South Africa's liquidation numbers have never stopped being published, but a quiet correction earlier this year revealed they'd been substantially undercounted, and the direction they seemed to show was backwards. Stats SA's own figures now show 2,904 businesses were liquidated in 2025, not the 1,534 the agency itself originally reported, and the first quarter of 2026 alone was revised from 377 recorded closures to 658. A few weeks later, a separate story broke in the same corner of the data, Stats SA's companion count of individual insolvencies had gone dark entirely for five years and only resumed in July 2026. None of that means the sky is falling this year specifically. June 2026's actual liquidation count, 245 businesses, is almost identical to June 2025's. But it does mean a lot of the commentary on South African business failure over the past few years was working from numbers Stats SA has since admitted were wrong, and it's worth going through what the figures, properly read, actually say.
What gets me is not the liquidations themselves, it’s that we’ve spent years arguing about the state of small business in this country using numbers that were wrong by nearly double. If the data can be that far off, how many decisions, funding calls, or op-eds were built on a mirage. Rural areas like ours barely get counted properly in anything, so this doesn’t exactly build confidence that the picture out here is any more accurate.
I sat in a lot of credit committee meetings over the years and I can tell you, this is exactly the kind of revision that quietly wrecks confidence in a book of numbers everyone assumed was solid. Banks price business risk partly off macro liquidation trends, so if Stats SA had 2025 wrong by nearly double, some of the caution or complacency in lending decisions over that period was built on sand. I’m less surprised than Lerato that the insolvency count went dark for five years, agencies here have a habit of quietly dropping the hard series and reviving the easy ones. Doesn’t mean panic now, June looks steady, but I’d want two or three more clean quarters before I trust the trend line again.