
The UK wealth exodus peaked in 2026. CitizenX projects a net 5,700 millionaire departures in 2027, with France overtaking Britain. HMRC data, the non-dom abolition record, and full methodology.
CitizenX projects the United Kingdom will lose a net 5,700 millionaires in 2027, with a P10–P90 interval of 2,950 to 10,050. That is a large outflow by any historical standard. It is also smaller than 2026, when our baseline puts the UK's net loss at 8,600, the largest in the world that year. The British wealth exodus is real. It is also past its peak, and in 2027 France takes the title nobody wants. This article sets out the evidence, the model, and every source behind those numbers. It is built on official statistics and real-time wealth migration data from CitizenX.
The UK holds roughly 3 million dollar-millionaires by UBS Global Wealth Report estimates. A net loss of 5,700 removes under a fifth of one percent of that stock in a year. What makes the British case worth a research note is not the raw scale. It is that the UK just ran the largest tax-driven migration experiment of the decade, published the data, and the data tells a more precise story than the public debate does.
Start with what is officially known, because the UK is unusual: it counts.
HMRC's annual publication on non-domiciled taxpayers put the combined population of non-doms and deemed-doms at 83,800 in the 2022/23 tax year, paying £12.5 billion in tax in the year to April 2024. Before the reform, roughly 9,500 individuals dropped out of non-dom status in a typical year, through departure, status change, or death. That churn number matters. It is the baseline against which any "exodus" has to be measured.
Then the regime was abolished, and two official reference points frame what happened next.
The first is the Office for Budget Responsibility's costing of the reform, published as a supplementary release in January 2025. The OBR built explicit behavioral assumptions into its revenue forecast: a share of the affected population leaves, and the departures follow a schedule, with 30% of individual departures assumed in 2025–26, 60% in 2026–27, and 10% in 2027–28. The government's own forecaster, in other words, expected the wave to crest in the 2026–27 tax year and fade after.
The second is what the early data showed. HMRC figures reported in August 2025 indicated non-dom departures running in line with, or below, the official forecast. The Times summarized it in five words: fewer non-doms left than feared. A separate assessment by the consultancy Chamberlain Walker in October 2025, working from the same official framework, estimated roughly 1,800 non-dom departures in the first months after abolition, somewhat above the OBR's implied pace. The two readings disagree about whether departures are tracking above or below the costing. Neither supports the tens-of-thousands figures that circulated in the press. The honest summary of the official record: departures in the thousands per year, elevated, front-loaded, and within sight of what the government priced in.
Our projection sits on that record, then scales beyond it, because non-doms are not the whole story. They are 84,000 households inside a millionaire population of three million.
The pressure on wealthy UK residents is not one policy. It is a stack, assembled over two decades and topped off twice in two years.
The long squeeze (2008–2017). The remittance basis survived two centuries before the 2008 introduction of the £30,000 annual charge for long-stayers. In 2017, deemed-domicile rules made anyone resident 15 of the past 20 years taxable on worldwide income. HMRC's series shows the non-dom population falling from roughly 90,000 toward 78,000 in the two years after that reform. That episode sits in our elasticity register as a measured, Tier A observation: a departure multiplier of roughly 1.9 on the affected population.
The abolition (2024–2025). Jeremy Hunt announced the end of the regime in March 2024. Rachel Reeves tightened it in the October 2024 Budget. From 6 April 2025, domicile left the tax code entirely. The replacement is the 4-year FIG regime: new arrivals pay no UK tax on foreign income and gains for four tax years, provided they were non-resident for the previous ten. After four years, worldwide taxation applies. Transitional sweeteners exist, a temporary repatriation facility at a flat 12% and asset rebasing to 2019 values, but the structural change is blunt. A regime that let wealthy foreigners live in Britain indefinitely on ring-fenced terms now expires after four years.
The part that moves people, not money: inheritance tax. The same reform made IHT residence-based. Long-term residents face UK inheritance tax on worldwide assets, with a tail of up to ten years after leaving. Excluded-property trusts, the standard shelter, lost protection. In our client conversations, and in the pattern of the corridor data below, the IHT change does more work than the income-tax change. Income tax moves portfolios. A 40% claim on a global estate moves families.
The stack keeps growing. The October 2024 Budget raised capital gains rates to 18% and 24% and capped agricultural and business property relief at £1 million from April 2026. The November 2025 Budget extended threshold freezes, added two points to dividend, savings, and property income rates, and announced a council-tax surcharge on homes above £2 million from 2028. And the November 2026 Budget arrives with the public finances still strained and a live political debate about wealth taxation. Our event register scores that risk; it does not need to guess.
Three independent lines point the same way.
First, the OBR's own departure schedule, 30/60/10 across three tax years, puts the modal departure in 2026–27 and the tail in 2027–28. Second, the 2017 precedent: the deemed-domicile shock produced a two-year adjustment, then the series stabilized. Policy shocks front-load, because the households most willing to move are the first to go, and the stayers have revealed something about themselves. Third, the mechanics of relocation. A household that decided in 2024 executed in 2025 or 2026. By 2027 the decided have left and the undecided face a smaller marginal incentive: the regime they knew is already gone.
Our severity score for the UK falls from 3.20 in 2026 to 2.40 in 2027, which the fitted elasticity converts to pressure multipliers of roughly 3.0 and 2.3. The result is the arc in the headline numbers: −8,600 in 2026, −5,700 in 2027. Still among the largest outflows in the world. No longer the largest. In 87% of our model runs, France's election-year surge pushes it past Britain.
One scenario would break the decay: a wealth tax, or a further IHT tightening, in the November 2026 Budget. That is precisely the kind of dated, scoreable event our quarterly revisions exist to absorb. If it happens, the 2027 number moves, publicly, with a changelog entry.
| Corridor | Projected movers 2027 |
|---|---|
| UK → UAE | ~2,750 |
| UK → Italy | ~1,250 |
| UK → Switzerland | ~1,000 |
| UK → United States | ~850 |
| UK → Spain | ~600 |
| UK → Australia | ~600 |
| UK → Cyprus | ~450 |
| UK → Monaco | ~350 |
| UK → Portugal | ~350 |
The UAE corridor is the largest single wealth-migration route in the world on our 2027 projections. Dubai Land Department buyer-nationality reporting shows UK buyers at their highest share in years, roughly a sixth of foreign purchases, and the UAE offers what the departing cohort is specifically fleeing: no income tax, no CGT, and no inheritance tax on the estate structures that the UK reform just exposed. Italy's €300,000 flat tax reads expensive until it is set against worldwide IHT exposure. Switzerland's lump-sum regime does the same arithmetic in French and German.
Two corridors deserve a note for what they say about the destination, not the origin. The US takes roughly 850 UK movers despite its own citizenship-based tax system; careers and capital markets still pull. And Portugal, the reflex answer of five years ago, takes only ~350, because the golden visa no longer confers what people assume: the NHR tax regime closed to new entrants, and our residence-conversion weighting strips out paper residencies that never become actual moves.
The exodus framing hides half the ledger. Our model projects roughly 3,150 millionaire arrivals to the UK in 2027: American professionals and founders (the largest single inbound group at ~1,500), Hong Kong families continuing on the BNO route, whose visa counts the Home Office publishes, South African and Turkish wealth, and the FIG regime's intended customers, wealthy new arrivals for whom four tax-free years is still a real offer. Net −5,700 is the difference between two large gross flows, 9,050 out and 3,150 in, not a one-way emptying. Countries with London's capital markets do not empty. They churn, and the composition of the churn is the real policy question.
Wealth-band composition of the projected departures: roughly 6,350 households at $1–5M, 2,150 at $5–30M, and 550 at $30M+. The IHT changes bite hardest at the top band, which is where trust structures concentrated. Detailed corridor and band analysis for the $30M+ tier is available to CitizenX Charter members.
Three reconciliations, briefly.
Against HMRC's counts. The official series tracks non-dom status holders, about 84,000 people. Our unit is the dollar-millionaire by investable wealth, about 3 million. Most millionaire leavers were never non-doms; they are British households responding to CGT, IHT, and the direction of travel. The HMRC series enters our model as a high-quality component signal for the most mobile subset, and it is why the UK carries a Tier A data grade.
Against the larger figures in circulation. Widely quoted claims put UK millionaire losses in the mid five figures. Those claims count differently than we do, and the official data gives no support at that scale. Where HMRC's series and the loudest numbers conflict, we side with HMRC, even though the smaller number is the less quotable one. A lab that only publishes dramatic findings is a marketing department.
Against "the exodus is a myth." The opposite error. A net loss of 8,600 millionaires in the peak year, front-loaded exactly as the government's forecaster scheduled, with a corridor to Dubai visible in the UAE's own property registry, is not a myth. It is a measured, medium-sized, policy-driven outflow. Both exaggeration and denial are failures to read the data.
The general framework is documented at citizenx.com/exodus/methodology. This section covers its application to the UK.
Definitions. Unit: an adult holding USD 1 million or more in investable wealth, excluding primary residence. Migration event: a change of primary tax residence across borders within the projection year. Household rule: the principal is counted; a spouse only if independently above the threshold. Visa approvals and property purchases are inputs, never counted as migration by themselves.
Structure. Projected gross outflow = millionaire stock × base emigration propensity × structural mobility factor × pressure multiplier. Outflows are allocated to destinations through corridor shares weighted by residence-conversion factors. Destination inflows are the sum of allocated corridor flows plus a rest-of-world component. Net flow = inflow − outflow.
Stock. UBS Global Wealth Report 2026 (data year-end 2025), which places the UK among the five countries above two million dollar-millionaires; we carry approximately 3.04 million for 2027 after trend growth, pending verification against the full databook.
Base propensity and mobility. A global base of 0.10% of millionaire stock per year, calibrated to the US Treasury's §6039G expatriation lists (formal-exit floor), HMRC's own pre-reform non-dom churn of roughly 9,500 exits per year on an 84,000 stock (mobile-subset ceiling), and US Census Bureau international-migration estimates. The UK's mobility factor is 1.30, the highest of any large economy in our set: English-language destinations, Commonwealth and Gulf networks, and no exit tax.
Pressure, from measured elasticity. The pressure multiplier is exp(β × S), with β = 0.344 ± 0.100 fitted to seven historical policy shocks with measured outcomes. Two are British: the 2017 deemed-domicile reform (HMRC's full official series; departure multiplier ≈ 1.9) and the 2025 abolition itself (interim multiplier ≈ 2.6, to be replaced with HMRC's post-reform releases as they publish). The register also includes FATCA, Norway's 2022–2023 wealth-tax episode, France's 75% supertax, and two reverse shocks, France's ISF abolition and Sweden's 2007 wealth-tax repeal, so the elasticity is trained on flows in both directions. The UK's 2027 severity score of 2.40 is built from dated events: the FIG and IHT regimes in full force, the OBR's 60%-in-2026–27 departure schedule tapering, the November 2025 Budget measures, and the scored risk of the November 2026 Budget.
Corridors and conversion. Anchored to Dubai Land Department buyer-nationality reporting, Italian Ministry of Finance flat-tax elector counts, Swiss lump-sum admissions, AIMA nationality data, and Home Office BNO statistics on the inbound side. Each destination carries a residence-conversion factor, high where regimes require presence (UAE, Italy, Switzerland), low where they do not (golden-visa destinations), so approval counts are never mistaken for moves.
Uncertainty. 2,000 Monte Carlo draws, vintage-locked seed. Every judgment-based parameter is drawn from a distribution; published intervals are the P10–P90. The UK's Tier A grade applies the narrowest dispersion in the model, which is why its interval, wide as it is, is tighter relative to its central estimate than any peer's.
Rounding and revision. Central estimates rounded to the nearest 50 above 1,000; raw values retained for accuracy grading. Quarterly revisions with a public changelog. Each November the prior vintage is graded against official series as they publish, and HMRC's next non-domiciled statistics release is the single most important test this model will face. We will publish the comparison either way.
Limitations, plainly. The interim post-abolition departure estimates come from a private economic consultancy's assessment of the official framework and will be replaced by HMRC data on release. The severity score for the November 2026 Budget is a structured reading of fiscal politics, not a prediction. Non-dom data illuminates the most mobile subset; scaling to all three million millionaires rests on the base-propensity calibration. The interval is wide because the truth is uncertain, and we publish the width.
How many millionaires are leaving the UK? CitizenX projects a net loss of 5,700 millionaires in 2027 (interval: 2,950–10,050), on roughly 9,050 gross departures and 3,150 arrivals. Our 2026 baseline, the peak year, is a net loss of 8,600.
Is the UK millionaire exodus exaggerated? The largest circulating figures are not supported by official data. HMRC statistics reported in 2025 showed non-dom departures in line with or below the government's own forecast. The outflow is real and large; the most dramatic numbers are not.
Has the non-dom regime been abolished? Yes, from 6 April 2025. It was replaced by the 4-year FIG regime: new arrivals pay no UK tax on foreign income and gains for four years, then worldwide taxation applies. Inheritance tax became residence-based, with up to a ten-year tail after departure.
Where are wealthy British people moving? The UAE first (~2,750 projected movers in 2027), then Italy, Switzerland, the US, Spain, and Australia. Dubai Land Department data shows UK buyers at a multi-year-high share of foreign property purchases.
Will the UK exodus get worse in 2027? Our central projection says no: 2026 was the peak, consistent with the OBR's own schedule that placed 60% of reform-driven departures in 2026–27. A wealth tax or further IHT tightening in the November 2026 Budget would change that, and our quarterly revisions would score it.
Governmental, parliamentary, and published-methodology research bodies, combined with real-time wealth migration data from CitizenX.
CitizenX advises clients on citizenship diversification and residence planning, including in destinations named above, and benefits commercially from interest in wealth migration. The method is built to check that incentive: the methodology is pre-registered before each vintage's numbers exist, projections are graded publicly against official series, revisions are logged whether they raise or lower the headline, and this Lab publishes findings that cut against its own book. This article contains two of them: the finding that the UK exodus is past its peak, and the finding that the largest circulating departure figures fail the check against HMRC's official data.