South African Stocks (2026): Gold, Cheap Banks and the JSE Discount

South Africa is the market that breaks the developed-versus-emerging framing. The Johannesburg Stock Exchange has been running since 1887, uses IFRS accounting, has deep institutional ownership and genuine corporate governance - developed-market plumbing by any standard. Its companies trade at emerging-market prices. This guide explains why, what you are actually buying, and where the discount is earned rather than free.

Developed plumbing, emerging price

Across the South African companies we track, the median trades near 12 times earnings and 5.6 times EV/EBITDA, on a median return on equity of about 17%. For comparison, the median United States stock in our data sits near 32 times earnings. Roughly a third of the multiple, at comparable profitability, on an exchange with none of the reporting opacity you might expect. The full, live table:

Three different things you are actually buying

"South African equities" is not one asset. It is three, and they behave nothing alike.

Gold and bulk miners. This is the market's global claim to fame. Gold Fields, AngloGold Ashanti, Harmony and Sibanye give you some of the world's largest gold production, mostly in the single-digit-to-low-teens P/E range. In bulk commodities, Kumba Iron Ore trades near 6 times earnings and under 3 times EV/EBITDA. These are leveraged bets on commodity prices first and South Africa second.

Cheap financials. Standard Bank, Absa, FirstRand, Nedbank, Old Mutual, Sanlam and Momentum span roughly six to fourteen times earnings. These are large, established banks and insurers serving a whole continent's worth of operations - priced as though the country's problems are permanent.

The holding-company discount. Naspers trades near 10 times earnings largely because it is a wrapper around a stake in Tencent, and the market persistently values the wrapper below the contents. That is a well-known structural discount, not a mispricing waiting to be noticed.

Why it is cheap: the rand, the grid, and politics

The discount is real compensation for real problems - name them honestly.

The currency. The rand has depreciated against the dollar over most long horizons. You can be right about the company and still lose in dollars. This is the single biggest drag on international returns from South Africa.

Electricity and infrastructure. Years of load-shedding from the state utility have directly suppressed output for miners, retailers and manufacturers. Rail and port constraints have at times stranded commodity exports that were otherwise sold. These are earnings problems, not sentiment.

Politics and policy. Policy uncertainty and governance concerns at state entities keep a permanent risk premium on the market.

None of this makes the market uninvestable - it makes the discount a price for a specific risk. The question is never "is it cheap", it is "am I being paid enough for the rand and the grid".

What the value evidence says - and its limits

Buying cheap has strong long-run support: Fama and French documented a value premium across decades and thousands of companies, and Templeton profited by buying markets at their point of maximum pessimism - a description that fits South Africa for much of the past decade. Our own data offers a narrower test: sorting the companies we track into thirds by valuation, the cheapest third beat the priciest on a median basis over the following year, but the average reversed it and controlling for earnings growth narrows the edge. One year, in a commodity-friendly period - suggestive, not proof.

We tested this on our own data. We took 190 issuers (commodity-heavy — that is where we have full price and filing history), ranked them by EV/EBITDA on 2025-09-05, and measured the next 12 months' return (to 2026-09-04). The cheapest third returned a median 46.0% versus 28.5% for the priciest third (median EV/EBITDA 5.3x vs 22.7x).

Honest caveats: this is one unusual year (a strong commodity and value rally), the sample skews to commodities, and on the mean (not median) the priciest third actually won, on a few tail winners. Once you control for EBITDA growth the cheap edge narrows (among high-growth names the cheaper half returned 31.4% vs 33.5% for the pricier half). One year is an illustration, not proof — the durable evidence is the multi-decade academic record above. This observation is recomputed daily.

What re-rating looks like: names that traded near ~2x EV/EBITDA a year ago and their subsequent return. For several, EBITDA barely grew — so the gains came from multiple re-rating, not earnings:

How to use this

The mistake is treating the three groups as one. A gold miner at 8 times earnings is a commodity-price call; a bank at 8 times is a call on South African credit and the rand; Naspers at 10 times is a call on Tencent and a discount that may never close. They need different questions.

Price is what you pay; value is what you get. The screen tells you where the price is low. The filings - and an honest view on the currency - tell you whether there is value behind it.

The full, ranked table of South African stocks by valuation - updated daily from filings - is above and on the South Africa market page, each name linking to its own card.