The exemption everyone talks about is only worth something if you win. The relief nobody talks about is only worth something if you lose, which is the more likely outcome.
What is the actual difference in the eyes of HMRC?
A spread bet is a bet. HMRC treats it as gambling, which for a retail client means it sits outside capital gains tax, outside income tax and outside stamp duty. The relevant guidance is BIM22020 and it has been stable for years. A contract for difference is a financial instrument, and gains on it fall inside the capital gains regime like any other disposal. That is the whole distinction. Economically the two products do almost the same job: leveraged exposure to a price without owning the underlying, margined, financed overnight, closed out at 50% of margin under FCA rules. The difference is a legal classification, and it produces two quite different outcomes at the end of the tax year.
Why does that make spread betting sound obviously better?
Because on the winning side it is better, and the arithmetic is not subtle. For 2025/26 the annual capital gains allowance is GBP 3,000. Above that a basic-rate taxpayer pays 18% and a higher-rate taxpayer pays 24%. So a CFD trader who nets GBP 20,000 in a year has GBP 17,000 in charge, and hands over GBP 3,060 or GBP 4,080 depending on band. The spread better with the same result hands over nothing.
Four thousand pounds is a real number and it explains why almost every comparison of the two products stops here. It is also, for most people reading those comparisons, completely irrelevant, because it describes a situation they will not be in.
What happens when you lose?
This is the half of the comparison that gets left out. A CFD loss is an allowable capital loss. You can set it against capital gains in the same year, and if you have none, you can carry it forward indefinitely against future gains, provided you report it in time. It is an asset on your tax return.
A spread betting loss is nothing. It relieves no gain, offsets no income and carries forward to no future year, because gambling losses are outside the tax system in exactly the same way the winnings are. The exemption is symmetrical, and traders who choose spread betting for the tax treatment are usually only thinking about one side of that symmetry.
How likely is each outcome, honestly?
An April 2026 analysis by The Investors Centre of the risk disclosures published by 14 FCA-authorised UK forex brokers put the mean retail loss rate at 68.4% and the median at 71.5%, with individual firms ranging from 50% to 76%. Those are the brokers’ own numbers, published because the regulator requires it, collected and averaged by somebody outside the industry.
A median of 71.5% across that forex sample is the number to hold on to: at the middle firm, more than seventy accounts in every hundred ended the reporting period down. If you are choosing a product on the strength of its tax treatment, weight the two outcomes by how often each one happens. The question that carries the weight is what becomes of your GBP 4,000 loss, and the answer under a spread bet is that it does not exist for tax purposes and never will.
One hole in that reasoning is worth flagging while the number is still on the page. Loss rates are disclosed per broker, never per wrapper. Nothing in any of those fourteen disclosures separates the clients who were spread betting from the clients trading CFDs, so the probability this whole argument leans on has been borrowed from a population that mixes the two. If spread betters and CFD traders lose at meaningfully different rates, the arithmetic below is calibrated on the wrong denominator, and no published figure anywhere can currently tell you whether they do. Anyone compiling these disclosures, including the site that compiled these, is working with a split the regulator never asked firms to report.
| Your year |
Spread betting |
CFD |
Which is better |
| Gain of GBP 2,000 |
No tax |
No tax, inside the GBP 3,000 allowance |
Neither, identical |
| Gain of GBP 20,000 |
No tax |
GBP 3,060 to GBP 4,080 due |
Spread betting, clearly |
| Loss of GBP 4,000, no other gains |
Relieves nothing |
Carry forward against future gains |
CFD |
| Loss of GBP 4,000, GBP 4,000 gain elsewhere |
Relieves nothing |
Offsets the gain, tax saved |
CFD, decisively |
| Break-even year |
No tax |
No tax |
Neither |
2025/26 rates: a GBP 3,000 annual exempt amount, then 18% at basic rate and 24% at higher rate. Losses have to be reported before they can be carried forward. The exempt amount has been cut more than once in recent years, and every figure in the CFD column moves with it.
The allowance does more work than people expect
Look at the first row of that table again. Below GBP 3,000 of net gain, the tax advantage of spread betting is precisely zero, because the CFD trader’s gain is covered by the annual exempt amount anyway. A great many retail accounts never clear GBP 3,000 of net profit in a year, which means a great many people have selected a product for a benefit they are not receiving.
It gets worse if you have other investments. The allowance is one allowance across all your capital gains, so if you have already used it selling shares or a fund, your first pound of CFD profit is taxable. That cuts the other way and genuinely favours spread betting.
Neither product wins outright. The answer depends on the rest of your tax position, and no comparison table can tell you which row of it you are standing in.
What does the exemption cost you in pricing?
Spread betting is almost always quoted with the cost built into the spread rather than broken out as commission. That is a pricing convention rather than a scandal, but it does make the two products hard to compare on cost because you are looking at one number instead of two. A CFD account quoting a raw spread plus a stated commission per lot is more transparent, whether or not it is cheaper. Whether spread betting carries a genuine cost premium is a question you have to answer broker by broker and instrument by instrument, and the only reliable way to do it is to fund both and compare what you were actually charged. Somebody has to pay for two deposits before that comparison exists at all, which narrows the field of who is in any position to publish one. That working method is what sits behind the comparison of spread betting brokers in the UK, put together by a UK investing research site that opens and funds live accounts with its own money to test trading platforms rather than ranking them from published fee schedules, and its spread betting pages set the FCA-regulated options side by side on what they charged in practice rather than on what they advertise.
What about the professional-trader question?
There is a persistent worry that trading full time turns spread betting profits into taxable trading income. In practice HMRC’s position has been that betting is not a trade and the exemption is not lost simply through frequency or scale. But this is genuinely an area where facts and circumstances matter, and if spread betting is your main source of income you should be paying an accountant rather than reading articles. The same applies in reverse to CFDs. Whether your activity is investment, generating capital gains, or a trade, generating income, is not something you elect. It follows from what you actually do. Most retail activity is comfortably the former, but nobody should treat a general article as a ruling on their own position.
Which should you actually choose?
If you are consistently profitable well beyond GBP 3,000 a year and have no other gains to shelter, spread betting is the better wrapper and the choice is easy. If you are starting out, trading small, or realistic about a loss rate running above 70% at the median firm in that sample, the CFD’s loss relief is worth more to you than an exemption on profits you may not make.
And if you hold other investments, work out where your annual allowance is going before you decide anything. That single fact will move the answer more than any argument about the products themselves.
Ten minutes with last year’s broker statement
The useful ten minutes here is not spent reading comparisons. Open the current HMRC rates alongside last year’s broker statement, take your own realised gains and losses, and run them down both columns of the table above. Most people who do that discover they have been choosing a wrapper for a benefit that lands several thousand pounds above where their trading actually finished, which is a more useful thing to learn from your own numbers than from anybody else’s.
…
