South Africa No Longer Setting the Development and Investment Pace in Africa

For three decades after 1994, South Africa was the continent’s default address for institutional capital. It had the deepest bond market, the most sophisticated banks, the largest industrial base and the presumption that reform, however messy, would eventually deliver. That presumption is now colliding with the data. In mid-September 2026 JPMorgan finalised the constituents of its new Government Bond Index–Emerging Markets Edge, a local-currency frontier benchmark tracking roughly $330 billion of debt across 26 countries. Africa will take almost 45 percent of the weight. Egypt and Morocco sit at the 8 percent country cap. Nigeria takes 7.4 percent. Kenya, Uganda, Zambia, Côte d’Ivoire, Angola, Senegal and Namibia are in. South Africa is not.

The omission is, in a narrow technical sense, unsurprising. GBI-EM Edge was built for markets that are not already in JPMorgan’s mainstream GBI-EM Global Diversified index, and Pretoria has long sat in that senior club. The political and economic signal is harder to dismiss. The new benchmark is a map of where global fixed-income money now expects African growth, issuance and reform to come from. It is a map that no longer has Johannesburg at the centre.

A Quarterly Contraction against a 4 Percent Continental Growth

Statistics South Africa reported on 8 September that real GDP contracted 0.2 percent in the second quarter of 2026, ending six consecutive quarters of expansion. Mining fell 3.0 percent, manufacturing 1.8 percent and trade 1.9 percent. Fixed-capital formation declined for a second straight quarter. On the labour market the official unemployment rate rose to 33.6 percent, with 8.5 million people actively looking for work.

Those are not the numbers of a regional locomotive. The African Development Bank estimated continental real GDP growth at 4.2 percent in 2025 and projects 4.3 percent in 2026. Afreximbank put 2025 growth at 4.5 percent. The World Bank has Sub-Saharan Africa at 4.1 percent in 2025 and about 4.0 percent in 2026. South Africa’s own annual performance remains stuck near 1 percent: 1.1 percent in 2025 on most official tallies, with 2026 forecasts clustered between 1.0 and 1.3 percent. The 0.2 percent figure that now defines the conversation is a quarterly print, not an annual collapse. The gap with the rest of the continent is still structural, not cyclical.

Ivory Coast, Uganda, Ethiopia, Rwanda, Benin and several smaller economies are growing at 6 percent or more. Zambia’s mining rebound and Ghana’s post-restructuring recovery are visible in bond markets. Nigeria’s return to a JPMorgan local-currency benchmark after eleven years is being sold in Abuja as a verdict on reform. South Africa, still Africa’s most industrialised economy, is growing at a third of the continental average and shrinking in the sectors that used to define its comparative advantage.

Where the Capital is Actually Landing

UNCTAD’s World Investment Report 2026 recorded about $70 billion of FDI into Africa in 2025, the third-highest level since 1990 even after a retreat from the exceptional $94 billion of 2024. The distribution is the point. Egypt absorbed $15.5 billion. Mozambique took roughly $5.7–6 billion on the back of LNG. Nigeria recorded about $4 billion, Uganda $3.4 billion, Morocco $3.3 billion, Namibia $1.4 billion and Zambia $1.3 billion. South Africa posted negative inflows of about $2.3 billion, driven by intracompany financing, profit repatriation and merger-and-acquisition accounting rather than a sudden investor boycott. The country remains a destination for announced projects. It is no longer a destination for net new equity on the scale its size would imply.

The Contrast with Neighbours is Concrete, not Rhetorical

In Angola, the National Oil, Gas & Biofuels Agency signed eleven upstream agreements at the Angola Oil & Gas conference in September 2026, covering new deepwater acreage and incremental investment in existing blocks. TotalEnergies and partners have flagged $10 billion of spending over five years. Chevron’s Block 0 concession has been extended to 2050. Shell signed three agreements. New names, including Pertamina, are evaluating entry.

In Botswana, MMG broke ground in February 2026 on the next phase of Khoemacau, three new copper mines designed to stretch the operation beyond twenty years and position the country as a meaningful copper exporter. Debswana’s mining licences have been extended to 2054. The government has signed energy partnerships that it says could mobilise P40 billion and, on an expanded case, more than P74 billion. Oman has committed to renewable projects of up to 3,000 MW and mineral exploration across a large share of unexplored territory.

Namibia is in a different league of optionality. TotalEnergies has been targeting a final investment decision on the Venus discovery, a development scoped at roughly 150,000 barrels a day and about 750 million barrels of oil equivalent. Shell continues to drill in the Orange Basin after mixed results. The EU partnership on green hydrogen and critical minerals has been rolled forward to 2030, with Hyphen, HyIron and related projects in the Team Europe pipeline. Andrada Mining secured competition-authority approval for a $51 million earn-in at Brandberg West. First oil is still a 2029–30 story. The capital and the geology are already being priced.

South Africa’s Department of Trade, Industry and Competition can point to R647 billion of “attracted” investment in 2025/26 and R415 billion of pledges at the latest South Africa Investment Conference. Pledges are not the same as cash that has crossed the border and stayed. Mining output is falling even as prices for several of the country’s commodities have been supportive. As one industry executive put it in September, the country is benefiting from prices, not volumes.

Licences at Geological Speed

The most measurable bottleneck sits in the Department of Mineral and Petroleum Resources. The electronic mining cadastre, promised for years as the fix for opacity and delay, was still only live in the Western Cape by mid-2026. National rollout has been pushed to 31 March 2027. The director-general has said the department receives about 2,800 applications a year and processes about 2,500, which means the backlog is a rolling stock, not a one-off pile. Between February 2025 and January 2026 the department granted 358 prospecting rights and 32 mining rights — a figure ministers present as evidence of confidence. It is also evidence of how thin the conversion from application to mine remains.

Courts are now doing the department’s job. In May 2026 the Pretoria High Court ordered the minister to consent, within days, to Afrimat’s transfer of a purchased mining right after a nine-month administrative silence that threatened the company’s allocation on Transnet’s iron-ore export corridor. That is not a development strategy. It is a confession that section 11 of the Mineral and Petroleum Resources Development Act has become a lottery.

Manufacturing tells a similar story. Factory output contracted for a third consecutive quarter in Q2 2026. Seven of ten manufacturing divisions were down. Logistics reform at Transnet has begun to show operational improvement, and Eskom’s load-shedding collapsed from 329 days two years earlier to a handful of days. Those are real gains. They have not yet produced a capex cycle. Gross fixed-capital formation is still going backwards.

The Regulatory Knot

South Africa’s problem is not a shortage of rules. It is a surplus of overlapping, slowly administered and frequently revised ones. The Mineral Resources Development Bill was still being “refined” in mid-2026 after years of consultation. Mining Charter obligations, ownership targets and the treatment of beneficiation remain moving targets. Expropriation legislation and the political temperature around property rights add a risk premium that no slide deck from an investment conference can cancel. Labour-market rigidity sits on top of a 33.6 percent unemployment rate that should, in any normal political economy, have produced a bias toward job-creating investment.

None of this is unique in Africa. Angola, Namibia and Botswana also impose local-content rules, state participation and environmental conditions. The difference is speed and predictability. An investor who can see a licence decision, a grid connection and an export path inside a defined window will accept a high royalty. An investor who cannot will go to the Kalahari Copper Belt, the Orange Basin or Block 32.

The GNU has improved the tone of economic policy. It has not yet rewritten the operating system. Cadastre delays, section 11 transfers, municipal dysfunction and the residual unreliability of rail and ports still tax every project that needs to move bulk commodities. That tax is now visible in the growth numbers.

What a Cleaner Framework Could still Unlock

South Africa does not need to become a frontier market to grow like one. It already has deep capital markets, a convertible currency, a large installed industrial base, world-class mining engineering and a legal system that, when it functions, still protects contracts. The prize from regulatory repair is therefore larger than for most of the countries now filling GBI-EM Edge.

A functioning national cadastre mining licence system, statutory time limits on licence and transfer decisions, and a mining code that is frozen for a decade rather than reopened every political cycle would not, by themselves, deliver a 4 percent growth. Combined with competitive electricity pricing, a Transnet that can move ore on contract, and a labour regime that prices young workers into jobs rather than out of them, they would change the investment calculus in the one African economy that still has scale.

The arithmetic is straightforward. Mining and manufacturing together are a modest share of GDP but a large share of exports, skilled employment and downstream industry. A return to even the volume growth of the mid-2000s, on today’s commodity slate of platinum-group metals, manganese, chrome, iron ore and coal plus a serious critical-minerals push, would lift the national growth rate by a multiple of the current 1 percent trend. Exploration spend, which has leaked to Western Australia, Canada and now Namibia and Botswana, would follow the licences. So would the processing plants that governments keep announcing and rarely commission.

The same logic applies outside the pit. Special economic zones only work if the rules inside them are actually different and actually enforced. The Richards Bay titanium-dioxide project and a handful of energy-storage and renewable bids show that capital will still write cheques when the grid, the port and the permit line up. The failure is that too few projects clear that filter.

A credible reform package would also change South Africa’s place in the bond universe. The country does not belong in a frontier index. It does belong in a conversation about whether its local-currency market is a growth market or a carry-trade residual. Higher trend growth, a declining debt-to-GDP ratio and a licensing system that produces mines rather than court orders would tighten spreads more effectively than another investment conference.

The GBI-EM Edge announcement is not an insult. It is a ranking. African states that have made issuance larger, more regular and more accessible are being paid for it in index weight and in the prospect of dedicated inflows. African states that are licensing new copper, oil and hydrogen projects are being paid in FDI. South Africa still has the institutions to recapture both. It no longer has the luxury of assuming that size and history will do the work that licences, rails and predictable rules now do everywhere else on the continent.