We have a self-assessment return open on the desk. Gulf-resident trader, British passport, FCA-regulated brokerage account. The 2024/25 tax year shows 247 closed positions across GBP/USD and XAU/USD. Net result: positive. The line that matters is not the profit figure. It is the page number. Capital gains go on supplementary page CG1. Trading income goes on supplementary page SE1. Those two pages route the same profit through two entirely different tax regimes, and the gap between them is not marginal. The trader ticked capital gains. We are not confident HMRC will agree.
The reason we are not confident requires examining what HMRC actually evaluates when a return arrives with forex profits attached. It does not look at the trader's LinkedIn bio. It does not care whether the brokerage account is labelled "personal" or "professional." What it examines — and what most self-assessment guides for forex traders conspicuously fail to explain — is a set of behavioural indicators that predate online trading by decades.
This matters disproportionately for Gulf-based British nationals. The desk sees this configuration routinely: UK passport, Dubai or Abu Dhabi residency, FCA-authorised broker, and a self-assessment obligation that survives the relocation because UK-source income or gains remain in scope. The classification question follows the passport holder, not the postal code.
HMRC Does Not Care What You Call Yourself on the Return
The distinction between capital gains treatment and trading income treatment hinges on what HMRC calls the "badges of trade." These are not codified in a single statute. They emerge from case law stretching back to the 1950s and are summarised across HMRC's own internal manuals — and here is where the desk found the friction. The Business Income Manual discusses trading activity through the lens of frequency, organisation, and profit motive. The Capital Gains Manual discusses disposal of assets with guidance that treats each closed position as a discrete chargeable event. Both manuals are current. Both are published by the same department. They pull in opposite directions for anyone with a triple-digit trade count in a single tax year.
The badges that matter for forex are specific. Frequency of transactions. The existence of a profit-seeking motive — which HMRC will assume in the affirmative for anyone trading leveraged instruments, because nobody opens an Exness FCA-regulated account to hold GBP/USD for long-term appreciation. Whether the activity is organised and systematic: tracked in spreadsheets, following documented rules, employing stop-losses and take-profits. And the holding period. A position held for forty-three minutes is not a capital asset. It is inventory.
The trader on our desk held positions for an average of 2.7 hours. That duration, combined with 247 closed positions across eleven months, satisfies at least three of the six recognised badges. HMRC does not require all six. Three is historically sufficient for a classification challenge.
What most UK forex content gets wrong is the baseline assumption. Articles routinely state that "most retail forex traders are classified under CGT." That claim is a relic of an era when retail meant placing a handful of spread bets per month. Spread betting occupies an entirely separate regime — spread bet profits in the UK are currently exempt from both CGT and income tax, which is precisely why so many UK-resident traders gravitate toward it. But our trader is not spread betting. The account is a standard FCA-regulated CFD account. Every position is a contract for difference, and CFDs do not enjoy the spread betting carve-out.
The result: 247 CFD trades through an FCA-regulated broker, held for hours not months, executed with leverage, tracked with precision. The self-assessment says capital gains. The badges say trading income. HMRC's two manuals offer contradictory frameworks for the same activity, and the trader is the one who bears the risk of guessing wrong.
The Annual Exemption Arithmetic Collapses Past a Few Hundred Trades
The appeal of CGT classification is obvious at small scale. The annual exempt amount means the first tranche of gains incurs zero tax. Basic rate taxpayers pay twenty percent on gains above that threshold. Higher rate taxpayers pay the same twenty percent on non-property assets — forex gains attract the lower CGT band, not the residential property surcharge. Compare that to trading income: no annual exemption, profits stacked directly onto the income tax computation, higher rate payers facing forty percent on the marginal pound.
On a modest profit, CGT wins. Unambiguously. The exemption absorbs a meaningful portion of the gain, and the residual rate sits below the income tax equivalent.
But the arithmetic inverts past a certain volume. Trading income classification offers a structural advantage that CGT cannot replicate: full deduction of expenses against gross profits. Platform subscriptions. Market data feeds. The cost of the VPS running an automated strategy. Under CGT, allowable deductions are narrow — essentially limited to acquisition and disposal costs directly attributable to the transaction. Under trading income, the deduction landscape opens wide, and for a trader with genuine infrastructure overhead, the effective rate after deductions can undercut the headline CGT rate despite the higher nominal percentage.
LBMA AM fix, 10 April 2026: $2,384.50. We reference this because our trader's portfolio is not pure forex. Thirty-one of those 247 positions were XAU/USD CFDs. Gold complicates the classification further than most guides acknowledge. HMRC has historically treated physical gold as a chargeable asset for CGT, but XAU/USD traded as a leveraged CFD is not physical gold. It is a derivative contract referencing the spot price. The LBMA fix anchors the price discovery mechanism, but the trader never touched a bar, never took delivery, never held an allocation at a vault. The instrument's nature — derivative, leveraged, closed within hours — pushes the gold positions toward trading income classification with the same force as the forex positions. A trader cannot credibly argue that GBP/USD trades constitute capital asset disposals while simultaneously filing XAU/USD CFDs under the same treatment. Both instruments were executed through the same Exness account, with the same strategy parameters, over the same holding durations. The classification must be consistent across the book, and the book reads like a trade.
The desk's view is blunt. Once trade count exceeds roughly 150 positions per year through a leveraged CFD account, the CGT classification becomes progressively harder to defend. Not impossible. Precedent exists for high-frequency investors receiving CGT treatment. But the burden of justification shifts materially, and HMRC does not need to prove the trader is carrying on a trade. It needs to demonstrate, on the balance of probabilities, that the activity has the character of one. Two hundred forty-seven positions in eleven months makes that demonstration straightforward.
FCA Brokers File Reports That Make the Classification for You
This is the dimension most tax commentary omits entirely. The trader's classification is not solely a matter of self-assessment arithmetic and badge-counting. FCA-regulated brokers have reporting obligations that construct an independent data trail, and that trail constrains the trader's filing choices whether the trader realises it or not.
Exness holds FCA authorisation. HF Markets holds both FCA and DFSA licences. Any broker operating under FCA regulation is subject to transaction reporting requirements under the UK's post-Brexit incarnation of MiFIR. Every CFD trade executed through an FCA-regulated account generates a transaction report filed with the FCA. These reports capture the instrument, the direction, the notional size, the execution timestamp, and a client identifier.
HMRC does not receive MiFIR transaction reports directly from the FCA as a matter of course. But it operates the Common Reporting Standard framework and the successor arrangements to EU information exchange. More critically, HMRC can and does request data from regulated firms when a self-assessment return triggers an enquiry. The 247-trade figure reported on supplementary page CG1 is not a number HMRC takes on trust. It is a number HMRC can verify — down to the timestamp and lot size — against the broker's regulatory filings.
Pattern is what triggers scrutiny. A return claiming CGT treatment for 247 leveraged CFD positions, average hold time under three hours, executed across both forex and commodity derivatives, is a return that reads differently from a genuine investor disposing of a handful of shareholdings. HMRC's enquiry selection is not random. It is risk-weighted. And the pattern visible in FCA-reported transaction data — short duration, high frequency, leveraged instruments, systematic execution — is indistinguishable from what an inspector would expect to see on a trading income return.
For Gulf-resident traders carrying UK tax obligations — and the desk encounters this configuration routinely, given the concentration of British passport holders across Dubai, Abu Dhabi, and Doha — an additional layer of complexity arises from residence status. A trader who qualifies as UK non-resident under the statutory residence test may fall outside HMRC's reach for trading income entirely, while remaining within scope for certain CGT disposals depending on the asset class and the period of non-residence. The interaction between the residence test and the trading classification demands its own analysis. But the FCA broker's reporting obligation does not condition on the trader's residence. The data trail accumulates regardless of whether HMRC ultimately has jurisdiction to tax it.
The practical implication for anyone reading this from a Gulf jurisdiction: if you hold a British passport, trade through an FCA-authorised broker, and submit a UK self-assessment — even a partial one reflecting UK-source obligations — the classification question is not a styling choice. The broker has already assembled half of HMRC's evidence base through routine regulatory compliance. Your return needs to be consistent with what that evidence shows.
This started as a comparison between two supplementary pages on a self-assessment form and turned into something more structural: the recognition that FCA broker reporting, HMRC's badges-of-trade framework, and the trader's own transaction data converge toward a classification that the trader may not have selected but cannot easily contest. The real question is not CGT or trading income in the abstract. It is what the data says about you, and whether your filing matches that portrait. Three signals worth tracking from here: first, whether HMRC updates the Business Income Manual to address high-frequency retail CFD trading explicitly, because the current guidance predates the volume of retail activity now flowing through FCA-regulated platforms. Second, whether the FCA's post-Brexit transaction reporting architecture expands the data surface accessible to HMRC without a formal information request. Third, whether the annual CGT exempt amount continues its recent downward trajectory — each reduction narrows the window in which capital gains classification offers any material tax advantage at all. The fork in the form is not closing. The distance between the two paths is growing.