Draft 1 · worked exemplar for discussion · this will not be the final allocation  ·  all 54 draft bundles →
Flag of the Kingdom of Eswatini
Kingdom of EswatiniBuy African Initiative · Right of Supply
Ministerial review · v4
15 July 2026
One member worked in full · a method for the other 53

The USD 620 billion Africa sends abroad

The government beachhead has two honest bands: USD 136.75 billion measured, already including parastatals, and about USD 207 billion government-influenced when the Minister's one-third ratio captures contractor-imported tenders. Eswatini's claim is deliberately smaller: two plants, 14 lines and a screened basket of about USD 1.92 billion.

USD 620 bn
Off-continent prize from USD 709 bn total imports
137 → ≈207 bn
Measured floor → government-influenced estimate
USD 1.92 bn
Post-review Eswatini basket
2 plants
Ceramics · niche steel
54
Members allocated by one method
The Minister’s brief · Eswatini · worked in full
Minister, Eswatini holds what almost no African state can offer — a finished intermediate the continent cannot yet make for itself: the Matsapha concentrate plant that has supplied sixty bottling companies across twenty African countries since 1987, your single largest export earner, anchoring the category that reached USD 650.6 million, some 27.1 per cent of exports, in 2024. Africa ships billions of dollars abroad every year to buy back processing it could own; your concentrate, your refined sugar from the Mhlume refinery, your value-added timber are Eswatini's rightful claim on that outflow. The Right of Supply gives you a twenty-five-year first right to serve that demand — never a subsidy, never a captive contract, because Match-or-Release keeps it yours only while you meet the market's best price and terms. This is Draft 1, deliberately provisional, built to be corrected — and your correction, Minister, is the next move.
Right of Supply · the member worked in full · a method for the other 53
01 · Correspondence
To the Minister

Your second review, answered in the architecture

15 July 2026 · Mbabane / Johannesburg

Minister,

Thank you for the second review. It sharpened the two places where this framework most needs to survive contact with government: the true size of the state-driven market, and the discipline by which a country is matched to a finished product.

On the first point: yes, parastatals are already inside the measured USD 136.75 billion floor. The sovereign database is built on named, fully state-owned or controlled buyers and monopoly functions. NNPC and EGPC sit in energy; GASC, OAIC and NCPB in strategic food; KEMSA, NMS and PCT in public health, alongside central banks, electoral commissions, defence ministries and public-works authorities. The deeper truth in your challenge is also right. Customs data sees the direct sovereign import, but it often misses the hospital tile, state-road rebar or utility equipment imported by a contractor. Applying your own one-third ratio to the USD 620 billion prize implies about USD 207 billion of government-influenced demand. This version therefore shows both bands everywhere: USD 136.75 billion measured, and the additional tender-driven layer clearly labelled as estimated.

On the second point, your Botswana copper example now becomes three allocation principles. We allocate the tender-specified finished good rather than an intermediate; we match the largest use to the largest endowment-holder; and we favour combinations of endowments that meet in one buildable plant. Eswatini has been re-screened accordingly. Porcelain tableware is out on tender-materiality; bulk structural sections are out; and the continental rebar anchor is ceded to larger iron-and-coal holders, with only a capped regional wedge retained to keep Eswatini's basket in balance. The core claim is now sanitaryware and tiles, fabricated structures, pipe and gas containers. Botswana's own action is logged for its build: electric motors in; bulk insulated copper cable anchored to Zambia.

Your first instruction also remains intact. Raw-material bases stay out. Kaolin, anthracite and iron only enter this document where they terminate in a plant and a finished product a public buyer can specify.

We ask for your reaction to two things: whether the measured-versus-estimated government layer now states the truth cleanly; and whether the three allocation principles capture the judgement you intended. Version five follows your word.

The Right of Supply research teamhuman analysts and AI research agents · for the Office of the Chair, AU STC-FMAEPI
02 · Executive summary
The whole case on one page

A measured beachhead, a fair allocation, a buildable claim

Right of Supply · in one read

The prize is continental. The claim is disciplined.

620USD bn leaving Africa
137USD bn measured floor
≈207USD bn influenced
2Eswatini plants

The prize. Africa imports USD 620 billion of goods from outside the continent. That is the market available for progressive import substitution.

The beachhead. USD 137 billion is measured direct sovereign demand and already includes parastatals. The tender-driven layer is larger. Using the Minister's ratio, total government-influenced demand is about USD 207 billion. The distinction is explicit: measured versus estimated.

The instrument. Allocation gives each member the first right to compete for a defined product. The Cell Phone Test protects the buyer: match price, quality, warranty and service, or release the order. Procurement is never pooled and no supplier receives a captive contract.

Eswatini. Two plants form the claim. Kaolin becomes ceramics and sanitaryware. Anthracite and iron meet in a niche steel route. The screened basket remains about USD 2 billion, close to what the kingdom imports. Balance is designed in.

What changed. The government layer now has two bands. Allocation now follows finished-product, scale-matching and endowment-combination principles. Tableware is removed. Bulk steel is ceded or capped. Raw-material bases remain outside the instrument.

03 · Version thread
The argument, improved in public

Four versions. Each review removed an easy claim and strengthened the method.

v1 · 13 Jul

“The USD 43.5 bn question.”

Full-demand scan: 114 HS6 lines, six endowment chains, every buyer named. The Minister's read: right ambition, wrong denominator — raw bases and private demand inflate the claim.

v2 · 14 Jul

“The focused basket.”

His four screens applied (government · off-continent · industrial · balance); raw bases removed; basket ~USD 2.2 bn ≈ Eswatini's own imports.

v3 · 14 Jul

“The continental prize.”

Reframed on the USD 620 bn off-continent prize; government layer measured at USD 136.75 bn (sovereign database, parastatals included); the funnel spine; Eswatini as the worked template for all 54.

v4 · 15 Jul

“The Minister's second review.”

Two-band government layer (~USD 207 bn government-influenced by his own ⅓ ratio); the three allocation principles; the basket re-screened by the tender test; this letter, summary and thread added.

The continental denominator

Start with what leaves Africa. Then narrow only where government can act.

04 · The size of the prize
709 → 620 → 137 / ≈207 → 1.92

USD 620 billion leaves the continent every year

Africa imports USD 709 billion. USD 620 billion comes from outside the continent; about USD 89 billion is sourced within Africa. The Right of Supply begins with the off-continent prize, then narrows to the government demand a signature can redirect.

USD 709 bn
Total continental imports — all buyers, all sources
The market
USD 620 bn
From outside Africa — the import-substitution prize
The prize
USD 137 bn
Measured direct sovereign imports · parastatals included
Band A
≈USD 70 bn
Tender-driven extension · estimated
Band B
USD 1.92 bn
Eswatini's screened, buildable basket
Worked case

Funnel widths are indicative; Eswatini is enlarged for legibility. Band A is measured at USD 136.75 bn (2024). Band B is inferred from the Minister's one-third ratio: one-third of USD 620 bn ≈ USD 207 bn total government-influenced demand, less the measured floor ≈ USD 70 bn.

SCREEN 01

Government first

Begin where a state, agency or parastatal controls the tender or directly imports the good.

SCREEN 02

Off-continent

Substitute imports from beyond Africa. Do not displace an existing African producer.

SCREEN 03

Industrial

Allocate a finished good that requires plant, capability and jobs—not a raw base.

SCREEN 04

Balance

Size the right near what the member buys so the continental allocation can net out.

05 · Government beachhead
Two bands · one honest market

USD 136.75 billion measured. About USD 207 billion government-influenced.

The measured floor already includes state-owned buyers. The estimate above it captures goods specified by government but imported through EPC contractors, construction firms and other delivery vehicles that customs cannot label as sovereign.

Band A · measured floor
USD 136.75 bn

Direct sovereign imports across 62 categories and 48 states. Parastatals and controlled agencies are already present.

NNPCGASCOAICNCPBKEMSANMSPCT
Band B · estimated tender layer
≈USD 70 bn

The implied extension between the measured floor and ≈USD 207 bn total government-influenced demand. Based on the Minister's ratio, not a customs measurement.

Parastatals, confirmed

The sovereign database uses monopoly, controlled, predominant and sole-lawful-buyer classifications tied to named agencies. Energy includes NNPC and national oil companies; strategic food includes GASC, OAIC and NCPB; public health includes KEMSA, NMS and PCT. Central banks, electoral commissions, defence ministries, public works and roads authorities complete the government layer. The question is not whether SOEs are counted. They are. The question is which contractor-imported tenders sit above the directly measured floor.

The seven sovereign pillars · USD bn, 2024

Energy & utilities71.65
Strategic agriculture & food27.03
Public health & pharmaceuticals14.02
Digital sovereignty & telecoms8.53
Infrastructure & transport6.60
Currency, governance & elections5.54
Defence & national security3.40

The direct sovereign floor

Fuel, grain, medicines, defence and other lines imported by a state or controlled agency.

The
tender

The contractor-imported layer

The hospital's tiles, the state road's rebar and the utility's pipe. Government specifies the finished material even when a contractor clears customs.

06 · Allocation logic
Three principles from the Minister's review

Allocation chooses the finished product—not merely the resource

The Minister's Botswana example turns judgement into a repeatable method. It decides which product belongs to which country before any demand total is claimed.

Principle one

The finished-product test

Allocate what the tender actually specifies. A government does not buy winding wire to repair motors; it buys motors. A hospital fit-out buys sanitaryware, not raw kaolin.

Pass: electric motors · sanitaryware
Fail: winding wire · raw kaolin
Principle two

Scale-matching

The largest use of a material anchors to the largest endowment-holder. Smaller holders take a niche, higher-value product rather than a continental bulk line.

Scale anchor: bulk copper cable → Zambia
Niche: motors → Botswana
Principle three

Endowment-combination

The strongest claim brings two endowments together in one plant. The allocation rewards the industrial combination, not the mere presence of either resource.

Eswatini: iron + anthracite → one niche smelter
EndowmentKaolin
IntermediateCeramic body
Tender-specified productSanitaryware
Named buyerHealth / public works
07 · Discipline
What came off Eswatini's list

A credible basket is defined as much by what it refuses

Every exclusion names the screen it fails. This is the visible evidence that the basket was chosen rather than padded.

Raw & refined sugar

Many African producers already supply it; Eswatini sells what it makes and cannot cheaply expand. No new industrial capability.

Regional incumbency · raw base

Sawn pine timber

A raw product already supplied on the continent. Shipping it as-is does not create manufacturing.

Industrial screen

Paper, pulp, board & tissue

Mondi and Sappi are world-scale incumbents; Eswatini's pulp mill is gone. Allocation would create conflict, not a defensible plant.

Incumbency · capability

Beef, meat & leather

Widely produced across Africa and over-supplied for this instrument. The route remains too close to a raw commodity.

Off-continent · industrial

Cotton, textiles & apparel

Predominantly private demand and already supplied regionally. Government cannot exercise the right at sufficient scale.

Government screen

Beverages, beer & concentrates

Consumer and private-sector demand. Public buyers do not procure these at the scale required by the instrument.

Government screen

Porcelain tableware · HS 691110

Institutional catering is real but marginal. It fails the tender-materiality test introduced by the second review.

Tender-materiality

Bulk structural sections

U, I, H and angle sections are scale lines better anchored to larger iron-and-coal holders. Eswatini keeps fabricated structures instead.

Scale-matching

Continental rebar anchor

Eswatini does not claim the bulk continental line. A capped regional wedge remains only because it brings the total basket close to balance.

Ceded anchor · capped wedge
08 · Endowment
What survives the screens

Three inputs. Two plants. No raw-material claim.

Cane and timber are deliberately absent. The surviving endowments matter only as feedstock, reductant or energy option inside a finished-product plant.

Illustrated map of Eswatini showing endowment geography
The terrain. Kaolin and mineral ground in the west and Highveld; Maloma anthracite and the Lubombo in the east. Stylised for orientation.

Kaolin

Emerging

Deposits identified, not yet worked at scale. Feedstock for sanitaryware and ceramic tiles.

Anthracite

Operating

Maloma colliery. Reductant for the niche steel route; possible kiln-energy option for ceramics.

Iron ore

Aspirational

Historic Ngwenya workings, partly depleted. A domestic smelter therefore remains a heavier, capability-building proposition.

Physical map of Eswatini
Physical geography. Maloma and the Lubombo anchor the coal-and-mineral east; historic Ngwenya sits in the north-west.

Kaolin + on-soil energy: an option, not a technical claim

Maloma anthracite could provide an energy-security or cost option for the ceramics line. Modern sanitaryware kilns are typically gas-fired, so fuel choice requires engineering and emissions validation. The document does not assume anthracite is the default kiln solution.

09 · Focused basket
Two beneficiation plays · 14 lines

Every HS6 line now terminates in a named public tender

Card values are gross 2023 continental imports. Addressable values apply the unchanged government-intensity and off-continent/substitution factors. They remain estimates pending the bilateral trade re-pull.

Plant one · near-term

Ceramics from kaolin

Ceramics & sanitaryware

5 HS6 lines · kaolin feedstock · government fit-out tenders
USD 1.84 bngross · ≈ USD 483 m addressable

The finished-product claim

Hospitals, clinics, schools, public housing and government buildings specify sanitaryware and tiles in their fit-out schedules. The allocation attaches to those finished goods—not to kaolin.

Ceramic sanitaryware

USD 400 mgross demand
EmergingGov highSubst high
Screened
≈195 m
691010Porcelain sinks, basins, baths, WCs and related fixturesHealth & public-works fit-outs263 m
691090Other ceramic sanitary fixturesHealth & public-works fit-outs137 m
Leading importing states
Nigeria 102 mLibya 28 mGhana 22 mMorocco 21 m

Ceramic tiles & paving

USD 1.44 bngross demand
EmergingGov mediumSubst medium
Screened
≈288 m
690721Low-absorption porcelain or vitrified tilesPublic works, schools, hospitals & housing739 m
690722Medium-absorption floor and wall tilesPublic works & housing fit-outs354 m
690723High-absorption floor and wall tilesPublic works & housing fit-outs348 m
Leading importing states
Morocco 208 mLibya 175 mCôte d'Ivoire 120 mSouth Africa 81 m
Plant two · ambition

Iron + anthracite in one niche smelter

Fabricated steel, pipe & a capped rebar wedge

9 HS6 lines · endowment-combination · infrastructure and utility tenders
USD 5.73 bngross · ≈ USD 1.44 bn addressable

Scale-matched, not inflated

Eswatini's core claim is fabricated structures, bridges, tubes, line pipe and gas containers. Larger iron-and-coal holders keep the continental bulk anchor. Rebar remains only as a capped regional wedge; its screened demand brings the total basket close to Eswatini's own imports without pretending the kingdom leads the continent.

Fabricated structures

USD 2.42 bngross demand
AspirationalGov highSubst medium
Screened
≈787 m
730890Structures and parts of iron or steel, n.e.c.Public works & transport structures2.33 bn
730810Bridges and bridge sectionsRoads and bridges authorities88 m
Leading importing states
DR Congo 306 mAlgeria 209 mEgypt 190 mCôte d'Ivoire 141 m

Capped regional rebar wedge

USD 2.31 bngross pool
CappedGov highSubst low
Maximum screened wedge
≤450 m
721391Hot-rolled bar and rod in coilRegional public works & roads861 m
721420Deformed reinforcing bar and rodRoads, schools, clinics & housing1.16 bn
721499Other hot-rolled bar and rodRegional public works projects282 m
Leading importing states
Senegal 208 mEthiopia 166 mDjibouti 145 mCôte d'Ivoire 133 m

Tubes, line pipe & gas containers

USD 999 mgross demand
AspirationalGov mediumSubst medium
Screened
≈200 m
730630Welded tubes and pipes, generalWater utilities & municipal works269 m
730690Other tubes, pipes and hollow profilesWater utilities & public works185 m
730511Large-diameter line pipeWater and energy utilities161 m
731100Containers for compressed or liquefied gasEnergy utilities & public gas systems385 m
Leading importing states
Angola 118 mSouth Africa 91 mNigeria 89 mEgypt 78 mTanzania 73 m
10 · Fairness test
Balance by construction

Eswatini supplies roughly what it buys

The re-screened basket lands at USD 1.92 billion of estimated addressable demand against Eswatini's USD 2.04 billion 2023 import bill. It is close because the allocation was designed to balance—not because the product list was inflated.

Gross
USD 7.57 bn

Continental 2023 import demand across 14 retained HS6 lines.

× government intensity
USD 4.31 bn

Product-specific factors: high .65 · medium .40.

× off-continent / substitution
USD 1.92 bn

High .75 · medium .50 · low .30. Estimated pending bilateral re-pull.

≈ balance
USD 2.04 bn

Eswatini's own 2023 imports from the same trade database.

The screened basket reaches 94.2% of Eswatini's import bill

USD 119 m below balance

Why the rebar wedge remains

Removing all bulk rebar drops the basket to about USD 1.47 bn. Retaining no more than USD 450 m of screened regional demand restores balance while leaving the continental anchor to larger producers.

A continental rule, not a prison

Balance is the adoption discipline. It need not freeze each country exactly at its current imports, but it prevents a small member from claiming a wildly disproportionate share of the fixed continental pie.

11 · Where GDP comes from
Small, real, buildable

A two-plant sequence—not a nine-times-GDP headline

The claim is sized to the customer book and to the kingdom's own economic weight. It is large enough to finance industrial capability and restrained enough to remain credible before 54 ministers.

USD 1.92 bn

Estimated addressable demand. About two-fifths of Eswatini's ≈USD 4.7 bn GDP.

≈ imports

94.2% of the kingdom's USD 2.04 bn import bill. The fairness test is met.

25 years

The allocation horizon that can make a plant financeable—subject to the Cell Phone Test on every order.

Sequence one · lower execution risk

Ceramics first

Kaolin is identified and a sanitaryware or tile line is the more modest industrial build.

Sequence two · higher capability lift

Niche steel second

Anthracite operates, but ore availability, smelting, capital and environmental performance require deeper proof.

12 · Customer book
Who buys the basket

The 2023 markets behind the two plants—ranked

Gross imports across the 14 retained lines, before the government and off-continent screens. North Africa, West African construction markets and major infrastructure economies lead; regional routes remain the logical first commercial wedge.

Leading importing states

Gross USD m · 2023
01Morocco482 m
02Côte d'Ivoire431 m
03Egypt430 m
04DR Congo415 m
05Senegal372 m
06Libya344 m
07Algeria338 m
08Ethiopia302 m
09South Africa279 m
10Tanzania269 m
11Nigeria267 m
12Ghana254 m
13Angola251 m
14Guinea184 m
15Mozambique167 m

Gross of intra-African trade. The off-continent share is applied at basket level and remains indicative pending the bilateral Comtrade re-pull.

13 · Devil's advocate
Surfaced, not buried

Where this could still be wrong

01

The two government bands are different kinds of evidence. USD 136.75 bn is measured. About USD 207 bn is estimated from the Minister's one-third ratio. The ratio is guidance, not a customs observation.

02

Parastatals are in the measured floor; contractor imports are not. The tender bridge is defensible only where government specifies the finished material and can enforce the match.

03

Tableware was removed on tender-materiality. Institutional catering exists, but it is not strong enough to carry a sovereign allocation.

04

Scale-matching deliberately limits Eswatini. Bulk sections are out, and larger iron-and-coal holders should lead continental rebar. Eswatini retains only a capped regional wedge.

05

The off-continent share remains estimated. Substitutability factors are reasoned, not yet isolated line-by-line. The bilateral trade re-pull is still required.

06

The niche smelter is the heavy lift. Anthracite operates, but Ngwenya ore is partly depleted. Feedstock, capital, energy, emissions and logistics must be proven before steel moves beyond aspiration.

07

Raw and incumbent sectors remain excluded. Sugar, timber, pulp, beef, textiles and beverages fail one or more of the government, off-continent or industrial screens.

14 · Synthesis
The honest read

One member worked in full—and a method for the other 53

The prize is USD 620 billion, but the instrument begins where government can act. That beachhead now has two explicit bands: USD 136.75 billion measured, including parastatals, and about USD 207 billion government-influenced when tender-driven contractor imports are estimated using the Minister's ratio.

Eswatini's result is intentionally modest. Four screens and three allocation principles reduce the claim to two plants, 14 HS6 lines and about USD 1.92 billion of addressable demand. Ceramics from kaolin is the near-term build. Niche steel combining iron and anthracite is the ambition. Tableware and bulk sections are out; the continental rebar anchor is ceded, with only a capped regional wedge retained.

The strength is the restraint. Each line terminates in a named tender type. Each value is labelled gross, measured or estimated. The basket lands close to Eswatini's own imports, so it can be offered to 53 other ministers as a fair method rather than a market grab.

The adoption standard

A disciplined claim, sized to balance—the version that survives a room of 54 finance ministers.

Your response — Minister, this exemplar is still Draft 1

Eswatini is worked in full to show the method, not to fix the answer. Mark where you stand and what you would change; the same seven questions sit on all 54 documents.

1

Where do you stand on this direction for Eswatini?

2

How strongly do you back the Right of Supply framework?

7/ 10
Not at allCompletely
3

How well does this draft bundle reflect Eswatini’s real strengths?

Tap to rate
4

In one word — are you happy with what you see?

5

Which single line matters most to your government?

6

What are we missing for Eswatini?

7

Your one instruction for Draft 2.

Collated live by the Office of the Chair of the AU STC-FMAEPI. Draft 2 will show, for each change, which member state asked for it.