Private regs as investments: the third that never sees a car

If you've ever wondered why the DVLA seems to release the same 24-plate combinations week after week and they all disappear without ever appearing on a car, the maths now backs the hunch. We pulled 85,075 plates that left the DVLA available list between 16 January and 28 March 2026, gave each at least 30 days to find a vehicle, then asked DVLA's Vehicle Enquiry Service¹ what car they ended up on. 29,952 came back unassigned — roughly one in three.

That's £11.3m of plates, bought and held. The shape of the unassigned half doesn't look much like the assigned half — and the cleanest signal is the year code.

What we mean by 'unassigned'

When you buy a plate from DVLA, there's usually a lag of a few weeks before it lands on a car — paperwork, a fresh V5C, the buyer deciding which vehicle to put it on. The median lag for plates that do get assigned is about 19 days, and by 30 days half the cohort is on the road. By 90+ days the assignment rate plateaus at 67–68 per cent and stops climbing. We've trimmed to the 30-day-old cohort here, so 'unassigned' means: the plate has had time to land on a car, and hasn't. We've already used the assigned 55,123-plate half of the same cohort to build the Top 10 cars wearing private registrations piece. This is its mirror — the half that never makes it onto a bumper.

The price ladder

The simplest cut first. Sort the cohort into price bands and the unassigned share climbs steadily:

Cheap plates skew toward drivers; pricey plates skew toward portfolios. By the time you're spending £1,000, almost half the buyers aren't fitting them.

The year code is the tell

Now restrict to current-format plates (AA00 AAA) and look at the year-code block — the two digits in the middle that tell you which six-month window the combination belongs to. The pattern is sharp:

And inside that 'New' bucket the two reigning champions are the most recent year codes of all. 56.3% of 24 plates and 59.4% of 74 plates in the cohort are sitting unassigned — well above any other slice we cut. Examples like ST25EVO, JO26ANN, RO26CCO and BR24ATT all changed hands at the £1.8k–£2k mark and none of them has touched a car yet.

The mechanism's obvious once you see it. A 24 or 74 plate has to wait for the right 24-registered vehicle to come up second-hand at the right time, and most of those cars are still on their original PCP. The plate is the speculative half of the trade — bought now in the hope someone wants it in three years' time.

Compare with 52-plate (28.8% unassigned), 62-plate (27.9%) and 65-plate (29.5%) — the cheapest, oldest, plainest year codes. They get fitted. They're tools.

Prefix is for drivers; NI is for investors

The three formats sit at very different unassigned rates:

Prefix plates are the closest thing in the data to a 'driver's' format — bought to fit, almost never bought to hold. That tracks: a prefix plate carries an old letter code, has no year-matching pressure, and tends to spell something the buyer already wanted on a car they already own.

NI dateless plates sit at the other end. They're 40.8% unassigned — the highest of any format — and the ones nobody bolted on actually cost £48 more on average than the ones that did get fitted (£478 vs £430). It's a small premium but it points the right way: when an NI plate is being acquired as an investment, the buyer's willing to pay slightly more for it. The top of that pile is a run of GIL- and OIL-coded Northern Ireland combinations at £1,299 each, all unsold to a vehicle 30+ days in.

The plateau, and the non-finding

Within the trimmed cohort there's a final time gradient. Plates that have been off DVLA's list for 45–59 days are 42.1% unassigned; 60–89 days, 35.7%; 90+ days, 32.6%. So even after the trim, the longer you wait the lower the rate falls — but it never falls past about a third. That's the floor. Roughly 32% of every batch DVLA releases is going to portfolios, full stop.

The non-finding is worth flagging too, because we expected the opposite. We labelled both halves of the cohort for first names, surnames, popular words and animal terms. Unassigned plates contain a first name 19.4% of the time; assigned plates 18.4%. Popular-word plates: 4.2% vs 3.9%. Almost no daylight. Investors aren't buying obviously 'generic' or 'resaleable' patterns at any higher rate than drivers are buying named plates — at least, not in a way our labels can see. The split between investment and ownership isn't really about what's on the plate. It's about the shape of the plate: the year code, the format, the price.

The takeaway

If you read the assigned cohort, you'd think private plates are mostly a car-customisation market. Read the unassigned half and a different story shows up: about a third of supply, £11.3m in 10 weeks, sat in safes against the day someone wants a year-matched plate for a new car. The marketplace is two markets — the one that ends on a bumper, and the one that ends in a spreadsheet.

If you're buying to drive, you're already shopping in the right half of the data: a prefix plate, an older year code, a number that means something. If you're buying to hold, the same data is telling you what your competition is paying for.