Oakwood Great Oak Ltd v HMRC [2026] UKFTT 1138 (TC) is not a free pass for every “project house”. It is a hard, fact-heavy application of the Court of Appeal’s test in Mudan [2025] EWCA Civ 799: when does a former dwelling stop being “suitable for use as a dwelling” for Stamp Duty Land Tax?
The First-tier Tribunal allowed the purchaser’s appeal. A £2.4m acquisition of “Great Oak”, Bushey Heath, was correctly returned as non-residential. HMRC’s closure notice charging residential rates failed. For CFOs buying through SPVs, PE propcos or development vehicles, the cash difference between residential and non-residential SDLT — plus corporate surcharges on high-value dwellings — is material. The case also shows how quickly HMRC will fight that line after Mudan.
Primary sources: the FTT judgment, Mudan UT, Mudan CA, Ridgway [2024] UKUT 36 (TCC), s.116 FA 2003, Schedule 4ZA, Quastels’ client note, Claritax’s case summary, HMRC’s SDLT Manual, and GOV.UK SDLT rates.
What was actually at stake
Oakwood Great Oak Ltd completed on 29 November 2022 for £2,400,000. The SDLT return treated the deal as non-residential. HMRC opened an enquiry in September 2023 and issued a closure notice on 22 December 2023 concluding the property was residential under s.116(1)(a) FA 2003. Review upheld that view in May 2024. The company appealed.
Section 116(1)(a) says residential property includes “a building that is used or suitable for use as a dwelling, or is in the process of being constructed or adapted for such use”. Schedule 4ZA uses the same language for higher rates on additional dwellings. If the building is still residential, corporate buyers can face higher residential rates and, depending on structure and value, further company charges. If it is not, the lower non-residential tables apply. On a multi-million pound site bought for demolition and redevelopment, the SDLT delta is often six figures — sitting in the completion funds flow, SPA tax warranties and fund model.
The legal frame after Mudan
Mudan killed the lazy argument that a house must be immediately habitable to be “suitable for use as a dwelling”. Lewison LJ was blunt: reading “suitable for use” as “suitable for immediate use” rewrites the statute. The focus is the fundamental characteristics and nature of the building, not a snapshot of whether someone could sleep there tonight.
Where the building was previously a dwelling, the live question is whether it has lost that identity by the effective date. The Upper Tribunal’s seven considerations — endorsed by the Court of Appeal — remain the working checklist: previous use and how long ago; where the building sits between “desirable house needing update” and “empty shell”; works for habitability versus cosmetic improvement; whether defects are fixable without works so hazardous or structurally compromising that remediation is not viable; safety of occupation (relevant, not decisive alone); avoiding empty labels like “minor” or “fundamental” as free-standing tests; and whether needed works mean the building no longer has the characteristics of a dwelling.
No single factor decides it. That multifactorial insistence is what Oakwood applied — and what HMRC’s “still standing, therefore residential” posture keeps under-weighting. Ridgway points the same way: no gloss on the statute; evaluative judgment of all circumstances.
The facts that moved the needle
Great Oak was a 1930s detached house with a 1960s extension. It had been a dwelling for decades. By completion it had been vacant for roughly three to four years. The tribunal accepted extensive deterioration: damp and mould, water ingress, cracking, damaged finishes and windows, service defects, vandalism signs, and structural concerns on the rear terrace, retaining wall and extension.
The decisive cumulative layer was asbestos. Licensed materials were widespread — basement, services, roof void — in a high-risk condition. Removal required licensed contractors, negative-pressure enclosures, air monitoring and clearance. Critically, remediation would not leave a house intact and ready to occupy: stripping contaminated materials would force removal of associated services and fabric, then substantial reinstatement before the building could function as a dwelling again. A costing exercise put remediation around £2.25 million. The tribunal treated the figure with caution on precision, but accepted it as evidence of exceptional scale.
HMRC’s core pushback was familiar: the house still looked like a house; it had not collapsed; asbestos can be removed; therefore it remained residential. The tribunal rejected the pure “theoretically fixable with unlimited resources” approach. Almost any standing structure can be repaired if cost and disruption are ignored. That reading empties s.116 of content. Practical consequences and cumulative condition matter to identity.
Factors that helped HMRC were real: prior long residential use, external recognisability, layout still readable as a house, no imminent total collapse, remediation not physically impossible. They were not enough once damp, structure, safety and asbestos-driven reinstatement were weighed together. The property had “crossed the line” contemplated in Mudan.
What CFOs should lock into deal process
1. Classify at the effective date. The EDT is completion (or substantial performance). The best evidence files exist at or around completion and are not reverse-engineered solely for a planning fight. In Oakwood, the director accepted the site was always a demolition/redevelopment play. Intent alone does not decide SDLT; condition does. Do not confuse the two in board papers.
2. Build a Mudan file, not a marketing brochure. For any SPV buying a vacant or distressed former dwelling, require: dated photographic survey; structural report addressing integrity; asbestos survey with risk category and method; independent costings separating strip-out/reinstatement from optional upgrade; vacancy and marketing history; and a short specialist SDLT memo applying the seven Mudan factors to this asset.
3. Drop both lazy models. Do not run “any fixable defect keeps it residential” in IC papers — that is closer to HMRC’s losing argument than to the Court of Appeal. Equally, do not assume every wreck qualifies as non-residential. Cosmetic wrecks, locked-up but structurally sound houses, and “needs a new kitchen and rewire” stock will still lose. The line is identity loss under cumulative assessment — not buyer preference for demolition.
4. Price the enquiry risk in the funds flow. Even with a strong non-residential return, expect HMRC interest where the asset still photographs as a house from the kerb. Hold a contingency equal to the residential-minus-non-residential SDLT delta through the enquiry window. Align SPA tax indemnities, completion accounts and seller disclosures on condition.
5. Separate planning strategy from tax classification. Seeking demolition consent is commercially rational. It is not itself the SDLT test. Instruct surveyors and asbestos consultants that their reports may be used in an SDLT dispute, so scope includes reinstatement consequences — not only “can asbestos be removed?” Everyone in Oakwood accepted removal was possible. The tribunal’s point was what removal did to identity and function.
6. Run the full rate matrix before funds release. Non-residential classification can also affect higher rates for additional dwellings and certain corporate residential charges. Check standard residential, higher rates, non-residential and any relief claims before you sign the funds flow.
Governance, not folklore
Oakwood sits downstream of Mudan and Ridgway. Earlier FTT illustrations such as Bewley remain fact-specific examples, not freestanding tests. HMRC’s operational stance after losing high-profile “derelict dwelling” arguments has not been to abandon residential classification — it has been to force taxpayers to prove the line has been crossed with technical depth. For PE and family-office real-estate programmes, add a standing control: any acquisition of a vacant former dwelling above a de minimis price needs a pre-filing SDLT classification memo signed by tax and the deal lead, with the Mudan checklist attached. No memo, no funds release for the SDLT line.
- Pull live deals and recent completions where the asset was vacant, fire-damaged, asbestos-heavy or marketed as a development opportunity.
- Re-read the SDLT return basis against Oakwood and Mudan. If you filed residential out of caution on a true wreck, quantify overpayment and limitation with advisers.
- If you filed non-residential, stress-test the evidence pack as if HMRC’s review letter arrived tomorrow.
- Update IC templates so “suitable for use as a dwelling” is a documented judgment, not a conveyancer tick-box.
- Brief boards that Oakwood helps genuine identity-loss cases; it does not licence aggressive non-residential filing on tired buy-to-lets.
Bottom line
Oakwood Great Oak confirms that a former house can leave the residential SDLT net without falling down. The price of admission is cumulative, contemporaneous evidence — structure, vacancy, safety and especially contamination that forces invasive reinstatement — measured against Mudan, not against hope or demolition preference. CFOs who treat SDLT classification as a last-minute land-tax formality will keep funding HMRC’s enquiry machine. CFOs who lock the Mudan file into the deal process will price risk correctly and defend the return.
This article is general information for finance leaders, not advice on any specific transaction. SDLT outcomes turn on the facts at the effective date and on current law and HMRC practice. Take advice on your deal.