XTB held about 87% of disclosed client assets in shares and ETFs while CFDs supplied 90% of Q3 instrument result

XTB’s transformation into a broader investment platform is real on the asset side but not yet on the earnings side. The company said clients held EUR 12.7 billion of assets at the end of September 2026, including more than EUR 5.8 billion in shares and almost EUR 5.3 billion in exchange-traded funds. Together, those two categories represented about 87.4% of the disclosed total using the published amounts. Yet XTB’s preliminary financial statement shows that contracts for difference generated PLN 544.1 million, or 90.5%, of the quarter’s PLN 601.3 million gross result from financial instruments. The remaining instruments generated PLN 57.3 million, or 9.5%. The app may increasingly store long-term investments, while the income statement remains overwhelmingly driven by leveraged trading.[1][2]

That gap does not make the company’s multi-asset positioning false. Customer assets and broker revenue measure different things. A share or ETF can remain in custody for years and create little transaction income in a particular quarter, while a much smaller pool of leveraged positions can generate spreads, swap points, commissions and market-making results repeatedly. XTB itself explains that its operating revenue comes from spreads, swaps, fees, the net result of its market-making activity and interest on client cash. The relevant finding is therefore a business-model dependency: the assets attracting and retaining customers are not yet the instruments producing most of the reported gross result.[1]

The divide is especially important because XTB says more than 84% of new customers began with shares, ETFs or an Investment Plan, and that stocks and ETFs together represented more than 86% of assets in custody. Those figures describe the entry route and accumulated asset base. They do not show the share of operating income attributable to those customers, how many later trade CFDs, or whether investment products are profitable after custody, execution, data, payment and marketing costs. A reader cannot infer that an 87% asset share should produce an 87% revenue share; but neither should a platform’s asset mix be used to suggest that CFD economics have become secondary when the financial table shows the opposite.[1][2]

The published chart and accounting table do not use the same categories

A second reconciliation problem sits inside XTB’s own result presentation. The accounting table divides the gross result into PLN 544.1 million from CFDs and PLN 57.3 million from other instruments. A nearby chart allocates 70.8% to commodity CFDs, 12.0% to index CFDs, 4.9% to currency CFDs and 12.3% to an “Other” category. Applying 12.3% to the PLN 601.3 million total gives about PLN 74.0 million—roughly PLN 16.7 million more than the table’s other-instrument line. The report says the chart’s “Other” category mainly comprises stocks and ETFs. The most plausible reconciliation is that it also includes equity- and ETF-linked CFDs, which remain CFDs in the accounting table. That is an inference from the definitions, not an undisclosed accounting error.[1]

The distinction matters because “other” can sound like the non-CFD business. XTB’s KPI footnotes explicitly exclude share and ETF CFDs from the reported CFD lot and nominal-volume measures, while the gross-result table classifies all CFDs together. The chart groups revenue by underlying economic exposure rather than legal instrument type. A reader using the chart alone could therefore overstate the contribution from cash shares and ETFs. The report does not provide a direct split between cash instruments, fractional rights and equity-linked CFDs, so the precise amount earned from unleveraged investing cannot be independently reconstructed from the preliminary release.[1]

Lower volume and sharply higher profitability per lot drove the quarter

The earnings surge was not produced by a corresponding surge in CFD activity. XTB reported 1.71 million CFD lots in the quarter, 18.3% fewer than a year earlier, while profitability per lot almost doubled from PLN 152 to PLN 301. Total operating income increased 65.9% to PLN 623.3 million and net profit rose from PLN 53.2 million to PLN 213.7 million. The result shows how sensitive the model is to the economic outcome on each unit of leveraged client activity. More clients and more assets provided scale, but the decisive quarterly variable was the amount XTB earned per lot rather than the number of lots traded.[1]

XTB describes its model as a combination of agency execution and market making. The company is the legal counterparty to client CFD transactions, although it says the majority of cryptocurrency CFD exposure is hedged with external partners and most share and ETF transactions, including CFDs on those assets, are executed on regulated markets or multilateral trading facilities. The company also states that it does not conduct proprietary trading in anticipation of market moves. Those explanations are relevant alternatives to the simplistic claim that high profitability per lot necessarily represents customers’ losses kept by an unhedged dealer. The public data still do not show the precise contribution from spreads, swaps, client P&L, hedging and execution for the quarter.[1]

What would prove that the transformation has reached the income statement

The evidence establishes three things. First, shares and ETFs dominate the disclosed client-asset base. Second, CFDs still dominate the gross result from financial instruments. Third, the quarter’s profit expansion came with lower CFD lot volume and much higher profitability per lot. None of those points establishes customer harm, mis-selling or an unsustainable model. XTB’s large and growing investment asset base may strengthen retention, cash balances and future fee opportunities even before it contributes heavily to transaction income. The company also recorded PLN 28.4 million of net interest income on client funds and PLN 6.3 million of fee and commission income outside the instrument result.[1][2]

A complete test of the multi-asset thesis would require a revenue and cost bridge for cash shares, ETFs, Investment Plans and fractional rights, separated from equity and ETF CFDs. It would also need customer-cohort data showing whether investors who enter through long-term products subsequently trade leveraged instruments, and whether the economics differ by jurisdiction. Until those data are published, the defensible conclusion is narrower than either the promotional or critical extreme: XTB has changed what customers hold, but the preliminary third-quarter accounts show that it has not yet changed what principally drives the gross result.[1][2]

There is an additional unit problem in the activity data. A CFD lot is not one economic unit across products: XTB notes that lot size varies and that currency and crypto CFDs use a 100,000-unit convention, while equity and ETF CFDs are excluded from the KPI because one lot equals one share. Nominal value and lot count therefore answer different questions and neither can be compared directly with the number or value of cash-share transactions. The rise in profitability per lot is meaningful within XTB’s recast series, but it is not a universal take rate that can be applied to all client activity.[1]

The change in profitability per lot supplies a useful counterfactual, although not a forecast. Applying the prior-year PLN 152 profitability measure to the reported 1.71 million Q3 lots would produce roughly PLN 260 million, around PLN 255 million below the result implied by PLN 301 per lot. The gap is about 42% of the quarter’s PLN 601.3 million gross result from financial instruments. Product mix, volatility and hedging mean this arithmetic cannot identify customer losses, but it quantifies how much the quarter depended on economics per unit rather than volume growth.[1][2]

The business-model description also narrows what the quarterly result proves about customer outcomes. XTB can earn from spread, swap, fee and market-making components while hedging some exposures and routing most share-related transactions externally. A high gross result can coincide with efficient execution, adverse client positioning, favourable hedging, or some combination. The preliminary report does not disclose client aggregate gains and losses, effective spreads or hedge P&L by asset class. Those missing data prevent the investigation from turning a documented earnings dependency into a claim that customers as a group lost the same amount.[1]

Report an error or submit an update