AfriCapital Review · Africa 30 Index · Methodology v1.3

September 2026 Rebalancing

Screen results for all thirty constituents, the eligible universe across five exchanges, the amendments adopted to the selection rule, and the constituent changes that take effect on 15 October.

Summary

Of the twenty-seven constituents we can screen, seventeen pass every criterion and rank inside the buffer. Two fail the market-capitalisation floor and are removed. Three fail a liquidity or free-float test and go on the Watch List for one review period, as the methodology provides. Three BRVM constituents cannot be screened because no data source we have tested carries BRVM fundamentals.

The material finding is structural. The selection rule ranks the whole continent by free-float market capitalisation, and the South African market is deep enough that it takes 24 of the 27 screenable seats on that basis. The 50% country cap then has to compress those names into half the index and hand the other half to five non-South-African constituents — producing an index in which a single Egyptian bank is the largest position at 12.5% and Naspers, with $34 billion of free float, carries 4.6%. The methodology's stated objective of a benchmark spanning seven exchanges and ten countries cannot be delivered by its own selection rule.

Current constituents

Criteria as published: free-float market capitalisation ≥ USD 500m (C1), free-float factor ≥ 15% (C2), 90-day average daily turnover ≥ USD 500k (C3). Rank is among eligible African-domiciled securities after removing inward-listed non-African companies. Under §3, a constituent is removed if it fails C1 or ranks outside the top 38; a constituent failing only C2 or C3 is placed on the Watch List for one review period.

Trading-days (C4) and listing-age criteria were not screened: no tested source exposes a trading-day count, and every current constituent has been listed well over six months. Dangote Cement's 11% free float is a structural fact about its ownership, not a data artefact.

What the rules as written would produce

Applying §3 mechanically — add any eligible non-constituent ranked in the top 22, remove any constituent ranked outside the top 38 — yields the left-hand outcome. The right-hand outcome allocates seats by exchange first and ranks within each; it is illustrative, using 12 South African seats, 6 Nigerian, 4 Kenyan, 3 Egyptian, 3 Moroccan and the 3 BRVM constituents carried. Both apply the published free-float bands and the 15% single-stock and 50% single-country caps iteratively.

The illustrative outcome on the right was not free of problems, and those problems shaped decisions C and D below. It fills the South African seats with three more miners, which lifts the mining sector to 42% and finance to 36% before the 35% sector cap is applied; and it drops Maroc Telecom for Marsa Maroc on a free-float difference of $12 million, which is exactly the kind of churn the buffer rules exist to prevent. Those are the reasons the seat allocation and a within-country buffer both need to be decided rather than assumed.

Decisions adopted

Five amendments to Methodology v1.2, adopted at this review and applied to the constituent list below. They will be incorporated into the published methodology as v1.3.

1. Seats are allocated by exchange, then ranked within each

   Eleven seats to the Johannesburg Stock Exchange (including South African companies whose primary line trades on the NYSE), six to the Nigerian Exchange, four to the Nairobi Securities Exchange, three each to the Egyptian and Casablanca exchanges, and three to the BRVM. Within each allocation, eligible securities rank by free-float market capitalisation. The 15% single-stock, 50% single-country and 35% single-sector caps continue to govern weights unchanged.

2. Cross-listed African companies are classified by domicile of principal operations

   AngloGold Ashanti is classified as South African, notwithstanding its 2023 re-incorporation in the United Kingdom. Country of incorporation remains the primary classification factor for companies incorporated in Africa; for an African company that has re-domiciled abroad while its operations, listing history and shareholder base remain African, the country of principal operations governs. This closes the loophole under which the index's largest constituent would otherwise sit outside the South African cap.

3. The buffer applies within each exchange

   An existing constituent is retained if it ranks within its exchange's allocation plus two. A non-constituent is added only when a seat is vacant and it ranks within the allocation. At this review the buffer retains Discovery (12th in South Africa) and Maroc Telecom (4th in Morocco), the latter $12 million behind Marsa Maroc — the churn the rule exists to prevent.

4. Caps are applied iteratively in the order stock, country, sector

   As §4 already provides. At this review the sector cap binds: Finance is held at exactly 35.00%. No constituent reaches the 15% stock cap; South Africa is held at exactly 50.00%.

5. BRVM constituents are carried at their existing combined weight

   Sonatel, Ecobank Transnational and Société Générale Côte d'Ivoire cannot be screened because no data source tested exposes BRVM fundamentals. They are carried at their current combined weight of 3.47%, and the remaining 96.53% is allocated across the twenty-seven screened constituents. If a source remains unavailable at the March 2027 review, the BRVM allocation will be reconsidered rather than carried indefinitely. Watch List constituents with a free float below 15% are weighted at the 15% band floor for the review period.

Constituent changes, effective 15 October 2026

The methodology requires changes to take effect no sooner than 30 days after announcement (§8). This notice was published on 13 September; the changes take effect at the open on Thursday 15 October 2026.

Shoprite, Sasol and Vodacom leave not because they fail any criterion — all three pass — but because they rank 16th, 17th and 21st among eligible South African securities against an allocation of eleven plus a buffer of two. Jumia ranks 59th. Nigerian Breweries and BUA Cement fail the market-capitalisation floor. The three Watch List names retain their seats for one review period and are reassessed in March 2027.

Weights at rebalancing

Fixed from 15 October until the March 2027 review. Free-float market capitalisation is at the 13 September reference date; weights are after banding and caps.

Largest position: AngloGold Ashanti, 9.97%. Smallest: East African Breweries, 0.78%, on the Watch List. Six of thirty constituents are new; the three Watch List names and the three BRVM names are the ones to revisit in March.

Data and method

Fundamentals — price, total and float shares, 90-day average volume, sector — from TradingView's screener for the JSE, NGX, NSE, EGX and CSE, and for the NYSE cross-listings, on 13 September 2026; 488 companies with market capitalisation above $100 million. All five local exchanges were queried directly, so the proxy-listing issue affecting the live index calculation (Maroc Telecom via Paris, GTCO and Seplat via London) does not apply to this screen. Currency conversion at Yahoo Finance spot rates on the same date. Free-float market capitalisation is closing price × float shares; turnover is 90-day average volume × price. Inward-listed non-African companies (BHP, Richemont, British American Tobacco, Glencore, AB InBev, Prosus, South32, Mondi, NEPI Rockcastle) were excluded under §2; Anglo American plc and Investec plc were excluded as borderline and are noted for review. Cross-listed entities were counted once, on the line with greater liquidity, per the §2 cross-listing policy. Weights use the published free-float bands. The screening code and outputs are retained with the index history.