NextFin News - Alex Gerko earned a record £895mn from his algorithmic trading firm XTX Markets in 2025, the largest payout in the company's decade-long history and an increase of roughly 31% on the £682mn he took home a year earlier, according to accounts filed with UK regulators. The windfall arrived as turbulent markets delivered XTX its most profitable year on record: combined revenue climbed 43% to £3.93bn and profit rose 33% to £1.71bn, up from £2.74bn and £1.28bn respectively in 2024.
The payout captures a central tension in modern market structure: while public companies and retail investors wrestled with valuation compression and policy uncertainty through 2025, the private quant firms standing on the other side of their trades were quietly harvesting the dislocations those same forces created. Gerko, 46, now consolidates his position as one of Britain's wealthiest people, with a fortune estimated at more than $17bn, and remains among the largest individual contributors to the UK tax base.
The Numbers: A Record Built on Dislocation
The 2025 accounts, filed at Companies House, show a firm firing on every cylinder. XTX Markets Technologies Limited, the group's intellectual property and services arm, drove the results: net revenue at the entity climbed to £3.02bn from £2.04bn, a 48% rise, while profit after tax reached £1.69bn. The two FCA-regulated trading entities told a different story. XTX Markets Trading Limited, which trades equity, fixed-income and commodity markets, saw net revenue rise to £849mn from £636mn, yet profit after tax slipped to £21mn from £23mn as variable costs scaled with revenue and administrative expenses climbed to £823mn. The original XTX Markets Limited entity, focused on equity and FX, posted net revenue of £62mn, roughly flat, with profit after tax falling to £791,000 from £9.2mn.
That divergence is the first important detail. The record was not made by market making alone. It was made by the technology company inside the trading firm — the entity that owns the models, charges the affiliates service fees, and employs 127 people at an average wage of roughly £457,000. Administrative expenses at that entity doubled to £766mn from £382mn, a 100% increase that tracks the build-out of the compute stack. XTX is not simply renting a market cycle; it is capitalizing the machine that trades it.
The cash distribution matches the profit surge. The technology entity paid £1.78bn in interim dividends to its Cayman Islands parent, XTX Holdings Limited, during 2025, up from £1.17bn the year before. After the reporting period closed, it paid a further £331mn in January 2026 and £261mn in March 2026 — bringing the total flow to the parent to roughly £2.37bn. Gerko controls the parent through his approximately 75% stake in the group. The £895mn figure, however, represents his reported earnings draw from the firm rather than a pro-rata slice of the dividend; it is the compensation pool payout, separate from the capital returned to shareholders.
The scale becomes clearer against XTX's own recent history. In 2024, a total of £1.28bn was distributed between Gerko and 30 quantitative traders — up 71% on the £747mn shared in 2023 — with the trader pool taking £597mn and Gerko the largest single share at £682mn. The number of profit-sharing traders rose from 25 to 30 over that period, a detail that matters: XTX's economics are not a one-person operation, even if one person captures the majority of the upside. The firm handles approximately $250bn of transactions daily across 35 countries, using machine-learning models to forecast price moves across more than 50,000 instruments — equities, fixed income, currencies, commodities and crypto — then trading on exchanges and alternative venues while offering liquidity directly to clients.
Why Turbulence Pays: The Mechanism Behind the Record
The transmission channel from market chaos to XTX's bottom line is direct and mechanical. Market makers earn the spread — the difference between the price at which they buy and the price at which they sell. When volatility rises, spreads widen: uncertainty commands a higher price, and the compensation for holding inventory risk increases. Proprietary trading desks, meanwhile, profit from the magnitude and frequency of price moves. A calm, range-bound market is the enemy of both businesses; a year of tariff shocks, shifting rate expectations and geopolitical flare-ups is the ideal hunting ground.
2025 delivered exactly that environment. The second year of a second Trump administration brought fresh tariff volatility; central-bank rate paths swung on every inflation print; and geopolitical tensions kept risk premiums elevated. The firm itself attributed the record to turbulent markets, and the accounts bear the label out: a 33% profit increase on a 43% revenue increase implies costs scaled with activity, but the absolute pound value of the dislocations grew faster than the expense of capturing them.
This is not the first time turbulence has filled XTX's coffers. In 2022, the year Russia invaded Ukraine and energy markets convulsed, XTX reported a 64% jump in profit to £1.1bn — the first time its UK profit crossed the £1bn threshold. The pattern is consistent: systemic stress, wider spreads and larger moves, higher market-making and proprietary revenue. The 2025 result, £1.71bn, is roughly 55% above that 2022 peak, a sign that the firm's capacity to monetize disorder has grown even as disorder has become a more familiar condition.
Here is the cyclical call, stated plainly: the £895mn payout is, at its core, cyclical. Volatility mean-reverts. The VIX does not stay elevated forever, and neither do the tariff headlines or the rate shocks that drove 2025's dislocations. A return to calm, range-bound markets would compress spreads and shrink the proprietary book's opportunity set. The payout is a function of a specific market regime, not a guaranteed annual escalator.
The Structural Leg: A €1bn Bet on Compute
To dismiss the record as pure volatility luck, however, would miss the structural shift underneath it. XTX has spent the past decade building a fixed-cost moat that turns temporary market dislocations into compounding advantage — and 2025 was the year that moat became visible in the accounts.
The firm's models run on roughly 25,000 AI chips, mostly sourced from Nvidia, fed by a geothermal-powered supercomputer in Iceland. Earlier in 2025, XTX committed €1bn to a data-center complex in Kajaani, Finland, where it owns a 478-acre site. The first 22.5-megawatt facility is scheduled for completion in 2026, with the wider project taking roughly five years. Finland's naturally low air temperature provides cooling, and the firm has explored channeling waste heat to the local community. Capital expenditure at the UK technology entity rose to £16.3mn in 2025 from £6.4mn in 2024, a 155% increase that is only the onshore down payment on the Finland build-out.
The logic is scale economics applied to prediction. Gerko summed up the strategy in an interview: the firm builds its own infrastructure ahead of demand, confident that applying more compute power to its machine-learning trading strategies will generate better returns.
"By building things ourselves, we can build ahead of our needs. Because of the way we use machine learning to build our trading strategies, we have been confident that we can apply more compute power to ultimately generate better returns."
This is the structural leg of the thesis: the compute arms race in quantitative trading is a winner-take-most dynamic. Once a firm crosses a threshold of data, chip count and research talent, the marginal improvement from each additional unit of compute compounds. XTX's €1bn data-center bet is not a bet on 2025's volatility; it is a bet that the firm will still be the lowest-cost, highest-precision predictor of price moves in 2030, regardless of whether volatility is high or low. In a calm market, the best-informed market maker still captures the spread — it just captures a thinner one, more times.
The talent side mirrors the compute side. The profit-sharing pool grew from 25 traders in 2023 to 30 in 2024, and the £1.28bn distributed that year — £597mn to the trader pool alone — is among the most aggressive compensation structures in global finance. That is the price of retaining the researchers who write the models that run on the chips. It is also a barrier to entry: a new quant shop cannot simply buy Nvidia GPUs and compete. It needs the people who know how to turn silicon into alpha, and those people are already paid.
The Adversarial Case: Why the Record May Not Repeat
The strongest argument against extrapolating the £895mn is the simplest one: mean reversion is the one reliable law in trading. If 2026 and 2027 bring a calm macro backdrop — inflation anchored, trade policy stable, no geopolitical shocks — XTX's revenue per unit of risk will fall. The firm's own history supports the caution. Between the 2022 spike and 2024, profit growth was positive but not parabolic; the step-change to £1.71bn required an unusually disorderly year. The doubling of administrative expenses at the technology entity, to £766mn, also raises the bar: higher fixed costs mean a calm market bites harder than it did when the cost base was smaller.
A second, underappreciated risk sits in the UK tax and regulatory ledger. Gerko lost a Court of Appeal case against HM Revenue & Customs in July 2024 over the taxation of a deferred payment plan from his GSA Capital years, with judges ruling the profits should face income tax rather than the lower corporation-tax treatment he had argued for. In mid-2025, he took a £22.5mn tax dispute to the UK Supreme Court, contesting how British authorities tax certain complex financial arrangements connected to the group. The outcome will not change 2025's headline, but it changes the after-tax reality of future payouts — and it feeds a wider debate about whether Britain remains hospitable to the ultra-wealthy who fund its tax base. Gerko has said he would support a wealth tax, a position that reads as pragmatic engagement with a political risk he cannot diversify away.
There is also a concentration argument. XTX is private, which means no public shareholders dilute the upside — but it also means the firm's fortunes are inseparable from one man's models, one man's capital allocation and one man's tax residency. Gerko renounced his Russian citizenship in 2022 after the invasion of Ukraine and is a British citizen living in north London, but the geopolitical optionality that made him wealthy also makes his base of operations a permanent subject of political scrutiny. The departure of co-chief executive Hans Buehler in mid-2025, to return to academia, leaves Gerko as the sole chief executive at a moment when the firm has never been larger.
The answer to the adversarial case is not that the payout will repeat — it may not. The answer is that the floor has risen. Even in a mean-reverting volatility regime, XTX enters the next cycle with more compute, more data and a deeper researcher bench than it had in 2022. A thinner spread captured with a wider edge is still a profitable trade. The cyclical leg says the £895mn is a peak; the structural leg says the next trough will be higher than the last one.
What Comes Next: Scenarios and Signals
Short term (6–12 months): sentiment and volatility. If macro volatility stays elevated — sticky inflation prints, renewed trade friction or a geopolitical shock — XTX's 2026 revenue should track at or above 2025 levels, and the payout pool will follow. The signal to watch is the firm's own dividend flow to XTX Holdings; a figure near or above the £2.37bn total would confirm the regime persists.
Medium term (1–3 years): the compute payoff. The Kajaani data center comes online in phases from 2026. If the €1bn bet translates into forecast-accuracy gains that outpace the cost of capital, XTX can grow profit even in a calmer market — the structural thesis validated. If compute spend outruns alpha generation, margins compress and the moat narrative weakens. The £766mn administrative expense line is the number to hold against the revenue line.
Long term (3+ years): structure over cycle. The secular shift toward electronic, AI-driven market making is intact. Human market makers have ceded share to algorithms for two decades; the question is how much of the remaining discretionary flow gets automated. XTX's bet is that the answer is most of it, and that the firm with the most compute captures the largest slice.
The falsifying signal is quantifiable: if XTX's combined profit for 2026 falls below £1.3bn — the 2024 level — while the VIX averages below 15 and global trade-policy uncertainty declines, the structural-moat thesis takes a hit. It would mean the 2025 record was regime, not capability, and that the €1bn compute bet has not yet bought durable advantage.
For the broader market, the implications split by side of the trade. Liquidity seekers — asset managers executing large blocks, corporates hedging FX exposure — benefit from XTX's willingness to commit capital in stressed markets, but they pay for it in wider spreads. Competing market makers face a higher bar: match the compute spend or lose flow. And for Britain's tax base, Gerko remains one of the largest individual contributors, a fact that makes his Supreme Court case a proxy for a larger question about how a post-Brexit financial center treats the people who fund it.
The record £895mn is not a story about one trader's luck. It is a story about what happens when a private firm with no public shareholders, an unlimited time horizon and a €1bn compute moat stands on the other side of a turbulent year — and collects the premium that uncertainty demands. Volatility will fade. The question is whether the machine built to harvest it has become too efficient to beat when it does.
Explore more exclusive insights at nextfin.ai.
