linkage requires cross-reference with AidData or BU Global Development Policy Center records, which we have not yet completed. This is flagged as an open verification item.

Rwanda — sustained single-firm dominance

China Road and Bridge Corporation (CRBC, a CCCC subsidiary, HKEX: 1800) has held approximately 27% of Rwanda's national roads agency (RTDA) works award value for eight consecutive years (2018–2026), totalling ~$169M on a base of ~$628M. 35% of CRBC's Rwanda awards used limited or direct-award procedures; the remainder were open tender.

Eight years of continuous presence at 27% share in a single agency, by a subsidiary of one holding group, is a concentration pattern that does not appear in any single-country procurement analysis.

Why this join is not available elsewhere

Every country's OCDS portal reflects its own suppliers, its own naming conventions, and its own registry fields. There is no cross-country supplier identifier in the OCDS standard. The parent–subsidiary linkage between "China Railway Seventh Group" (Uganda) and "China Railways Seventh Group Company Ltd (CRSG)" (Kenya) is not in any OCDS field anywhere.

Surfacing this pattern requires:

1. A unified multi-country OCDS corpus (we hold 52M records, 68 countries)
2. Normalized supplier names across country-specific formatting
3. A structured parent-entity map built from public HKEX filings
4. Temporal join: award date against contract period, not just registry presence

The infrastructure that makes this analysis possible is the same infrastructure that allows us to run equivalent queries for any entity group — BRI-related, sanctions-adjacent, or otherwise — across jurisdictions, in minutes rather than months of manual research.

Who this matters to, and why

Development finance institutions (World Bank, AfDB, AIIB, US DFC, UK BII): DFIs use government procurement data to assess market competition in sectors they co-finance. A market that appears competitive on single-country data may show significant concentration when parent entities are resolved. Debt-dependency risk and EPC+Finance co-financing patterns are measurable from the procurement layer — before credit exposure is booked.

Geopolitical and BRI research (CSIS, AEI, Chatham House, IISS, Mercator): BRI footprint estimates typically rely on loan databases (AidData, Boston University GDC) or news aggregation. Procurement award data adds a distinct signal: what was actually contracted, at what value, by which buyer agency, under which procedure. It complements debt-flow analysis and is independently verifiable from public registries.

Sovereign risk and country-risk functions: Concentration of critical-infrastructure award value in a single national-ecosystem supplier group is a structural risk factor. The Zambia energy-stack pattern — physical infrastructure and monitoring layer procured from the same origin within one budget cycle — is the kind of signal that informs infrastructure-dependency assessments.

Investigative and data journalism: Named state-owned enterprises, public procurement registries, verifiable OCDS identifiers, and HKEX disclosure cross-reference. Every data point in this analysis is reproducible from public sources. The join is the work; the underlying records are open.

Coverage: a data floor, not a ceiling

This analysis covers 8 African countries in our corpus (Kenya, Uganda, Rwanda, Zambia, Nigeria, Ghana, South Africa, Liberia). Chinese state-enterprise presence was confirmed in 5.

The following countries are not in our current corpus:

Tanzania, Ethiopia, Mozambique, and Angola represent the heaviest documented Chinese infrastructure investment on the continent. Their absence from this analysis means the $1.56B figure is a minimum baseline, not an estimate of total footprint — the true procurement footprint across these countries is materially larger, though we do not assign a multiplier without the underlying records.

We are explicit about this because the coverage gap is material to any analytical conclusion. What this data shows is what is visible in eight countries with functioning OCDS portals. It does not show what is not yet in the corpus.

Methodology and honest limits

Entity resolution approach: Parent-subsidiary mapping is drawn from HKEX Annual Report subsidiary schedules for CRCC (2023), CCCC (2023), PowerChina (2022), and CSCEC (2023). These are audited public filings. Mapping is conservative: entities were attributed to a parent only where the subsidiary name appeared in an annual report or is a well-documented trade name (e.g., "China Road and Bridge Corporation" = CCCC subsidiary, documented in HKEX filings and CCCC's own investor communications).

What the resolution does not establish: Whether a specific subsidiary was operating independently, whether its contracts were directed by the parent entity, or whether procurement decisions by African buyer agencies were influenced by state-level agreements. The data establishes contractual concentration; it does not establish coordination, direction, or impropriety.

False-positive controls applied: South Africa was excluded after identifying that "SINO"-prefix supplier names in ZA records are predominantly local South African entities (Sinolwazi = Zulu-language name). Nigeria small-value Huawei records were individually verified. Generic tokens ("China" alone) were not used as matching criteria.

FX note: All USD values use a single current exchange rate applied to historical local-currency amounts. For KES, UGX, and ZMW, historical FX divergence versus current rates is estimated at ±30–40%. Values are directional indicators, not precise contract amounts.

Procedural data gaps: 56% of Uganda UNRA award records in our corpus have a blank procedure field. Competitive-versus-direct breakdown for Uganda is therefore not fully verifiable from OCDS data alone.

This is a concentration signal, not an adjudication. Identifying procurement concentration by state-enterprise affiliates does not establish that procurement rules were violated, that contracts were improperly awarded, or that the contracting entities engaged in any unlawful conduct. Analytical conclusions require legal and regulatory review in applicable jurisdictions.

For full entity-level award trail, parent-subsidiary mapping, country breakdown, and reproducible query methodology: desk@data.heyvaql.com

Source: Open Contracting Data Standard (OCDS) public procurement registries, 5 African countries, 80 award records. Corporate hierarchy: HKEX Annual Reports (CRCC 1186, CCCC 1800, PowerChina 2727, CSCEC 3311) and SASAC public registry. All entity matches are corporate/institutional; no natural persons are named or profiled. AqlData is an independent data compilation; not affiliated with any government, regulator, or official body. This document does not constitute legal, compliance, or investment advice.