Key takeaways
- Nothing about capital gains tax has been announced for the Budget on 28 October 2026. The rates in force for 2026 to 2027 are 18% and 24%, above a £3,000 exempt amount.
- On a £50,000 gain, someone earning £70,000 pays £11,280 today. Taxed at income tax rates of 20%, 40% and 45%, the same gain costs £18,800.
- HMRC's ready reckoner, published 24 June 2025, costs a 1 percentage point rise in the higher capital gains tax rate at minus £15 million in 2026 to 2027.
- A 1p rise in the income tax basic rate is costed at £6,900 million in that same year, roughly 460 times the size of that CGT row.
- CGT liabilities hit a record £24.2 billion across 584,000 taxpayers in 2024 to 2025, and 45% of the tax came from people with gains above £5 million.
What capital gains tax at Budget 2026 could cost on a £50,000 gain
You want to know whether the Budget 2026 puts capital gains tax up, and what that does to a gain you're holding. The position as at 26 September 2026 is that no change has been announced for it, and GOV.UK still gives the rates in force as 18% and 24%. There's no new rate to apply to anything, and anyone quoting one is quoting a guess.
What can be priced is the option that comes up most often: taxing gains at income tax rates instead of the separate schedule they have now. The Office for Budget Responsibility said in July 2026 that "the Chancellor has commissioned the OBR to produce our next forecast on Wednesday 28 October 2026". That date is the confirmed part. The table below runs three gains through both rate schedules, for someone on £35,000 and someone on £70,000, after the £3,000 annual exempt amount comes off.
| Gain | Other income | CGT at 18% and 24% | At income tax rates | Difference | Share of the gain |
|---|---|---|---|---|---|
| £10,000 | £35,000 | £1,260 | £1,400 | £140 | 12.6% to 14.0% |
| £50,000 | £35,000 | £10,364 | £15,746 | £5,382 | 20.7% to 31.5% |
| £200,000 | £35,000 | £46,364 | £81,089 | £34,725 | 23.2% to 40.5% |
| £10,000 | £70,000 | £1,680 | £2,800 | £1,120 | 16.8% to 28.0% |
| £50,000 | £70,000 | £11,280 | £18,800 | £7,520 | 22.6% to 37.6% |
| £200,000 | £70,000 | £47,280 | £85,893 | £38,613 | 23.6% to 42.9% |
The £140 row and the £38,613 row are the same policy change. What separates them is where the gain lands once it's stacked on top of your income, which is the mechanism the rest of this piece takes apart.
Only the Budget date is settled, and the 2026 to 2027 rates are 18% and 24%
Four things are confirmed, each with a date on it.
- The rates. GOV.UK states that "if you're a higher or additional rate taxpayer, you'll pay 24% on your gains from 6 April 2026", and that a gain inside the basic band is charged at 18%.
- The allowance. "For the 2026 to 2027 tax year the allowance is £3,000." Trusts get £1,500.
- Business Asset Disposal Relief. The relief rate reached 18% on 6 April 2026, the second of two steps announced at Autumn Budget 2024, up from 10% and then 14%. The OBR records a lifetime limit of £1 million on it.
- Carried interest. On this one class of gain, alignment has already happened. GOV.UK says you pay "Income Tax and National Insurance contributions on carried interest you receive from 6 April 2026 instead", with a multiplier that CenTax describes as reducing taxable profits by 27.5%.
That leaves the capital gains tax rates for 2026/27 at 18% and 24%, with the £3,000 allowance underneath. What isn't confirmed is anything about Budget 2026 capital gains tax changes. The Institute for Government's Thomas Pope and Jill Rutter wrote on 7 September 2026 that the Chancellor's "task for the budget is threefold", naming a course for the parliament, reassuring markets, and funding additional spending. That's a description of pressure, not of a measure. Reporting about what a Budget might contain is reporting, and none of the arithmetic here rests on it.
How the 18% and 24% split works, and why unused basic rate band decides it
Capital gains tax isn't charged on a separate ladder of its own. It's charged on top of your income, which is why two people with identical gains can owe very different amounts.
HMRC's own steps are: work out your taxable income after the £12,570 Personal Allowance, work out your total taxable gains, deduct the £3,000 allowance, then add what's left to your taxable income. Whatever falls inside the basic rate band of £37,700 is charged at 18%. Everything above it is charged at 24%.
HMRC gives two worked examples. On £20,000 of taxable income and a £12,600 gain, the bill is £1,728, all at 18%. On the same income and a £52,600 gain, £17,700 is charged at 18% and £31,900 at 24%, and the bill is £10,842. One more mechanical detail is worth knowing: the allowance is used "against the gains that would be charged at the highest rates", so it shelters 24% money before 18% money.
So the £35,000 earner in the table has £22,430 of taxable income and £15,270 of basic rate band spare. The £70,000 earner has £57,430 of taxable income and none spare, which is why a £10,000 gain costs that person £1,680 rather than £1,260. Working out which lots you sold, and at what cost, comes first; the UK share matching rules decide that, and LedgerTouch applies them when it computes a lot-level gain.
Alignment wouldn't be a flat rise, it would be a steepening
The last column of the table is the interesting one, and the chart plots it as the jump in the share of the gain taken in tax. The rises run from 1.4 percentage points to 19.3 percentage points across the same six cases.
At the bottom, a basic rate taxpayer with a £10,000 gain sees 12.6% of it go in tax today and 14.0% under alignment. That's the whole effect: 1.4 percentage points, because 18% becomes 20% and nothing else moves. At the top, a £70,000 earner with a £200,000 gain goes from 23.6% to 42.9%, because most of the gain crosses into the 40% band and then the 45% band above £125,140.
The steepening is not a quirk of the design. It's what happens whenever a flat two-rate charge is replaced by a progressive one, and it's why a single headline rate doesn't describe the effect on anybody. The UK already runs a third schedule on the same principle: dividends above a £500 allowance are taxed at 10.75%, 35.75% and 39.35% for 2026 to 2027. That schedule sits below the 18% charge on gains at the basic rate and well above the 24% charge higher up. A fuller comparison of the three is in our table of UK investment tax rates.
HMRC's own ready reckoner costs a large capital gains tax rise as a net loss
The strongest objection to any of this comes from HMRC's own model. It publishes a ready reckoner of illustrative tax changes, last updated on 24 June 2025, built on the OBR's March 2025 forecast and assuming an April 2026 start.
Its CGT rows are extraordinary reading. A 1 percentage point rise in the higher rate is costed at minus £15 million in 2026 to 2027, £80 million in 2027 to 2028 and minus £30 million in 2028 to 2029. A 10 percentage point rise gets worse rather than better: minus £540 million, minus £2,060 million and minus £3,565 million. A 10 point rise in the lower rate is costed at minus £135 million by 2028 to 2029. Going the other way, raising the annual exempt amount by £500 is costed at nothing in 2026 to 2027 and £35 million in 2027 to 2028. That tells you how little revenue sits in the bottom of the base.
HMRC is explicit about the signs and the reason. "Negative signs are used in the CGT illustrative tax changes to indicate net losses." And the note attached to the rows says the changes are "non-linear and asymmetrical", adding that "very large tax rate rises can reduce exchequer yield due to taxpayer behavioural impacts".
Put that beside the income tax page of the same document. A 1p change in the basic rate is costed at £6,900 million in 2026 to 2027, and 1p on the higher rate at £1,600 million. Against minus £15 million for a percentage point of CGT, the basic rate penny is roughly 460 times the size, and it points the other way. On HMRC's own arithmetic, the rate on gains is close to the worst lever in the building for raising money.
CenTax says £11.3 billion, and that doesn't contradict the ready reckoner
Two credible sources appear to disagree here, and refereeing them is the useful part. The Centre for the Analysis of Taxation published a technical note in August 2025. It estimates that its capital gains package raises £11.3 billion, or £11.8 billion if carried interest is included, after behavioural response and uprated to 2026 to 2027. An earlier CenTax report, dated 11 October 2024, put the same package at £14 billion against an OBR baseline of £16.2 billion for 2025 to 2026. The smaller figure in the 2025 note is not a retreat: its baseline already contains the October 2024 rate rise, so there is less left to collect.
The difference is in what each one prices. The ready reckoner moves a rate and leaves the base alone. CenTax's reform, in its own words, would "introduce an investment allowance for the 'normal rate of return', equalise the marginal rate on gains with the marginal rate on income, remove death uplift and apply rebasing on arrival and deemed disposal on departure".
That list matters because capital gains tax is charged on realisation. Nobody pays it until they sell, so a higher rate on an unchanged base makes waiting more attractive. CenTax notes that accrued gains "currently are exempted because the asset is transferred at death", which makes waiting indefinitely a settled outcome rather than a deferral. Raise the rate on its own and the behavioural response has somewhere to go. Remove the exit, and it has less. CenTax makes that point about HMRC's numbers directly, saying they "crucially assume that the tax rate is raised without any other changes", and that "if rates were increased without other reforms, then revenues would indeed be much lower". CenTax's own distribution table shows where the money would come from: 22.2% of it from the top 0.1% of the income distribution and 32.7% from the 90th to 99th percentile.
So the two numbers are answers to different questions. One says a rate rise by itself raises nothing. The other says a rate rise plus base reform raises £11.3 billion. Both can be right, and together they say something more useful than either alone: in capital gains tax, the rate is the headline and the base is the revenue. CenTax's own worst case is £9.6 billion. Its 2024 report also found that "over half (51%) of CGT payers in 2020 would have been better off under the proposed package", because the investment allowance cuts the effective rate for many of them. An alignment described as a tax rise on investors is not, on the only detailed independent modelling of it, a tax rise on most of them.
584,000 people pay it, and 45% of it comes from fewer than 1% of them
The OBR's November 2025 forecast put CGT at £20.3 billion in 2025 to 2026, "1.6 per cent of all receipts", equivalent to £705 per household. HMRC's statistics, updated on 27 August 2026, show why that total is so concentrated. In 2024 to 2025, 45% of CGT came from people with gains of £5 million or more, a group HMRC puts at "less than 1% of CGT taxpayers each year". And 52% of gains came from the 17% of CGT-liable individuals with taxable incomes above £125,140.
HMRC's asset-level sample for 2023 to 2024 puts financial assets at 79% of gains, so capital gains tax on shares held by UK investors does most of the work, not property. The base is narrow too. The 584,000 taxpayers are "around 1.5% of the number of people who pay Income Tax". Cutting the annual exempt amount from £12,300 in 2022 to 2023 to £6,000 and then £3,000 pulled up to 163,000 extra taxpayers into the charge. HMRC says the 76,000 added by the April 2024 cut alone "contributed less than 1% of the overall increase in gains and CGT liabilities". That is the awkward fact under any reform. Broadening the base adds tax returns much faster than it adds revenue, while the revenue itself sits with the people an alignment would move to 40% and 45%.
Speculation moves the numbers before any rate does
There's a documented precedent for what happens in the weeks before a Budget that people expect to touch CGT, and it comes from HMRC rather than from commentary. Its statistics say plainly that "there was public speculation ahead of Autumn Budget 2024 that the rates of CGT would be increased". Part of what followed, HMRC adds, can be attributed to "taxpayers bringing forward the timing disposals to before the Budget date of 30 October 2024".
The scale is in the same release. Gains reported in 2024 to 2025 hit £127 billion, an 82% rise, liabilities hit £24.2 billion, an 89% rise, and the taxpayer count rose 45% to an all-time high of 584,000. The OBR describes the same pattern in its forecast, putting a temporary rise in 2025 to 2026 receipts down to "people bringing forward asset disposals in advance of anticipated rate rises at the October 2024 Budget". HMRC notes the equivalent effect on Business Asset Disposal Relief, where announcing the increases early gave "an opportunity to bring forward disposals to be taxed at the lower 10% rate".
That has a cost as well as a saving, and the cost is rarely stated. A disposal moved forward crystallises a bill in the current tax year that wasn't otherwise due, and uses up an annual exempt amount that would otherwise have covered a future gain. HMRC expected the 2024 rate rise alone to leave 264,000 individuals paying more in 2025 to 2026. Where the disposal is a fund holding moving into a wrapper, the trade-off is the one set out in our piece on Bed and ISA costs.
What these figures cannot tell you
The table is a rate swap and nothing else. It applies the 2026 to 2027 income tax bands to a gain under today's rules. So it ignores every base change a real reform would carry: an investment allowance, relief for inflation, loss treatment, business reliefs and anything done to death uplift. CenTax's own package includes an investment allowance precisely because equalisation without one taxes a return that is partly just inflation.
The ready reckoner is an illustration, not a costing of a policy. It runs on the OBR's March 2025 forecast, and HMRC warns that adding its components together should "be considered only as a general guide". CenTax's estimate comes from a sample of 2019 to 2020 administrative data uprated to 2026 to 2027, which is a model rather than an outturn. The 2024 to 2025 CGT statistics are a poor baseline for anything, because they are the year forestalling distorted.
The bands used here are the ones on the GOV.UK income tax page, which also notes that "income tax bands are different if you live in Scotland". And none of this is a forecast of the Budget. It's the arithmetic of one option, priced against the rules actually in force.
What would change these numbers
Whatever the Budget 2026 capital gains tax decision turns out to be, the informative part is whether it touches the base or only the rate. A rate move on its own runs straight into the ready reckoner's minus signs. A rate move packaged with death uplift, an investment allowance or a deemed disposal on departure is a different measure with a different yield. The £11.3 billion estimate only holds for the full package.
Three smaller things would move the table directly. A change to the £3,000 annual exempt amount rescales every row, and hits the £10,000 gain hardest in proportion. A change to the £50,270 or £125,140 thresholds moves where the gain lands without touching a CGT rate at all. And a change to reporting, such as extending the 60-day property window that has applied to completions since 27 October 2021, alters when the bill falls due rather than how big it is.
The tell to watch on the day is which document the numbers come from. A measure with a certified Exchequer impact behind it has been scored; the 2024 rate rise carried +£1,370 million for 2026 to 2027 in its impact note. A measure without one is a direction of travel, and the distance between the two is the distance between the ready reckoner's minus signs and an £11.3 billion estimate.