In 2017 a technical change to mortgage underwriting reshaped how landlords buy. The Prudential Regulation Authority forced lenders to stress-test a landlord’s whole portfolio, not just the property in front of them — and the result was both a sharp fall in lending and a stampede into limited companies. GalimAI’s data is the clearest measure of where that stampede ended up.
What GalimAI’s own data reveals
The shift into companies is not a forecast in GalimAI’s data — it is the dataset. The 463,022 property-owning companies we map are dominated by the limited companies and SPVs that became the standard way to hold a portfolio once individual lending tightened. Each is linked to its named owners, its financing and its distress signals, so the regulation’s long shadow is visible at owner level.
That matters because the same rules created two reachable groups: the new companies formed to keep buying, and the heavily-geared portfolios that tightened lending now makes hard to refinance. Both are named lists in GalimAI, not abstractions.
What changed: the PRA portfolio rules, in plain terms
The PRA tightened buy-to-let underwriting in two phases. From January 2017, lenders had to apply stricter affordability and interest-rate stress tests. From 30 September 2017, ‘portfolio landlords’ — those with four or more mortgaged buy-to-lets — had to have their entire portfolio assessed on every new application, not just the property being bought.
The effect on individual lending was immediate; the effect on structure was lasting. Borrowing through a limited company sidesteps much of the personal-affordability squeeze and keeps mortgage interest deductible, so incorporation became the default for serious landlords.
The public backdrop
| Indicator | Figure | Note |
|---|---|---|
| Annual BTL loans | 117,000 (2015) → 72,400 (2018) | Lending falls as rules bite |
| Phase 2 start | 30 Sept 2017 | Whole-portfolio underwriting (4+ properties) |
| BTL purchases via company | 7.5% (2018) → 43% (2025) | Incorporation becomes the norm |
| 35% (2024) | Interim step | Share still climbing |
The two curves tell one story: individual lending down, company ownership up. GalimAI’s 463,022-company map is the destination of that shift, seen in full.
The most plausible mechanism
The channel is underwriting. By forcing whole-portfolio stress tests on individuals, the PRA made it harder for geared landlords to keep buying in their own name — while company structures, taxed and assessed differently, remained workable. Layered on top of Section 24, incorporation became the rational response, and the timing of the company surge from 2017 onward fits closely. We read this as a strong correlation with a clear mechanism, alongside tax and stamp-duty changes pushing the same way.
Sources
The proprietary figures in this study (the 463,022 companies, 1,000,000+ owners and the distress signals) are GalimAI first-party data. The public background figures are drawn from:
- Update on the PRA's portfolio-landlord underwriting rules - Mortgages for Business
- The buy-to-let sector and financial stability - Bank of England
Frequently asked questions
What did the PRA portfolio rules do?
From 2017 the PRA forced lenders to stress-test a landlord's whole portfolio, not just the property being bought, and apply tougher affordability tests. Annual buy-to-let loans fell from about 117,000 in 2015 to 72,400 in 2018.
Did the rules push landlords into companies?
Yes - alongside Section 24. The share of mortgaged buy-to-let purchases made through a company rose from about 7.5% in 2018 to 43% by 2025, because company structures sidestep much of the individual-affordability squeeze.
What does GalimAI's data show?
GalimAI maps 463,022 property-owning companies and 1M+ owners - the company-and-SPV market tighter lending created - each linked to its owners, financing and distress signals.
How can investors use this?
The newly formed companies still buying, and the heavily-geared portfolios that tighter lending makes hard to refinance, are both reachable named lists in GalimAI.