The Missing Link in the UK Housing Supply Data

Good morning,

Yesterday, HM Land Registry quietly announced something that sounds like pure admin: from 28 August, every sale record gets two new ID codes attached.

From 28 August, every sale record gets two new ID codes, making it much easier to link properties to data on things like flood risk, conservation areas and planning restrictions.

Sounds boring, I know.

But it could reveal something the "stock at a 12-year high" headlines are missing: how much of that supply is actually usable.

Let’s dive in.

In this newsletter, you'll find...

• The boring announcement that isn't boring

• If you remember one thing, make it this

• Feedback

The boring announcement that isn't boring

Right, so here's what's actually happening.

From 28 August, Land Registry starts publishing UPRNs (Unique Property Reference Numbers) and INSPIRE parcel IDs alongside Price Paid Data — the record of every registered sale in England and Wales, going back to 1995.

Translation? Every sold price can now be pinned to one exact address and one exact plot of land, instead of a fuzzy postcode match.

Doesn't sound like much. It's actually the join key the property data world has been faking with string-matching for two decades.

For context on scale: Land Registry safeguards nearly £9 trillion of ownership across England and Wales, and processed 103,398 transactions for value in July alone. That's the size of the dataset that just got a lot easier to trust.

Why this matters right now

Remember Rightmove's number from a couple of weeks back, stock for sale at a 12-year high, asking prices down 2.0% in a month? Everyone read that as "more choice, softer prices."

Here's the bit nobody's connecting: once a sale joins cleanly to a UPRN, it also joins cleanly to flood risk zones, conservation areas and listed-building status sitting on the government's Planning Data platform, 100+ datasets, all queryable by that same ID.

Which means we can finally ask the question headline stock figures can't: how much of that "12-year high" is genuinely sellable, mortgageable stock — and how much is sitting in a flood zone, a conservation area, or behind an Article 4 direction that kills your renovation plans?

The catch hiding in the fine print

Before you get too excited, a proper catch. Land Registry isn't backfilling this for anything published before 28 August. Three decades of history, still unmatched. Only new monthly files get the clean join from here.

So the "hidden supply risk" story isn't fully visible yet. We're at day one of being able to build it properly, not day done.

Worth noting too: the data comes under the Open Government Licence, so anyone (you, me, a rival newsletter, a mortgage lender's risk team), can start building this join today. This isn't a niche advantage for long.

So what do you do with this?

If you're viewing anything right now, before you get excited about "lots of choice," check three things: EPC rating (data's had UPRNs since 2021, so this join already works), flood risk zone, and any conservation area or Article 4 flag on the address.

Two identical £350,000 semis (one clean, one flagged Flood Zone 3 and Article 4 restricted), aren't the same asset, even if the portal lists them side by side at the same price.

If you're screening a city, this is your excuse to stop trusting postcode averages. They blend the clean stock with the encumbered stock into one number. You want to know which one you're actually being offered.

And if you're a landlord thinking about that EPC C deadline in 2030, this same join is your shortcut — one lookup instead of chasing three separate registers to work out what you're dealing with.

If you remember one thing, make it this

A 12-year high in "homes for sale" isn't the same as a 12-year high in homes you'd actually want to buy.

From 28 August, for the first time, we've got the tools to tell the two apart (property by property, not postcode by postcode).

Give it a month for the join-key data to properly fill out. Then run the numbers on your own patch and reply, I'd genuinely love to know what you find.

And if you need any help compiling this together, be free to reach out to me.