Visual has been suspended, cannot afford the audit of its February 2026 accounts, and has been shopping for money since at least September. It has now signed a non-binding term sheet with Tongshang Feizhou (Hainan) International Trade for R55 million: R18 million for 450 million new shares at 4 cents, and a R37 million secured facility for Phase 1 of the Stellendale retirement development. Three million rand of the equity goes straight at the audit and the historic creditors, which is the difference between resuming trade and quietly disappearing.
Read the small print before you get excited. It is non-binding. It needs definitive agreements, board approvals, regulatory approvals and a JSE ruling because the shares are suspended. And 450 million new shares against roughly 1.2 billion already in issue means existing holders are handing over a bit more than a quarter of the company at 4 cents a share to a counterparty with no track record in South Africa. You are not being rescued, you are being recapitalised — and the price of survival is the equity. This is worth watching, not buying.
Sanlam already owns 61.9% of Santam, which means it already sets the dividend, the capital and the strategy. What it does not own is the share of the earnings that leaks out to minorities every year. Buying that back for R21.2 billion in cash hands the group the whole cheque — and it is paying R505 a share for a stock that closed at R485, so the control premium is roughly what you would spend on a decent lunch. Irrevocables and indications of support already cover 48.3% of the scheme shares, and Sanlam's own stake does the rest of the arithmetic. This is not a negotiation. It is a formality with a date on it.
The interesting part is what it says about where Sanlam's capital is going: out of spread businesses and into the short-term insurer trading on 12.7 times earnings and 3.3 times book. Santam shareholders are being offered 4% to hand over a business that earns a premium multiple, and the honest answer is that they are getting a thin deal. Which is precisely why it is a good one for Sanlam. Buying in something you already control, at a price barely above the market, is the cheapest consolidation there is — and the only people who should feel short-changed are the ones on the other side of the trade.
Off a base of 326 cents, Omnia is guiding to 407-440 cents of earnings per share for the half year — a 25% to 35% lift — with headline earnings per share up a more modest 18% to 28% to 378-410 cents. Now look at the gap between those two numbers. Roughly seven percentage points of the growth is coming from below the headline line, which is accountant for 'something got sold, revalued or paid out that will not repeat'. The operations did the heavy lifting, but not all of it.
At R122.80 you are paying 14.3 times trailing and 12 times forward for a business that sells explosives to mines and fertiliser to farms — volumes that follow the commodity cycle and pricing that follows the rand. The direction of travel is right and the auditors have signed off on the range, so this is not a hopeful trading statement. But the real test is 9 November: if Agriculture is carrying the growth, this re-rates from 12 times forward. If it is mining volumes and a soft rand doing the work, the multiple is already fair and you are buying the cycle, not the company.
Ferrochrome production collapsed 46% to 61kt for the nine months to September, because Boshoek went off in May 2025 and Wonderkop followed a month later when Eskom's electricity price made smelting a donation. Two furnaces, one utility bill, a year of lost output. Chrome ore only fell 7% and PGMs 8%, which tells you the mine kept working and it was the smelting side that got taken out back.
Then in June the new electricity agreements with Eskom were concluded, Boshoek restarted in September, Wonderkop in October, and the ramp-up runs through the fourth quarter. So the nine-month number you just read is a photograph of the worst of it. At R1.36 the stock is on 8 times trailing but under 4 times forward earnings and two-thirds of book — the market has these assets priced as if the furnaces never come back, and they are back. The risk is not the restart, it is the chrome price and whether that Eskom tariff survives the next round of increases. The trough is in.
The court found the Minister's advisory committee was never provided for in the Request for Proposal and set the award aside on that procedural point alone. It accepted the Minister acted in good faith, made no adverse findings against Sizekhaya, its shareholders, management or the quality of its bid, and dismissed the EFF's separate application outright. Sizekhaya keeps running the National Lottery, the draws keep happening, the tickets keep selling. On the face of it, barely a scratch.
Here is the problem. The asset you own — 59.4% of Goldrush Group, which holds 50% of Sizekhaya pending the government's shares — has just been put back in a politician's in-tray with a twelve-month clock on it. 'Eligible to be awarded' is not the same sentence as 'awarded', and you cannot build a valuation on a maybe. Ithuba and Ringeta get another bite at a process they have already spent two years litigating, and the Minister gets to reconsider with the committee's report in the bin. At 79 cents the market marked it down 1.25% — no panic, just a quiet repricing of a lottery ticket as a lottery ticket. That is the right call.
Read the reasons in the order the company wrote them. Distributable income is up because net finance costs fell and the expected credit loss charge came down. It went up despite lower net operating income and higher administrative expenses. That is not a landlord getting better at landlording. That is a balance sheet that stopped bleeding and a debtor book that stopped getting worse. The buildings are still collecting less rent than they were a year ago.
At 36 cents, with 11 cents of distribution in a half, the market is offering you a yield north of 60% and telling you exactly what it thinks of it. A REIT whose net operating income is shrinking is a melting ice cube — every distribution is a slice off the capital, not a return on it, and 0.1 times book says the market does not believe the asset valuations either. It jumped 9% on the statement, which tells you how little was expected. Cheap is not the same as good, and nothing here says the office portfolio has found a floor.
Jebb McIntosh died on Monday. He had been on the CMH board since 1976 — half a century in a business where the average executive lasts five years and the average car dealership changes hands every decade. For the motor trade he and CMH were the same thing, and the board's announcement reads like people who have lost a colleague rather than a line in a governance file. That counts for something on the JSE, where founder-led businesses are rarer than they should be.
The uncomfortable part, the part no SENS announcement will say: a man who ran the business for fifty years is the key-man risk nobody was ever charged for. At R37.50 and 6.6 times earnings, CMH is priced like a company that will quietly keep doing what it always did — sell cars, finance them, take the margin, pay the dividend. The market now has to work out whether that was the man or the machine, and the board has not yet told anyone who is holding the wheel. NEUTRAL, and not because the numbers are bad. They are cheap. Because the next two quarters are a succession test and the answer is not in the announcement.