XTX Markets posted £1.28 billion net profit in 2024 — up 54% from £835 million in 2023. Revenue across three UK entities reached £2.74 billion (+37% YoY). The London algorithmic trader employs ~250 people globally, with 113 staff at its largest UK entity (XTX Technologies) earning £435,814 average compensation. The firm’s performance demonstrates how computational infrastructure converts market turbulence into systematic profits.
August 2024: The Carry Trade Unwind
Bank of Japan raised rates to 0.25% on July 31. US employment data missed expectations August 2 (114,000 jobs vs. 175,000 forecast). Combined catalyst: forced liquidation of yen carry positions.
BIS estimates ~$160 billion in hedge fund FX forward positions unwound; cross-border bank claims potentially exceeded $500 billion. USD/JPY appreciated 6–7% during the critical late-July to early-August period. Nikkei 225 fell 12.4% on August 5 — worst single-day decline since 1987. S&P 500 dropped 3%. Correlated sell-offs hit Mexican peso, Brazilian real, South African rand as leveraged positions unwound.
Market maker profitability: P&L = (δ — α) × r × v × 0.5, where δ = bid-ask spread, α = execution costs, r = fill rate, v = volume. August delivered simultaneous expansion in δ and v — multiplicative, not additive.
Infrastructure: 25,000 GPUs, 650PB Storage, $250B Daily Volume
XTX reports a research cluster of ~25,000 GPUs paired with ~650 petabytes usable storage (company statement). Industry sources suggest the cluster includes significant deployments of Nvidia A100 and V100 GPUs, though XTX has not published a verified per-model breakdown. The firm processes >1 trillion data points daily across 50,000+ instruments. Daily volume: $250 billion spanning FX, equities, fixed income, commodities, crypto.
Critical differentiation: proprietary infrastructure eliminates cloud bottlenecks. Two Sigma reported 20% GPU availability pre-cloud migration during capacity shortages. During August volatility spikes, cloud capacity constraints risk model staleness — milliseconds of recalibration delay = adverse selection or missed fills.
XTX committed €1 billion to Finland data center (2026 operational, 22.5MW initial capacity). Investment preceded 2024 volatility surge — positioning, not reaction.
P&L Mechanics: Three Drivers
1. Optimal Spread Expansion
Avellaneda-Stoikov framework: optimal spreads scale with √(σ²γ/2kA), where σ = volatility, γ = risk aversion, k = order arrival rate, A = fill intensity. August’s multi-asset volatility spike justified wider spreads as optimal pricing for inventory risk, not market dysfunction.
2. Volume Surge
FX markets remained well-functioning during the unwind. Market makers executed high fill rates during forced liquidations. Volume multiplication across correlated assets: USD/JPY, EUR/JPY, emerging market currencies, equity indices.
3. Cross-Asset Capture
XTX’s multi-asset capability captured spreads across synchronized moves. Carry trade unwind hit yen pairs, EM currencies, equity indices, volatility products — simultaneous across 50,000+ instruments requiring real-time recalibration.
Comparative Performance: The Market Maker Boom
XTX Markets: £2.74B revenue, £1.28B profit (~250 employees globally)
Citadel Securities: $9.7B revenue (+55% YoY), $5.2B EBITDA (+87%)
Jane Street: $20.5B revenue (+94% YoY), $13B net income
Revenue per employee (using global headcount): XTX ~£11M, Jane Street ~$7.6M. XTX’s UK entity (113 staff) generated £2.04B revenue, yielding £18M per UK employee — but this excludes global operations and should not be extrapolated to firm-wide metrics.
Non-bank market makers captured substantial market share from traditional banks. Regulatory constraints and capital requirements disadvantage banks. Market makers operate with lean structures, proprietary capital, algorithmic execution — advantages amplified during volatility.
Quantitative Edge: Computational vs. Human Capital
Market-making models require microsecond-latency predictions across thousands of instruments. XTX’s ML systems extend Stoikov framework: optimize reference prices considering execution probability, price impact, inventory risk, cross-asset dependencies, real-time order flow imbalances. These calculations require massive compute unavailable via shared cloud infrastructure.
The architectural choice matters critically. Proprietary data centers eliminate capacity constraints catastrophic during volatility spikes. August carry trade unwind + cloud bottlenecks = inventory losses or missed opportunity. Industry commentators note cloud capacity constraints can matter during spikes; XTX’s privately owned infrastructure reduces this risk.
Key Insight: Nonlinear Returns to Volatility Preparation
Market-making profitability exhibits convexity in volatility. Spreads widen with vol; volumes increase simultaneously; outcome = multiplicative profit expansion. Firms with real-time ML forecasting quote competitive spreads even in volatile markets — capturing flow while managing inventory algorithmically.
XTX’s infrastructure investment during lower-volatility 2023 positioned the firm for outsized returns when conditions reversed in 2024. The firm’s £1.28B profit from ~250 employees reflects structural reality: market-making alpha increasingly derives from computational infrastructure vs. human expertise alone.
Actionable Takeaway
For quant-driven market makers: Infrastructure investment creates option-like payoff during volatility. Downside: fixed costs during low-vol periods. Upside: nonlinear profit expansion when vol returns. XTX’s 2023–2024 trajectory demonstrates this payoff structure empirically.
For portfolio managers: Recognize non-bank market makers as systemically important liquidity providers. Understand their operating constraints (proprietary capital, algorithmic risk limits) during stress events.
Sources & Methodology
Financial Data:
Financial Times: “Alex Gerko earned £682mn from trading firm XTX in 2024”
Finance Magnates: “XTX Markets Posts 50% Profit Jump to £1.28 Billion”
Global Trading: “XTX Markets’ profits skyrocket in 2024”
Companies House filings (UK regulatory submissions)
August 2024 Crisis:
Bank for International Settlements: Bulletin №90 — “The market turbulence and carry trade unwind of August 2024”
Bank of Japan: Policy announcement July 31, 2024
Reuters: BoJ rate decision coverage
CNBC/CNN: US employment data August 2, 2024
MarketWatch: Nikkei 225 plunge coverage
Kyodo News: August 5 market crash
Infrastructure:
XTX Markets: Official company website
Lex (Substack): “AI supercomputers powering XTX Markets and DeepSeek’s trading empires”
A-Team Insight: “XTX Markets Commits €1 Billion to Finnish Data Centre Complex”
Bloomberg: €1 billion data hub announcement
The Trade News: Finland data center coverage
Competitor Data:
Bloomberg: “Citadel Securities’ $9.7 Billion Trading Revenue Passes Barclays”
Bloomberg: “Jane Street’s $20.5 Billion Trading Haul Tops Citigroup, BofA”
Academic Framework:
Avellaneda, M. & Stoikov, S. (2008): “High-frequency trading in a limit order book”, Quantitative Finance, 8(3), 217–224
Key Data Corrections:
USD/JPY movement: 6–7% appreciation during late-July to early-August period (verified via SIFMA, Acuity KP, FSG Journal)
Revenue per employee: UK entity (113 staff) vs. global headcount (~250) clarified per FN London
All core financial figures triple-verified through regulatory filings, company statements, and independent financial press. Market microstructure claims grounded in peer-reviewed academic literature.
Methodology Note: This analysis synthesizes regulatory filings, central bank publications, and verified industry intelligence. Figures represent best available public data; private trading firms selectively disclose financial information. Where claims could not be independently verified from authoritative sources, they have been removed or appropriately qualified.
Cover photograph: Acabashi, CC BY-SA 4.0, via Wikimedia Commons.