The Road to November

South Africa spent the first half of 2026 proving that its national institutions can be repaired, and that repairing them is not the same as delivering anything.

The evidence for the first half of that sentence is unusually strong. On 5 June Fitch raised the sovereign rating to BB, its first upgrade of South Africa in almost twenty one years, seven months after S&P did the same. In May, Moody's moved its outlook to positive, the first such revision since 2007. The economy grew 0.5 percent in the first quarter, a sixth consecutive quarterly expansion. The current account posted a surplus of R190.7 billion, 2.4 percent of GDP and the widest since 2021. On 16 May the grid completed a full year without load shedding, and by early July the run had reached 413 days.

Now the second half of it. Over the same six months the official unemployment rate rose to 32.7 percent as 345,000 people lost work. The Department of Water and Sanitation published assessments showing nearly half of the country’s wastewater systems in a critical state, and confirmed that water lost before billing is costing roughly R26 billion a year. In Johannesburg a single January failure at a bulk supply plant emptied reservoirs across the city for weeks. Several of us went those weeks without water at home.

Those two paragraphs are usually published as separate stories, in separate sections, for separate audiences. We put them side by side because they describe one mechanism. South Africa has repaired the institutions that answer to Treasury and to the Reserve Bank, and has not repaired the institutions that answer to a municipal manager. Everything a household actually consumes is delivered by the second group.

In a country producing this much daily news, that is the only job a quarterly can usefully do. The headline is almost always accurate and almost always partial. It carries the announcement and drops the mechanism, and the mechanism is where the money and the risk sit.

Rail is the clearest test of whether the gap can be closed. On 13 May the Transnet Rail Infrastructure Manager concluded access agreements with eleven private train operating companies, taking the network from one operator to twelve across five corridors and adding an initial 24 million tonnes of capacity, with 52 million possible over five years. That is genuine structural change, executed. It is also five years from the original policy decision, and the first private trains are only expected in the second half of this year.

We are not going to pretend reform is failing. It is not. What we would say is that the programme has now delivered most of what can be delivered from Pretoria, and that the remaining constraints, water, refuse, billing, roads, clinics, are held by municipalities whose finances and staffing only entered the programme at Phase II, and which are reconstituted by ballot on 4 November.

So the question we would put to any organisation reading this at the midpoint of the year is not whether South Africa is reforming, because on the record it is. The question is where your own exposure actually sits. If it sits at national level, the last twelve months have been good. If it sits at municipal level, and for most firms and almost all households it does, the last twelve months have been something else, and November decides the next five years of it.