What Is Jump Trading?
Jump Trading is a privately held proprietary trading firm headquartered in Chicago, Illinois. Founded in 1999 by Bill DiSomma and Paul Gurinas - both former floor traders at the Chicago Mercantile Exchange (CME) - the firm has grown into one of the largest and most secretive high-frequency trading operations in the world.
The name "Jump" reportedly comes from the trading pit hand signal for spread trades at the CME. DiSomma and Gurinas started by trading futures on the exchange floor, and the firm's early success came from applying quantitative methods to markets that were still transitioning from open-outcry to electronic trading. That transition gave Jump a structural advantage: they understood the mechanics of the products from floor trading, and they had the technical vision to build systems that could trade those products faster electronically.
Today, Jump Trading operates from offices in Chicago, New York, London, Singapore, Shanghai, Bristol, and Sydney. The firm employs roughly 1,500 people globally, spanning quantitative research, software engineering, FPGA development, and trading. Unlike publicly traded competitors such as Virtu Financial or Flow Traders, Jump discloses very little about its operations, revenue, or strategies. This secrecy is deliberate - in high-frequency trading, any information about your edge can erode that edge.
Jump sits alongside firms like Citadel Securities, Jane Street, Tower Research Capital, and DRW Trading as one of the most significant proprietary trading firms globally. If you're exploring the wider prop trading space, our guide to prop trading firms covers the full range of top firms and how they compare.
What Does Jump Trading Do?
Jump Trading makes money by trading financial instruments using quantitative strategies and purpose-built technology. The firm trades across multiple asset classes - futures, equities, options, fixed income, commodities, and digital assets - on exchanges around the world.
Their core activities fall into several categories:
High-Frequency Trading (HFT)
Jump is widely regarded as one of the top five HFT firms globally. High-frequency trading involves executing a large number of orders at extremely high speeds - often in microseconds - to capture small price inefficiencies. The profit per trade is tiny, but the volume is enormous.
Jump's HFT strategies rely on ultra-low latency infrastructure (more on this below), co-located servers at major exchanges, and custom hardware including FPGA-based trading systems. The firm competes directly with firms like Citadel Securities, Virtu Financial, and Tower Research Capital for speed and execution quality.
Market Making
Jump acts as a market maker on multiple exchanges, continuously quoting bid and ask prices for various instruments. Market makers earn the bid-ask spread while managing inventory risk. This requires sophisticated pricing models, real-time risk management, and the ability to update quotes in microseconds as market conditions change.
Statistical Arbitrage
Beyond pure speed-based strategies, Jump runs statistical arbitrage programmes that identify pricing relationships between related instruments. These strategies are less dependent on raw latency and more reliant on research and modelling. They might exploit temporary mispricings between futures contracts and their underlying assets, between correlated equities, or between options and their theoretical values.
Futures and Options Trading
Jump has deep roots in futures trading - the firm was literally born on the CME floor. They're active across equity index futures, interest rate futures, commodity futures, and options on all of these. Chicago remains the global centre for futures trading, and Jump's headquarters location gives it proximity to the CME Group (which operates the CME, CBOT, NYMEX, and COMEX exchanges).
Crypto and Digital Assets
Jump expanded into cryptocurrency trading through its digital assets division, initially branded as Jump Crypto and later reorganised as Jump Digital. This arm traded crypto spot and derivatives, provided liquidity on decentralised exchanges, and invested in DeFi protocols and blockchain infrastructure projects. The crypto division brought Jump significant attention - not all of it positive - which is covered in detail below.
Jump Trading's Technology Stack
Jump Trading invests more heavily in trading technology than almost any other firm. Their infrastructure isn't an afterthought - it's the core competitive advantage. If you're interested in the technical foundations of low-latency systems, our guide to C++ in quantitative finance and hardware acceleration for quant articles cover the key concepts.
C++ for Core Trading Systems
Jump's trading engines are written primarily in C++. The firm uses modern C++ (C++17 and C++20 features) for its performance-critical execution paths, where every nanosecond of latency matters. Engineers at Jump are expected to understand memory allocation, cache behaviour, template metaprogramming, and lock-free data structures. Python is used for research and prototyping, but production trading code runs in C++.
FPGA-Based Trading
Jump was one of the early adopters of Field Programmable Gate Arrays (FPGAs) in trading. FPGAs allow the firm to implement trading logic directly in hardware, bypassing the overhead of a traditional CPU and operating system. This can reduce latency from microseconds to nanoseconds for specific operations like market data parsing and order generation.
The firm employs dedicated FPGA engineers who design and implement trading algorithms in hardware description languages such as VHDL and Verilog. This is a relatively niche skill set, and Jump competes with a small number of firms - including Citadel Securities, DRW, and IMC Trading - for this talent.
Microwave and Millimetre-Wave Networks
One of the most distinctive aspects of Jump's technology investment is its microwave tower network. In 2013, Jump (through a subsidiary called World Class Wireless) purchased a microwave tower in Richborough, Kent - a former NATO installation - to transmit market data between London and Frankfurt faster than fibre optic cables allow.
The physics is straightforward: microwave signals travel through air at close to the speed of light (~299,792 km/s), while light through fibre optic cable travels at roughly 200,000 km/s due to the refractive index of glass. Over the roughly 400 km between London and Frankfurt, this translates to a latency advantage of several hundred microseconds - an eternity in HFT. Jump has invested in similar microwave links between other financial centres, including Chicago and New York (via a corridor across New Jersey and Pennsylvania). For more on why these differences matter, see our network speeds and latency guide.
Co-location and Custom Networking
Jump co-locates servers directly in exchange data centres (such as the CME's Aurora data centre in Illinois, and the London Stock Exchange's Basildon facility) to minimise the physical distance between their trading systems and the exchange matching engine. Inside these data centres, the firm uses custom network interface cards, kernel-bypass networking (DPDK, Solarflare OpenOnload), and hand-tuned operating system configurations to shave microseconds from the order path.
Data Infrastructure
Jump processes enormous volumes of market data in real time - every quote, trade, and order book update across dozens of exchanges globally. This requires distributed systems capable of handling millions of messages per second with minimal jitter. The firm builds much of this infrastructure in-house rather than relying on third-party vendors.
Jump Crypto and Digital Assets
Jump's venture into cryptocurrency became one of the most high-profile chapters in the firm's history - and one of the most controversial.
The Formation of Jump Crypto
Jump Trading began trading cryptocurrency markets around 2015, initially applying the same quantitative market-making strategies it used in traditional markets. By 2021, the firm formalised this as a separate division called Jump Crypto, staffed with dedicated teams focused on blockchain trading, DeFi protocol development, and crypto venture investing.
Jump Crypto was among the most active institutional participants in decentralised finance. The team contributed to projects including Wormhole (a cross-chain bridge), Pyth Network (a DeFi oracle), and various Solana ecosystem initiatives. At its peak, Jump Crypto had invested in over 100 blockchain-related projects and was one of the largest liquidity providers on both centralised and decentralised crypto exchanges.
The Wormhole Bridge Incident
In February 2022, Wormhole - a bridge protocol that Jump Crypto had helped develop and maintain - was exploited in one of the largest DeFi hacks at that time. An attacker exploited a vulnerability to mint 120,000 wrapped Ethereum (wETH) on the Solana side of the bridge, worth approximately $320 million.
Jump Crypto stepped in and replenished the stolen funds to make Wormhole users whole. This was widely seen as an extraordinary move - few firms would absorb a $320 million loss to protect users of a DeFi protocol. Whether it was motivated by reputational concerns, strategic positioning, or genuine commitment to the ecosystem, the decision drew significant industry attention.
Regulatory Scrutiny and Restructuring
Jump Crypto faced increasing regulatory scrutiny in 2023 and 2024, particularly from the US Commodity Futures Trading Commission (CFTC) and Securities and Exchange Commission (SEC). Reports emerged of an investigation into Jump's crypto trading activities, and the firm's president, Kanav Kariya, resigned in June 2023.
Following this, Jump significantly scaled back its crypto operations. Jump Crypto was reorganised as Jump Digital with a reduced scope, and the firm redirected its focus towards traditional asset classes. As of 2026, Jump maintains some presence in digital asset markets but at a fraction of its previous scale.
Careers at Jump Trading
Jump Trading hires across several technical disciplines. The firm is smaller than some competitors - roughly 1,500 employees versus Citadel Securities' ~4,000 - but compensates this with some of the highest pay in the industry. If you're evaluating where to apply, our quant trader career guide covers how to think about role selection across the industry.
Quantitative Researcher
Quant researchers at Jump develop the mathematical models and statistical methods that drive trading strategies. This involves time series analysis, signal research, portfolio optimisation, and backtesting. Candidates typically hold PhDs in mathematics, statistics, physics, electrical engineering, or computer science.
The research process at Jump is closely integrated with production trading. Researchers don't just publish papers or run backtests in isolation - they work directly with engineers and traders to get strategies into production. Research tools are primarily Python and R, with production implementations in C++.
Algorithmic Developer / Software Engineer
Software engineers at Jump build and maintain the trading infrastructure: execution engines, market data systems, risk management platforms, and internal tools. The work is heavily C++ focused for anything on the trading path, with Python, Go, and Java used for supporting systems.
Jump's engineering culture emphasises performance. Engineers are expected to think carefully about memory layout, system calls, and hardware behaviour. This isn't typical web-scale engineering - it's more akin to embedded systems or game engine development in terms of the performance constraints.
FPGA Engineer
FPGA engineers design trading logic that runs directly on hardware. This is a specialist role requiring expertise in VHDL or Verilog, digital signal processing, and an understanding of financial markets. FPGA engineers at Jump work closely with quant researchers and C++ developers to implement the most latency-sensitive components of the trading stack.
This is one of the hardest roles to fill in quantitative finance. The intersection of FPGA expertise and financial domain knowledge is small, and Jump competes aggressively with firms like DRW, IMC, and Citadel Securities for these candidates.
Quantitative Trader
Quant traders at Jump sit between research and execution. They manage live trading strategies, monitor performance, adjust parameters, and make real-time decisions about risk and capital allocation. Traders need strong quantitative skills (probability, statistics, game theory) combined with the ability to act decisively under pressure.
Jump's trading teams operate across asset classes and geographies. A trader in the London office might focus on European equity index futures, while a counterpart in Chicago covers US interest rate products.
Data Scientist / Machine Learning Engineer
Jump employs data scientists who work on problems ranging from alternative data analysis to predictive modelling for market microstructure. Machine learning is applied selectively - not as a black box, but in contexts where the data supports it: pattern recognition in order flow, regime detection, and execution optimisation.
Jump Trading Salary and Compensation
Jump Trading offers compensation that competes with - and in some cases exceeds - other top-tier prop firms. The firm uses a base salary plus bonus structure, with bonuses heavily tied to firm performance and individual contribution. Compensation data for private firms is inherently imprecise, but the ranges below reflect 2026 estimates based on industry reports and candidate disclosures.
| Role | Experience | Base Salary (USD) | Total Comp (USD) | Total Comp (GBP, approx.) |
|---|---|---|---|---|
| Quantitative Researcher | Entry (PhD) | $150,000 - $200,000 | $300,000 - $500,000 | £240,000 - £400,000 |
| Quantitative Researcher | Senior (5+ yrs) | $200,000 - $300,000 | $500,000 - $1,500,000+ | £400,000 - £1,200,000+ |
| Software Engineer | Entry | $130,000 - $175,000 | $250,000 - $400,000 | £200,000 - £320,000 |
| Software Engineer | Senior (5+ yrs) | $180,000 - $250,000 | $400,000 - $800,000 | £320,000 - £640,000 |
| FPGA Engineer | Entry | $140,000 - $180,000 | $280,000 - $450,000 | £220,000 - £360,000 |
| FPGA Engineer | Senior (5+ yrs) | $200,000 - $280,000 | $500,000 - $1,000,000 | £400,000 - £800,000 |
| Quantitative Trader | Entry | $150,000 - $200,000 | $300,000 - $600,000 | £240,000 - £480,000 |
| Quantitative Trader | Senior (5+ yrs) | $200,000 - $300,000 | $600,000 - $2,000,000+ | £480,000 - £1,600,000+ |
| Data Scientist / ML Engineer | Entry | $130,000 - $170,000 | $250,000 - $400,000 | £200,000 - £320,000 |
GBP figures assume an exchange rate of approximately $1.25 to £1 and should be treated as rough estimates. London-based roles may have slightly different compensation structures due to UK tax and benefits considerations.
Bonuses at Jump can represent 50% to 80% of total compensation in strong years, making the variable component larger than at most hedge funds. The firm doesn't publicly disclose compensation data, so these figures are compiled from Glassdoor, Levels.fyi, and candidate-reported data across interview preparation communities.
The Jump Trading Interview Process
Jump's interview process is rigorous and heavily technical. The firm hires slowly and deliberately, and the acceptance rate for most roles sits well below 5%. Expect the full process to take four to eight weeks.
Stage 1: Application and CV Screen
Jump posts roles on its website and recruits at target universities including MIT, Stanford, Carnegie Mellon, Cambridge, Oxford, Imperial College London, and ETH Zurich. Referrals carry significant weight. Your CV should emphasise quantitative skills, competitive programming results, research publications, or relevant engineering experience.
Stage 2: Online Assessment
Most candidates receive a timed online test. For engineering roles, this typically involves 2 to 3 algorithmic coding problems (usually in C++ or Python) within 60 to 90 minutes. For quant research roles, expect probability questions, statistical reasoning, and mathematical problem-solving. FPGA roles may include hardware-specific questions on digital logic and timing analysis.
Stage 3: Technical Phone Screens (1-2 Rounds)
Phone screens last 45 to 60 minutes and are conducted by team members. Engineers face live coding (often in C++) with emphasis on data structures, algorithms, and systems thinking. Quant researchers get probability and statistics problems. Quant traders face rapid-fire mental arithmetic, probability puzzles, and market reasoning questions.
Stage 4: On-site or Virtual Final Round
The final round consists of four to six interviews over a full day. Each interviewer covers a different area:
- Technical depth: advanced C++ (for engineers), probability theory (for researchers), or market microstructure (for traders)
- Problem-solving: novel problems you haven't seen before, testing how you think rather than what you've memorised
- Systems thinking: how you'd design a trading system component, optimise a data pipeline, or debug a latency issue
- Cultural fit: Jump values intellectual curiosity, low ego, and a willingness to collaborate across disciplines
Offers typically arrive within one to two weeks after the final round. Jump is known for making competitive offers quickly once a decision is made.
How to Prepare for a Jump Trading Interview
Preparation for a Jump interview should start at least six to eight weeks before your target application date. The firm's bar is high, and the technical depth expected varies significantly by role. Our quant interview questions guide covers the foundational material you'll need.
For Software Engineering Roles
C++ proficiency is non-negotiable. Jump's engineering interviews go well beyond typical LeetCode problems. You'll need to understand move semantics, smart pointers, template metaprogramming, memory models, and low-level optimisation. Practise writing clean, performant C++ - not just correct C++.
Study data structures and algorithms thoroughly, with emphasis on problems that involve time complexity trade-offs and cache-friendly design. Be ready to discuss system design questions specific to trading: order book data structures, real-time market data processing, and low-latency messaging.
Recommended resources: Effective Modern C++ by Scott Meyers, C++ Concurrency in Action by Anthony Williams, and competitive programming on Codeforces or LeetCode (focusing on medium to hard problems).
For Quantitative Research Roles
Focus on probability theory, statistics, stochastic processes, and time series analysis. Jump's research interviews test whether you can derive results from first principles, not just recall formulas. Expect multi-step probability problems, hypothesis testing scenarios, and questions about experimental design.
You should be comfortable with Python (NumPy, pandas, scikit-learn) and able to write code to simulate and test ideas quickly. Some interviews include a take-home research project where you're given a dataset and asked to find patterns or build a predictive model.
Recommended resources: Probability and Statistics for Engineering and the Sciences by Jay Devore, All of Statistics by Larry Wasserman, and practice problems from Glassdoor and QuantNet.
For FPGA Engineering Roles
Prepare for questions on digital logic design, timing analysis, pipelining, and resource utilisation on FPGAs. You should understand the differences between FPGA and ASIC design flows, and be comfortable writing synthesisable VHDL or Verilog. Expect questions about how you'd implement specific trading operations (market data parsing, order generation) in hardware.
Knowledge of Xilinx or Intel (Altera) toolchains is expected. Familiarity with high-level synthesis (HLS) is a plus but not required - Jump tends to prefer hand-written HDL for maximum control.
For Quantitative Trader Roles
Trader preparation overlaps with quant research but adds emphasis on mental arithmetic, game theory, and decision-making under uncertainty. Practise estimation questions (Fermi problems), expected value calculations, and scenarios where you need to update beliefs with new information (Bayesian reasoning).
Jump's trader interviews may include a mock trading game where you quote prices, manage risk, and respond to changing market conditions. Practise with friends or through online trading simulations to build comfort with real-time decision-making.
General Tips
- Research Jump specifically. Read their (rare) public talks and blog posts. Understand their technology investments, market presence, and culture.
- Prepare questions for your interviewers. Asking thoughtful questions about the team's work, technology choices, or research process signals genuine interest.
- Practise under time pressure. Jump interviews move quickly. If you can't solve problems under a clock, add timed practice sessions to your routine.
Jump Trading vs Other Prop Trading Firms
Jump competes for talent with a small group of elite proprietary trading firms. Here's how they compare across dimensions that matter to candidates. For a more comprehensive breakdown, see our prop trading firms guide.
| Jump Trading | Citadel Securities | Jane Street | DRW Trading | Tower Research | Optiver | |
|---|---|---|---|---|---|---|
| Headquarters | Chicago | New York | New York | Chicago | New York | Amsterdam |
| Founded | 1999 | 2002 | 2000 | 1992 | 1998 | 1986 |
| Employees (approx.) | ~1,500 | ~4,000 | ~2,500 | ~1,200 | ~800 | ~1,800 |
| Primary languages | C++, FPGA | C++, Python, Java | OCaml, Python | C++, Python | C++, Python | C++, Python |
| Asset classes | Futures, equities, crypto, options | Equities, options, fixed income | ETFs, bonds, equities, options | Futures, crypto, options | Equities, futures, options | Options, futures, ETFs |
| HFT focus | Very high | High | Moderate | High | Very high | Moderate |
| FPGA usage | Extensive | Extensive | Minimal | Extensive | Moderate | Moderate |
| Microwave networks | Yes | Yes | No | Yes | No | No |
| Interview difficulty | 5/5 | 5/5 | 5/5 | 4/5 | 4/5 | 4/5 |
| Entry total comp (USD) | $250K - $500K | $250K - $500K | $300K - $550K | $200K - $400K | $200K - $400K | $200K - $400K |
| UK offices | London, Bristol | London | London | London | London | London |
| Secrecy level | Very high | Moderate | Moderate | High | High | Low (public co.) |
Key Differences to Consider
Jump vs Citadel Securities: Both are top-tier HFT firms with extensive FPGA and low-latency infrastructure. Citadel Securities is larger and more diversified (particularly strong in equities market making), while Jump has deeper roots in futures and has been more aggressive in crypto. Citadel Securities has a slightly higher public profile due to its equity market share.
Jump vs Jane Street: Different philosophies. Jane Street is known for its collaborative culture, OCaml-based technology stack, and strength in ETF/bond trading. Jump is more secretive, more hardware-focused (FPGAs, microwave networks), and more oriented towards pure speed. Jane Street's interview process emphasises probability and trading games; Jump's leans more towards systems-level technical depth.
Jump vs DRW: Both are Chicago-based firms with strong futures trading businesses. DRW (founded by Don Wilson in 1992) is slightly older and has a broader business that includes real estate and venture investing. Jump tends to invest more in hardware and low-latency infrastructure, while DRW has a reputation for giving traders more autonomy. Both were early movers in crypto.
Jump vs Tower Research Capital: Tower is another latency-focused firm that competes directly with Jump in HFT strategies. Tower is headquartered in New York and is known for its quantitative trading research and technology-driven approach. The firms often compete for the same engineering talent, particularly in C++ and FPGA roles.
Frequently Asked Questions
Is Jump Trading a good company to work for?
Jump consistently ranks among the most desirable employers in quantitative finance. The compensation is at the top end of the industry, the technical challenges are genuinely interesting, and the firm invests heavily in infrastructure that engineers find compelling to work with. The culture is described as intellectually intense but collaborative - people are hired for their ability, and there's relatively little of the hierarchy you'd find at a bank. The main downsides cited by former employees are the secrecy (you can't talk about your work publicly) and the intensity of the working environment during volatile markets.
How hard is it to get hired at Jump Trading?
Very hard. Jump's acceptance rate for technical roles is estimated at 1% to 3%, making it comparable to firms like Jane Street and Citadel Securities. The firm recruits primarily from top-tier universities and expects exceptional quantitative or engineering skills. That said, Jump does hire candidates from non-traditional backgrounds if their technical abilities are strong enough. Competitive programming achievements, open-source contributions, and research publications can all help your application stand out. If you're building towards a quant career, our guide to becoming a quant outlines the typical preparation path.
Does Jump Trading have offices in the UK?
Yes. Jump has two UK offices - one in London and one in Bristol. The London office focuses on trading European markets and houses quant researchers, traders, and engineers. The Bristol office is a technology hub that handles a significant portion of Jump's engineering and infrastructure work. Both offices are actively hiring in 2026, and UK-based roles offer competitive compensation relative to the London quant finance market.
What happened to Jump Crypto?
Jump Crypto was Jump Trading's digital assets division, active from approximately 2021 to 2023. The division traded cryptocurrency markets, invested in blockchain projects, and contributed to DeFi infrastructure including the Wormhole cross-chain bridge. Following regulatory scrutiny from the CFTC, the resignation of Jump Crypto president Kanav Kariya in mid-2023, and the broader crypto market downturn, Jump scaled back its crypto operations significantly. The division was reorganised as Jump Digital with a reduced mandate. As of 2026, Jump maintains limited cryptocurrency trading activity but at a fraction of its 2021-2022 scale.
What programming languages does Jump Trading use?
C++ is the primary language for performance-critical trading systems. FPGA engineers use VHDL and Verilog for hardware implementations. Python is widely used for research, data analysis, prototyping, and internal tooling. Go and Java appear in supporting infrastructure. Jump's technology culture prioritises performance above all else, so candidates for engineering roles should have strong C++ fundamentals - this isn't a firm where you can rely solely on Python.
Does Jump Trading offer internships?
Yes. Jump runs internship programmes for undergraduate and postgraduate students, primarily during the summer. Intern roles are available across quantitative research, software engineering, and FPGA engineering. The internship interview process mirrors the full-time process (online assessment, phone screens, and a final round), and a strong internship performance is one of the most reliable paths to a full-time offer. Jump recruits interns from universities including MIT, Stanford, CMU, Cambridge, Oxford, Imperial, and ETH Zurich, though applications from other institutions are considered.
Want to go deeper on Jump Trading: What They Do, Careers & How to Get Hired in 2026?
This article covers the essentials, but there's a lot more to learn. Inside Quantt, you'll find hands-on coding exercises, interactive quizzes, and structured lessons that take you from fundamentals to production-ready skills — across 50+ courses in technology, finance, and mathematics.
Free to get started · No credit card required