Metro elections, 4 November 2026
Eight metros, eight questions, one very uneven map
In one South African city the next mayor is close to decided. In another, two candidates are two points apart. The gap between them says more about how this country is governed than any manifesto will.
South Africans vote for metro councils on 4 November. Eight metros, eight mayors, and eight completely different levels of certainty about who ends up with the job.
These are live prices, not opinions. Each number is what people are currently willing to pay for a contract that pays out if that person becomes mayor. When money moves, the number moves.
Next mayor — live odds
25 SepCape Town
Geordin Hill-Lewis
Buffalo City
Princess Faku
Johannesburg
Helen Zille
Tshwane
Cilliers Brink
Mangaung
Werner Pretorius
Nelson Mandela Bay
Retief Odendaal
Ekurhuleni
Jongizizwe Dlabathi
eThekwini
Haniff Hoosen
You don't vote for a mayor
This trips up almost everyone, and nothing on this page makes sense without it.
South Africans do not elect mayors directly. You vote for a council. The council then elects the mayor. So the winner isn't whoever the public likes most, it's whoever can assemble half a chamber after the votes are counted.
Now look at the board again. The cities at the top aren't the ones with the best mayors. They're the ones where a single party can govern alone.
Certainty tracks coalition arithmetic, not popularity. Every city where one party can win outright sits near the top. Every city that needs a coalition sits near the bottom.
Tshwane — the biggest move of the month
On 30 August this page had Cilliers Brink at 28%. He is now the clearest favourite outside the Cape. Nothing about Tshwane's politics became simpler in that month; what changed is that the market stopped treating the race as a free-for-all. Nasiphi Moya, who briefly looked like the story in early September, has settled back into second.
eThekwini — two points apart
Durban is now the closest race in the country. Two points separate the top two, and both of them sit in the thirties, which is another way of saying that two-thirds of the probability is spread across everyone else. In the country's third-largest metro, the honest answer to "who will run this city" is still that nobody knows.
Cape Town — the control group
Geordin Hill-Lewis remains the closest thing to a settled outcome in the country, though the market has eased him back from the mid-nineties over the past month. Cape Town is the only metro where one party holds an outright majority. No coalition, no arithmetic, no kingmakers.
Johannesburg — a favourite with a deep challenger
Helen Zille leads comfortably. What's interesting is underneath: ANC veteran Frank Chikane sits well behind her on price, but carries by far the deepest pool of money in the race. More conviction is parked on the challenger than on the favourite.
Mangaung — a different name at the top
A month ago the leader here was Gregory Nthatisi. Today it's Werner Pretorius, and Nthatisi has fallen into single digits. It is the only metro on the board where the favourite has changed identity rather than just price, and it happened on a market thin enough that it deserves the caveat below.
These markets are not equally reliable and we'd rather say so than let the percentages look interchangeable. The figure under each row is the money standing behind that particular price.
Where it reads very thin, a single participant can move the number by tens of points, or withdraw and move it back. Cape Town's price is backed by around R170,000. Ekurhuleni's is backed by about R8,000. Those two numbers should not be read with the same confidence, even though they're printed in the same font.
One more thing worth knowing: some candidates sitting at 1–3% show very large liquidity figures. That's automated market-making, not conviction. Genuine interest in this country's mayoral races is measured in tens of thousands of rands, not millions.
Treat every number here as a starting point for your own thinking, not a verdict.
How to read a percentage
A price of 35% doesn't mean "no". It means that if you could run this election a hundred times, that outcome wins about thirty-five of them. Long shots land. Favourites lose. It's a measure of confidence, not a forecast of fact.
It also means something practical: at 35%, you're paying 35c for a contract that pays R1 if you're right. If you believe the true chance is 50%, that price is wrong in your favour, and backing that belief is how the price moves toward what you think.
New to this? Read the 3-minute explainerWhat moves these numbers before 4 November
Candidate confirmations. Several parties still haven't finalised metro candidates. Every confirmation resolves uncertainty and reprices a race.
Coalition signals. Small parties declaring who they'll work with matters more than anything a mayoral candidate says. In a hung council, the kingmaker holds the last few seats.
Turnout. Metro outcomes turn on who shows up, not who is persuaded. It is the hardest thing to poll and the most important thing to know.
Anything that lands on a record. An audit outcome, a parliamentary finding, a court date. These cities are governed under scrutiny, and each new fact reprices the people responsible for it.
Taking a position
If you think one of these prices is wrong, you can say so with money instead of a comment thread.
- Sign up with your email. No wallet, no seed phrase, no crypto experience needed.
- Add money in rands. Normal EFT from your bank with a short reference code. From R50.
- Pick a side on any market at the price showing. Trades start at R20.
- Cash out to your bank account whenever you want, or hold until the market resolves.
You're trading against other people's opinions, not against a house. The odds you see are the odds you get.
40 days. Eight cities. Eight open questions.
Predika is built in South Africa, priced in rands, on the outcomes that decide how this country actually runs.
Trade the metro markets Add money in rands