Quant Researcher vs Trader vs Developer vs C++ Engineer
The four most common quant roles, how their day-to-day work and interviews differ, what each pays, and how to decide which one you should actually target.
"Quant" is not one job. It is a family of roles that attract different kinds of people, pay differently, and demand different skills. A quant researcher, a quant trader, a quant developer, and a C++ low latency engineer all work at the same firms, often on the same desk, but their days look nothing alike.
This guide breaks down the four most common roles, how their interviews differ, what they pay, and how to figure out which one you should chase.
A note on the numbers before we start. Every compensation figure below is a candidate-reported range for top-tier firms, drawn from levels.fyi, Wall Street Oasis, and published base-salary ranges in pay-transparency job postings, as of 2026. Treat them as reported ranges, not quotes. Two caveats matter more than the headline: first-year totals are inflated by one-time signing bonuses, so steady-state pay in year two is often lower than year one; and the gap between the elite tier and everyone else is enormous, so these numbers describe maybe two dozen firms, not the industry.
Quant researcher
Researchers build the models that decide what to trade. They spend their days on statistics, data, and signal research: finding patterns in market data, testing hypotheses, and handing strategies to traders who execute them.
The day to day is closer to a data scientist than a trader. You will write a lot of Python for prototyping, read papers, run backtests, and argue about whether a result is real or overfit. The work is intellectually intense and more collaborative than the stereotype suggests.
Interviews skew mathematical: probability, statistics, linear algebra, and brainteasers, plus a coding round that is usually Python friendly. A strong background in math, physics, statistics, or CS is the norm, and many researchers hold PhDs, though masters and exceptional undergraduates do land roles.
New grad researchers at top firms report roughly 400k total in the first year, on a base commonly in the 200k band. Mid-level runs about 700k, and senior researchers with a track record report 1.2M and occasionally beyond, though that ceiling requires shipping research that demonstrably makes money.
Quant trader
Traders — titled quantitative trader at most firms, and simply trader at some of the older prop shops — are the ones making decisions in real time. They manage risk, quote prices, and react to what the market is doing right now. It is fast, high pressure, and deeply collaborative in its own way: you are constantly talking to researchers and other traders.
The day to day is a mix of monitoring positions, tuning parameters, and thinking about edge. Some of it is mental math on your feet, and some of it is careful post trade analysis.
Interviews emphasize probability, expected value, game theory, mental arithmetic, and market making games. The mental math bar is higher here than anywhere else, and there is a performance element: you have to think well while someone watches.
New grad traders start in a similar band to researchers, or slightly above it: roughly 450k total at the top firms, with Jane Street and Citadel Securities at the upper end and reported year-one packages there running higher still once a six-figure signing bonus lands. Citadel Securities' own 2026 postings list a 300k base for university-graduate quantitative traders, which is among the richest published figures anywhere in quant. The real upside comes later, from performance-based bonuses: mid-level traders report 800k, and a strong senior trader can out-earn almost anyone at the firm.
Entry compensation for traders and researchers is close enough that it should not decide your choice. The divergence comes with seniority, and it is about variance rather than level: trader pay tracks the P&L of a book directly and swings hard year to year, while researcher pay is somewhat more insulated from any single desk's bad year.
Quant developer
Developers build the systems the researchers and traders depend on. This is a software engineering role with a finance context: building the research platform, the data pipelines, the execution tools, and everything in between. You do not need to discover alpha, but you need to understand enough about what the quants want to build it well.
The day to day is software engineering: code reviews, systems design, debugging, and shipping. The languages vary by firm, but Python, C++, and increasingly Rust and OCaml all show up.
Interviews are standard software engineering loops with a quantitative flavor. Expect data structures, algorithms, and systems design, plus a conversational check that you can understand the domain. The math bar is lower than research or trading.
New grad compensation is strong, commonly 350k total at top firms, with a steadier, less bonus-dependent structure than trading. It is also the most underrated number in this article: at mid and senior levels, reported quant developer packages at firms like Citadel run well into the mid six figures, which puts a strong developer ahead of many researchers without the same bonus variance. The ceiling is lower than a star trader, but so is the floor risk.
C++ low latency engineer
This is the specialist role, and it is in constant demand. Low latency engineers optimize the hot path: the code that reacts to market data and sends orders. They care about cache misses, lock free data structures, network stack tuning, and every nanosecond between packet and order.
The day to day is performance engineering. Profiling, reading assembly, benchmarking, and occasionally rewriting a hot loop in a way that saves microseconds. It is a deep craft and rewards people who genuinely enjoy the machine.
Interviews are the most systems heavy of the four. Expect C++ deep dives: move semantics, memory model, concurrency, lock free queues, and cache behavior, plus a design round where you build something like an order book and defend its latency. Our C++ roadmap covers the preparation in detail.
New grad compensation is similar to quant developers, roughly 350k total, and top low latency engineers are genuinely hard to hire, which gives you real leverage on the way in and at every renegotiation after. The work is also less likely to be automated away than almost anything else in finance.
How the roles compare
The cleanest way to think about it is where you sit on two axes: math versus engineering, and speed versus depth.
Researchers and traders are math first. Researchers work in depth over days and weeks; traders work in seconds and minutes. Developers and low latency engineers are engineering first. Developers build broadly across the stack; low latency engineers go deep on the performance critical slice.
If you love statistics and open ended research, aim for researcher. If you want fast decisions and thrive under pressure, trader. If you love building systems and shipping, developer. If you care about nanoseconds and know what a cache line is, low latency.
Compensation and career trajectory
The headline numbers only tell part of the story, and they vary a lot by firm, city, and market conditions. At top prop firms and hedge funds in New York or London, the bands above hold. At smaller shops or in lower cost regions, expect 20 to 40 percent less.
The career trajectories differ too. Researchers and traders can move toward portfolio management, where the compensation is tied to P&L and the upside is enormous. Developers and low latency engineers can rise to lead infrastructure teams, or move into technical leadership with less variance but a lower ceiling.
One thing that is consistent across all four: the first job is the hard one to get, and internships are the most reliable door. Firms recruit heavily from their intern classes.
How to decide and how to apply
Do not pick based on the highest headline salary. Pick based on what you would want to do for ten years.
Start with honest self assessment. Do you reach for a notebook or a terminal first? Do you prefer a problem you can solve in an afternoon or one you chew on for a month? Do you want your output to be a model, a decision, or a system? Your answers point at the role.
Then validate with practice. Try a market making game and see if the pressure energizes you. Solve some probability problems and see if you enjoy the reasoning. Write a small C++ program and see if the machine interests you. PuzzledQuant's games and problem set are a low commitment way to test your instincts before you commit to a path.
When you are ready to apply, use the PuzzledQuant jobs board, which aggregates open quant roles across firms, filters by position type and location, and includes the full spread from researcher to low latency engineer. It is the fastest way to see who is hiring for the role you picked.
The industry rewards people who know which of these four they actually are. Figure that out first, and the rest is just preparation.