South Africa · ZAF · snapshot 2026-08-16
The rand went the other way
Reviewed
Currency movement
-9.4%
The local currency bought more euros at the end of this window than at the start — a more expensive entry for a euro-denominated buyer, and a better exit for anyone already holding the asset.
What it does not show. It is not a forecast and not a property return. A currency move and a price move are different facts and routinely point opposite ways; nothing here says which way either is going next.
Foreign-buyer rules
1 of 1 published rules apply
These are the restrictions a foreign buyer meets before price is even discussed — approval regimes, surcharges, ownership limits.
What it does not show. It is not a completeness claim. A rule we have not read is not a rule that does not exist, and this row counts what is cited, not what is in force.
Cost of money
7%
The price of money to banks in this jurisdiction, which sets the floor everything else is priced from.
What it does not show. It is emphatically not a mortgage quote. What a non-resident is offered is a different number, usually much higher, and the gap between them is where a market's financing story actually lives.
Supply against absorption
Licence unconfirmed
Only one national statistics agency has been read and cleared for publication so far. For every other market the data exists and the permission to republish it does not.
Residential price cycle
90.9 against a 2015 base of 100
House prices have risen more slowly than consumer prices since 2015, or fallen — the market is below its own inflation-adjusted level of that year.
What it does not show. It is a position, not a direction, and it is not a valuation. Deflating by consumer prices says nothing about what the asset yields or what credit costs, and a market can sit above its 2015 level for years without that meaning anything is due.
Valuation against its own history
94.0 to income
Housing costs less relative to incomes than it did in 2015, on this country's own index. Affordability has improved against the local salary, which can mean prices fell or that incomes rose faster — the ratio does not say which.
What it does not show. It is not comparable to another country's figure. Each index is 100 in its own 2015, so two of these side by side are two different base years wearing the same label — a market at 120 is not dearer than one at 90, it has merely moved further from its own past. It is also not a forecast, and a ratio can stay stretched for a decade.
On the market now
Not measured
No listings platform has been read for this market. The one platform this site can read covers Russian cities only, and inventing a count for anywhere else would be worse than an empty row.
Thesis
South Africa is on this list because it is the market these conversations started from, and because the honest thing to do with a market you discussed months ago is to check what actually happened. Over the twelve months to July 2026 the rand did not weaken. It strengthened 9.4% against the euro, moving from 20.75 to 18.80 per euro on ECB monthly reference averages.
Counter-thesis
This page makes no claim that the screen would have predicted that, and it must not be read as one. The currency series used here is not revised, which is the only reason a retrospective reading of it is legitimate at all; every other layer on this site comes from series that are revised, and a backtest across those would be reading today's corrected numbers into a past decision. What this page demonstrates is narrower and more useful: a currency assumption is checkable, and this one would not have survived the check.
The series behind those figures, and the workings
Currency
-9.4%
Twelve-month move in ZAR per euro, 20.7477 in 2025-07 to 18.8035 in 2026-07. Positive means the local currency weakened. Source: European Central Bank monthly reference averages.
18.6978
Latest daily fixing, ZAR per euro on 2026-08-14. Shown because it is what a reader would see if they went and looked; the window above uses monthly averages, which one volatile day cannot move.
Against other base currencies
A market is cheap or dear relative to the capital entering it, so the same twelve months read differently depending on what the buyer holds. Both legs are quoted per euro, so the euro cancels and no third source is involved.
| Buyer holds | ZAR per unit, 2025-07 | ZAR per unit, 2026-07 | Twelve-month move |
|---|---|---|---|
| US dollar (USD) | 17.7682 | 16.4691 | -7.3% |
| Pound sterling (GBP) | 23.9945 | 22.0204 | -8.2% |
| Swiss franc (CHF) | 22.2503 | 20.3156 | -8.7% |
| Singapore dollar (SGD) | 13.8687 | 12.7514 | -8.1% |
| Australian dollar (AUD) | 11.6154 | 11.4686 | -1.3% |
Cost of money
7%
Central bank policy rate, 2026-07. Source: Bank for International Settlements.
This is the price of money to banks, not the price of money to a non-resident buying a building. The gap between those two is where a market’s financing story actually lives, and it is not in this number. A mortgage quote for a foreign buyer depends on the lender, the asset and the borrower, and no public dataset carries it.
BIS terms permit this figure on a free page and state that nothing in their statistics constitutes investment advice. They also allow inclusion in a commercial product only where it adds no charge, so this series is published here and is not priced into any paid work.
Background
Annual national statistics from the World Bank, under a licence that permits reuse with attribution. They describe the country; they do not time an entry, and they are not counted as measurements of this market — otherwise every territory on earth would claim to be measured on the strength of one dataset.
| Indicator | Value | Year |
|---|---|---|
| GDP per capita | 6,598 US$ | 2025 |
| Consumer price inflation | 3.2 % | 2025 |
| Population | 64,747,319 | 2025 |
| Urban population | 63.8 % of total | 2025 |
| Official exchange rate | 17.8887 per US$ | 2025 |
What the rules do to a foreign buyer
Each entry cites the instrument and the authority page it was read from, with the date it was retrieved. Facts of law change; the retrieval date is how you know how old this reading is.
| Topic | In force | When it bites | Applies to | What applies | Instrument | Read on |
|---|---|---|---|---|---|---|
| Withholding on a sale by a non-resident | In forceNot stated by the source | On the way out | All property | A purchaser paying a non-resident for immovable property in South Africa must withhold part of the price and pay it to SARS: 7.5% where the seller is a natural person, 10% a company, 15% a trust. It does not apply where the amounts payable total R2 million or less — but above that line the percentage is charged on the whole price, not on the excess, so a R2.1 million sale is withheld in full. The money is an advance against the seller's own income tax for that year, not an extra tax, and the seller may apply to SARS for a directive reducing or removing it. For a foreign buyer this reads twice: it is an obligation you carry when you buy from a non-resident, and it is what happens to your own proceeds when you come to sell. | Income Tax Act No. 58 of 1962 (South Africa), section 35A. Retrieved from the South African Revenue Service's own external guide IT-PP-02-G01, "Amounts to be Withheld When a Non-Resident Sells Immovable Property in South Africa", effective 3 September 2025, and from the SARS page for non-resident sellers. | 2026-08-11 |
What supports this, what would make it wrong, and how it could go
What supports this
- Rand per euro moved 20.75 → 18.80 between the July 2025 and July 2026 monthly averages (ECB reference rates), a 9.4% strengthening of the rand.
- The latest daily fixing on 7 August 2026 was 18.71 per euro, consistent with the monthly series.
- ECB reference rates are fixings and are not revised, so the twelve-month reading is the same figure a reader would have computed at the time.
What would make this wrong
- The rand resumes weakening, in which case the twelve-month window closes on a different answer and this page changes with it.
- A reader treats this as evidence of forecasting skill; it is evidence of an assumption being checkable, which is a different and smaller claim.
Outcome branches
- A buyer entered on a weak-rand thesis a year agoThe currency worked against them on entry cost relative to waiting, before any consideration of the local asset.
- The screen is applied before the next such conversationThe currency assumption is tested against a non-revised series in minutes rather than assumed, which is the whole operational claim being made here.
Record
| Why this market is on the list | On watch. Discussed with the same relationships two months before the Japan and Singapore enquiries, which makes it the natural test of whether this screen would have said anything useful at the time. Recorded in ADR-0010 §9. |
|---|---|
| Dossier | Approved by founder on 2026-08-10 |
| Next review | 2026-11-10 |
| Evidence snapshot | 2026-08-16 |