What separates quant researchers, developers and traders at a market maker, how Jane Street, Optiver and IMC define them, pay data and which to hire first.
A quant researcher finds and proves the edge, a quant developer builds the systems that turn it into live orders, and a quant trader owns the risk and P&L once it is in the market. At a market maker the three seats overlap far more than the titles suggest, so the hiring mistake is rarely picking the wrong title. It is writing a brief for one seat when the problem you actually need solved belongs to another.
This guide sets out what each role does inside a market maker or prop trading firm, how the leading firms describe the split, what public pay data shows and how to decide which hire comes first.
The simplest way to separate them is by what each person is accountable for.
In a bank or a large multi-manager fund these can be distinct careers with distinct reporting lines. In a market maker, they tend to sit in the same room, often on the same desk, working on the same problem.
The leading firms are explicit that the three disciplines operate as one team, and their own descriptions are worth reading before you write a job description.
Jane Street says its trading, research and machine learning roles combine to form a collaborative quantitative team, and that most of its people write code as part of their regular work. Its departments page describes traders as focused on enhancing models, analysing their performance and finding opportunities for automation, and notes that the majority of its trading is fully automated. Its quantitative research page makes researchers responsible for creating new strategies and writing the code that implements them.
Optiver describes bringing traders, researchers and engineers into one loop to build models and strategies, with traders expected to design, test and deploy automated trading systems. Its technology page says engineers build the data, pricing, risk and execution systems that power both trading and research, and work in close partnership with traders and researchers.
IMC organises around cross-functional teams of traders, engineers and quant researchers. Its quant research roles centre on building and maintaining algorithms, strategies and models, while its technology roles focus on designing and optimising the firm's global trading system.
Two practical conclusions follow. First, the line between trader and researcher is thin at an electronic market maker, because most trading decisions are expressed in code and parameters rather than clicks. Second, the developer is not a support function. At these firms, engineering quality is part of the edge.
A market-making researcher's work is usually narrower and faster than at a systematic hedge fund. Instead of searching for multi-day alpha, they tend to work on:
Typical backgrounds are mathematics, physics, statistics, computer science and engineering. Optiver's graduate researcher posting, for example, asks for a degree in engineering, physics, maths, econometrics or computer science and programming experience. At senior level, the question that matters is not the PhD. It is whether the person's research has actually reached production and made money, and whether they can explain why some of it did not.
"Quant developer" covers several quite different jobs, and confusing them is one of the most common briefing errors in the sector.
Job ads across the leading firms show the spread, from C++ and FPGA infrastructure roles to on-desk developers who blend statistical modelling, data engineering and applied research. Those are different candidates, and a single job description that asks for all of them will attract none of the best.
At an electronic market maker the trader manages a portfolio of automated strategies rather than executing orders by hand. The core of the job is:
Coding is increasingly expected. IMC's graduate trader posting lists developing and improving trading strategies and algorithms among the duties, and treats programming experience as a plus rather than a requirement. At senior level, judgement under pressure remains the scarcest quality. Plenty of people can describe a risk limit. Far fewer have cut a position at the right moment with real money on the line.
Public data is thin because the bulk of pay is discretionary bonus, which firms do not disclose. What is public is mostly base salary from US visa filings and occasional reporting on firm-level pay pools.
Some practical points for anyone setting a package:
For broader benchmarks across market infrastructure roles, see our salary guide.
It depends on where the constraint sits, not on which title sounds most senior. A useful test is to ask what is currently stopping the firm making more money.
The most expensive mistake is hiring a star researcher into a firm whose systems cannot deploy their work. They leave within a year, and their non-compete means they cannot be replaced from the same pool quickly.
In the UK, yes. Under the FCA's certification regime, approving the deployment of a trading algorithm, or having significant responsibility for monitoring whether one remains compliant, is a certification function under SYSC 27.8.23R. The FCA's guidance makes clear this is not limited to high-frequency trading and applies whether the firm builds the algorithm or buys it from a third party.
That has hiring consequences. A senior quant trader or head of a quant team at an FCA-authorised firm may need to be certified as fit and proper, so references, regulatory history and conduct checks belong in the process, not after the offer. The FCA has also said that compliance and risk functions must keep pace with technology, which means the people overseeing quants need enough technical depth to challenge them. This is general information, not legal advice; take specific advice on your firm's regime.
Most market makers run rigorous technical screens for graduates. The weaker point is usually the senior hire, where firms rely on reputation and the logo on the CV. A better process tests the job itself.
Verify what can be verified. Ask for the scope of P&L owned, the size of team and book, and the systems actually built, then take references that test those claims. Treat claimed P&L figures with care: they are rarely transferable between firms, because the infrastructure, capital and flow that produced them stay behind.
The deepest pools are other market makers and prop firms, the electronic trading desks of banks, systematic hedge funds and, for developers, low-latency technology firms and exchanges. Each source has trade-offs. Bank candidates bring product knowledge and control discipline but may be used to slower release cycles. Hedge fund researchers bring rigour but may never have worked at market-making time horizons. Exchange and technology candidates bring engineering depth but may need time to learn how a trading P&L actually behaves.
Most strong candidates are not looking. They are paid well, locked in by deferred compensation and restricted by non-competes. Reaching them takes a specific pitch: what problem they would own, what resources they will have and why the move makes sense for their career, not just their pay.
That is the kind of search Spearpoint runs for market makers and prop trading firms. We will tell you if the brief is really for a different seat, and if the package will not move the people you want. If you are on the other side of the table, our candidate page explains how we work with quants considering a move, and hiring firms can start on our clients page.
For Sam's deeper take on why strong technical candidates fail weak interview processes, read The Better The Candidate, The Less Likely They Pass Your Process in Talent Before Capital.
At the same level, base salaries are often similar. The difference is in the bonus, which is discretionary and rarely disclosed. Traders and researchers whose work maps directly to P&L tend to have the most variable pay, so in a strong year either can earn far more than base. Public H1B filings put the average US quant researcher base at about $190,000 since the start of 2025.
Increasingly, yes, especially at electronic market makers where most trading is automated. Jane Street says most of its people write code as part of their regular work. IMC treats programming as a plus for graduate traders rather than a requirement. At senior level, a trader who cannot read and question the code running their book is at a real disadvantage.
The labels overlap. A quant developer usually sits closer to the trading desk and works on strategy code, pricing and research tooling. A software engineer may work further from the desk on core infrastructure, connectivity or low-latency systems. Many market makers simply call everyone a software engineer and define the work by team instead.
Hire for the constraint. If good ideas are stuck waiting to reach production, a developer will add more value than another researcher. If the systems are sound but strategies are stale, hire a researcher and make sure the data and tools are ready for them. Hiring a star researcher into a firm that cannot deploy their work usually ends in a quick departure.
Often. Under SYSC 27.8.23R, approving the deployment of a trading algorithm, or having significant responsibility for monitoring whether one stays compliant, is an FCA certification function. That covers many senior quant traders and heads of quant teams at FCA-authorised firms, so fitness and propriety checks belong in the hiring process. This is general information, not legal advice.