The index took in 2,775 more cars in August than it lost. That is an ordinary month. Where they went is not.
Where August’s cars went
Five of the six price bands grew. The £5k–£10k band took the largest share of new stock — 2,530 cars, most of them three- to five-year-old hatchbacks coming off finance. At the top, the £30k-and-up band grew by 1,195, the fastest it has moved since March.
Every figure below is the index as it stood at 00:05 on 1 September, counted against the same moment a month earlier. Nothing here is sampled and nothing is estimated.
Six bands sum to452,120· a month earlier they summed to449,345
Segment widths are each band's share of the index, drawn to scale. The six deltas sum to +2,775, which is the month's net change.
One band is going the other way
The £10k–£15k band lost 2,460 cars. It is the only band to shrink in both July and August, and it is the band most people are actually shopping in — 22 per cent of everything the index holds sits inside it.
The reason is not mysterious. Cars that would have been listed at £11,000 last year are being listed at £9,500, and the band below has absorbed them. £5k–£10k grew by almost exactly what £10k–£15k lost.
What that does to a £15,000 budget
Less than the headline suggests, and more than nothing. There are still 99,480 cars in the band, so nobody is short of choice. But the cars leaving it are leaving downwards, which means the median car inside the band is a little older and a little higher-mileage than it was in July.
The practical move is to search the band below as well as your own. Set the range from £8,000 rather than £10,000 and you pick up 2,530 cars that were priced above you a month ago.
Method
Counts are taken from the live index at 00:05 on the first of each month. A car is counted once, at its current asking price, whether it is listed by a dealer or a private seller. Cars withdrawn during the month are not counted; a car whose price changed is counted in the band it ends the month in. No sample, no weighting, no seasonal adjustment.



