Four UK brokers, four profit engines: what regulated accounts reveal behind the brands

A broker’s brand can look like the business. Its statutory accounts often show something more complicated. FXBrokers.ai examined the latest filed accounts of four UK-regulated companies—Trading 212 UK Limited, Pepperstone Limited, Capital Com (UK) Limited and OANDA Europe Limited—and reconstructed how each entity reached its reported result. The exercise found four materially different profit architectures beneath a common retail-broker label. One company produced a large local operating surplus. Another received an intra-group profit allocation that was almost nine-tenths of pre-tax profit. A third had moved much of its workforce and activity into a separate group service company. The fourth recorded a statutory loss during a change in ownership.[1][2][3][4]

This is not a ranking of the four brands. Their reporting periods, products, group structures and accounting policies differ. Pepperstone’s latest period ended on 30 June 2025, while the other three accounts examined cover calendar 2025. Nor does the legal company named on a UK customer agreement necessarily contain every employee, hedge, technology asset or economic exposure associated with the global brand. That limitation is the central finding rather than a reason to abandon the comparison: a headline such as “UK broker profit” can refer to fundamentally different things.[1][2][3][4]

The legal-entity perimeter is the investigation

Each company in this comparison is the authorised UK counterparty to at least part of its brand’s customer business, but none should be presumed to equal the entire brand. A regulated subsidiary can receive technology, hedging, marketing, treasury or staff services from affiliates. It can also provide services to them. Revenue may therefore arise from customers, group companies or both. The statutory accounts identify the reporting boundary; they do not automatically trace the full economic chain outside it.[1][2][3][4]

FXBrokers.ai treated the boundary as evidence rather than inconvenience. For each company, the analysis asked four questions: where was trading or customer revenue recognised; which risks remained in the entity; which material income or cost came from affiliates; and what operating resources remained locally? The answers produced four incompatible models. Combining them into one margin ranking would remove the most informative differences in the records.[1][2][3][4]

This approach also prevents a common entity error. A global website may use one brand while UK customers contract with a named company and group earnings are reported by a parent in another jurisdiction. Enforcement, insolvency, complaints and client-money questions begin with the contracting entity. Investor analysis often begins with the consolidated parent. Investigative reporting has to connect the two without substituting one for the other.[1][2][3][4]

Trading 212: a large local profit pool

Trading 212 UK Limited reported turnover of £277.6 million for 2025, up from £161.7 million in 2024. Pre-tax profit rose from £52.9 million to £123.1 million, while profit after tax reached £92.2 million. FXBrokers.ai calculates that turnover increased 71.7% and pre-tax profit 132.6%. The pre-tax margin consequently widened from 32.7% to 44.4%. Those figures describe the regulated UK company itself, not a group-wide allocation presented as if it were local operating revenue.[1]

The balance sheet reinforces the scale of the regulated entity. Cash stood at £141.9 million and net assets at £194.3 million at the year end. Average employee numbers rose from 53 to 122, an increase of about 130%. A business can increase profit faster than staff for several legitimate reasons, including operating leverage, higher interest income, greater trading activity and services supplied by other group entities. The accounts nevertheless show a company whose local revenue, workforce and capital base all expanded substantially in the same period.[1]

That matters when comparing Trading 212 with a broker whose UK subsidiary is primarily a distribution, risk-transfer or service node. A 44% pre-tax margin at one legal entity cannot automatically be applied to the wider group, but it is still an economically meaningful result: the entity carrying the UK authorisation generated more than £123 million before tax and retained a substantial net-asset base. The figure is not merely the profit of a marketing shell with negligible operations.[1]

Pepperstone: the affiliate allocation dominates

Pepperstone Limited’s accounts tell a different story. The company reported £15.0 million of trading revenue and £22.9 million of other income for the year to 30 June 2025. Pre-tax profit was £24.1 million. The accounts identify a £21.4 million profit-repatriation allocation from a group entity within other income. FXBrokers.ai calculates that this allocation alone was equivalent to 88.9% of the UK company’s pre-tax profit; all other income was equivalent to 95.1% of pre-tax profit.[2]

The filing also explains the regulated company’s risk arrangement. Pepperstone Limited’s permissions do not allow it to retain market risk in the way a dealing business might; market risk is borne by an affiliated group company under the disclosed structure. This does not make the UK company artificial or its profit invalid. It means its statutory result reflects an intra-group mechanism designed around licensing, risk transfer and group remuneration. The economic question is therefore not simply how much UK customers traded, but how the group allocated the resulting economics back to the regulated subsidiary.[2]

The same accounts reported £26.2 million of segregated client cash, broadly stable against £27.0 million a year earlier, and an £8.5 million dividend. Client money is not company revenue and should not be compared with profit. Its inclusion here serves a different purpose: it shows that a regulated customer-facing business can hold a substantial safeguarding responsibility while its reported earnings are driven mainly by an affiliate allocation. Operational custody, risk ownership and profit recognition may sit in different parts of the group.[2]

Capital.com: the result after a service-company migration

Capital Com (UK) Limited’s revenue fell from £40.9 million in 2024 to £18.8 million in 2025, a decline of 53.9%. Profit after tax fell from £4.45 million to £1.49 million, or 66.5%. Average employee numbers dropped from 116 to 30, a 74.1% reduction. Read without the notes, those changes might look like a sudden collapse in the brand’s UK customer franchise. The filing instead links the movement to a group reorganisation in which staff and activities were moved to a separate service entity.[3]

That distinction is material. When a group centralises technology, compliance, customer support or other functions, the regulated broker may buy services rather than employ the people directly. Revenue and expenses can also move as commercial agreements are rewritten. The legal entity’s lower headcount therefore cannot be treated as a direct count of people serving UK customers, and its revenue decline cannot be converted mechanically into a decline in customer trading. What changed was not only business volume, but the boundary of the reporting company.[3]

Capital Com (UK) Limited ended 2025 with net assets of £9.3 million, down from £11.8 million. Its continued profitability shows that the company remained economically active after the migration, but the accounts no longer describe the same operating perimeter as the prior year. A trend chart that fails to mark the reorganisation would compare two different organisational designs and present the result as ordinary year-on-year performance.[3]

OANDA: a loss during a change of control

OANDA Europe Limited reported turnover of £14.9 million for 2025, down 19.3% from £18.5 million. The company moved from a pre-tax profit of £1.51 million to a pre-tax loss of £2.07 million—a negative swing of £3.58 million. The loss after tax was £1.59 million. It nevertheless ended the year with £10.9 million of net assets and £6.4 million of cash, so the filing does not by itself indicate insolvency or an inability to meet regulatory capital requirements.[4]

The period also covered the completion of FTMO’s acquisition of the OANDA group after regulatory approvals. A change of control can bring transaction costs, integration work, revised group charges and strategic changes, but the public accounts do not permit every pound of the loss to be assigned to the acquisition. The disciplined conclusion is narrower: the UK regulated company’s revenue contracted and it recorded a loss in a year when its ownership changed. Causation beyond the disclosures would require further evidence.[4]

Why the four profits are not interchangeable

Put side by side, the figures expose a category error common in broker analysis. Trading 212’s £123.1 million pre-tax profit arose inside a rapidly expanding UK legal entity with £277.6 million of turnover. Pepperstone’s £24.1 million result was heavily shaped by a £21.4 million group allocation and a disclosed transfer of market risk to an affiliate. Capital.com’s smaller profit followed a migration of staff and functions outside the regulated company. OANDA’s loss occurred in a contracting entity during a change of control. Calling all four numbers “UK broker earnings” without those qualifiers obscures more than it reveals.[1][2][3][4]

The accounting periods also matter. Pepperstone’s June year end means its numbers overlap only partly with the calendar-year results of the other companies. Turnover definitions can differ; one entity may recognise financing, spread income, commissions or group-service income in lines that another presents elsewhere. Tax rates, hedging arrangements and foreign-exchange translation add further differences. FXBrokers.ai therefore did not aggregate the four profits or calculate a league-table margin across them. Such a total would imply comparability that the filings do not support.[1][2][3][4]

Four plausible but wrong comparisons

The first wrong comparison would divide every company’s pre-tax profit by turnover and declare the highest margin the most efficient. Pepperstone’s group allocation and Capital.com’s service-company migration mean the numerator and denominator do not represent the same operating perimeter as Trading 212’s. OANDA’s transition-year loss adds acquisition and reorganisation uncertainty. The ratios are mathematically valid inside each set of accounts but not economically interchangeable across the set.[1][2][3][4]

The second would treat headcount as a productivity score. Trading 212’s employee growth and Capital.com’s decline reflect different organisational changes. Staff can move to a group service entity while continuing to support the same customer base; outsourced technology and shared functions can also reduce local employment. Revenue per employee would therefore reward or punish legal structuring unless group-service costs and personnel were restored to a common perimeter.[1][3]

The third would compare segregated client cash with revenue or capital. Client money is a safeguarded customer balance, not a sale and not ordinarily distributable equity. A company can have stable client money and rapidly changing profit, as Pepperstone’s figures illustrate. The fourth would read a dividend as cash extracted from current customers. Dividends are distributions from lawfully available reserves and can reflect prior-year accumulation, capital planning and group ownership; the accounts do not support a direct customer-harm inference.[2][3]

Why the finding has national significance

The four entities serve UK retail investors under the same broad regulatory system, yet their public accounts expose different dependencies on affiliates, staff structures and risk locations. That is nationally significant because customers, policymakers and even market commentators routinely discuss “UK brokers” as if the regulated company were a standardised operating unit. It is not. The same authorisation category can sit inside markedly different group designs.[1][2][3][4]

That variation does not require regulators to prohibit group structures. It does strengthen the case for entity-level operational-resilience and wind-down disclosure. A subsidiary that transfers market risk may depend on an affiliate’s continued performance; a subsidiary that buys most services from a group company may need substitutability plans; a locally profitable entity may still rely on shared technology. The accounts identify where further supervisory questions should begin without alleging that any disclosed model has failed.[1][2][3][4]

The customer and regulatory consequence

For customers, the legal-entity map determines which company owes the contractual obligations, which rules apply and where complaints or compensation questions begin. It does not necessarily identify the entity that owns the technology, carries market risk or earns the largest share of group profit. For regulators, that separation increases the importance of intra-group service agreements, risk-transfer arrangements, capital resources and operational-resilience dependencies. A well-capitalised regulated subsidiary can still depend heavily on affiliates; a low-headcount company can still be responsible for a large customer base.[1][2][3][4]

For investors and competitors, the accounts demonstrate why brand-level revenue estimates need a legal-entity reconciliation. A group can move staff, functions or risk without losing customers. It can also remit profit to a regulated subsidiary through a disclosed allocation. Those choices may be entirely legitimate, but they change what statutory growth and margin mean. The right first question is not “Which broker made the most?” It is “Which company made it, from which activity, under which agreement, and for which period?”[1][2][3][4]

Capital, dividends and audit opinions answer different questions

The retained capital and dividend figures also resist simple ranking. A large dividend can reflect accumulated surplus and group capital planning rather than current-year operating strength; a small dividend can reflect expansion or regulatory buffers rather than weakness. Net assets are measured inside each legal perimeter and are not the same as excess regulatory capital. None of the four accounts supplies a common measure of capital available above firm-specific requirements, so statutory equity should not be presented as a like-for-like safety score.[1][2][3][4]

An unmodified audit opinion provides assurance on whether the financial statements are prepared, in all material respects, under the applicable reporting framework. It does not certify that intra-group prices are commercially optimal, that every operational dependency is low-risk or that customers will never suffer a service failure. Conversely, the presence of a loss does not invalidate a clean audit. Audit status, profitability, regulatory capital and customer protection are related but distinct tests.[1][2][3][4]

A more useful public comparison would require each broker group to publish a legal-entity bridge: customer accounts and assets by contracting entity; local and group revenue; services bought from affiliates; risk transferred to affiliates; regulatory capital and required buffers; dividends; and the largest operational dependencies. That bridge would not eliminate legitimate group structures. It would make clear where the economics, obligations and resilience actually sit.[1][2][3][4]

What remains unresolved

The filings do not provide a complete, consistently segmented view of customer numbers, trading volumes, average balances and group-wide economics for all four brands. Nor do they disclose every transfer-pricing calculation or operational dependency. A definitive cross-brand profitability comparison would require consolidated figures built on common definitions and matched periods, plus an entity-by-entity map of revenue, expenses, hedging and capital. Public statutory accounts are strongest at showing what happened inside each reporting perimeter—not at creating a perimeter the companies did not report.[1][2][3][4]

The investigation’s central result is therefore structural. Four companies performing broadly similar customer-facing functions produced four different kinds of statutory result. Trading 212 demonstrates a large local profit engine; Pepperstone a group-remuneration model linked to risk transfer; Capital.com the accounting effects of moving operations into a service company; and OANDA a loss-making transition year. The records do not establish wrongdoing at any company. They establish that the regulated legal entity—not the brand name—is the necessary unit of serious broker reporting.[1][2][3][4]

Report an error or submit an update