When people say they want to work in quant finance, they usually have a vague image in mind: Bloomberg terminals, quantitative models, and a very large salary. What they often miss is that "quant" is not one job. It is three deeply different careers that happen to share a building.
Choosing the wrong path is not fatal, but it wastes time, focus, and prep energy. A candidate who spends six months grinding LeetCode for a researcher role, or studying stochastic calculus for a developer position, is preparing for the wrong interview. This guide gives you a clear-eyed view of what each role actually involves so you can stop preparing in general and start preparing for something specific.
The core difference
Think about where each role sits in the trading lifecycle.
Quant Researchers build the strategy. They mine data, develop models, backtest hypotheses, and identify signals that could be profitable. Their output is a set of rules or a model that says "under these conditions, do this."
Quant Traders execute and manage that strategy in live markets. They make real-time decisions, adjust positions, manage risk, and own the daily P&L. Their output is performance.
Quant Developers build the infrastructure that makes all of this possible: the trading systems, data pipelines, execution engines, and risk frameworks that researchers and traders rely on. Their output is the platform.
At large firms, these roles are distinct. At smaller shops, they overlap significantly. But the underlying skill sets and interview requirements diverge enough that you should pick a primary target before you start preparing.
Quant Trader
The quant trader is the most misunderstood role in the industry. The popular image involves shouting and hand signals. The 2026 reality is someone sitting quietly in front of multiple monitors, watching positions, adjusting hedges, and making split-second decisions based on models they understand deeply and sometimes override.
At a market-making firm like Jane Street, Optiver, or Citadel Securities, a trader's job is to quote bids and offers on financial instruments and profit from the spread while managing the accumulated inventory risk. This requires constant awareness of Greeks exposures, volatility regimes, and order flow dynamics. At a prop trading desk or systematic hedge fund, traders manage portfolios built by researchers, monitor positions, execute trades, and make discretionary calls when markets behave in ways the model didn't anticipate.
The best quant traders combine mathematical rigor with psychological resilience. Losing $2M on a bad morning and making correct decisions in the afternoon is a skill. Interviews are partly designed to test whether you have it.
What the interview actually looks like
The trader interview is uniquely intense. Probability puzzles come at pace and you're expected to answer quickly. You'll price options on the spot. Market-making exercises ask you to quote a two-sided market on something deliberately strange: the number of golf balls in a 747, the probability of rain in London tomorrow. The interviewer will trade against you, probing whether you can update prices rationally under pressure.
Coding matters much less here than in the other two roles. The emphasis is on quantitative intuition, speed, and composure.
Who thrives in this role
The trader path suits people who are energized, not paralyzed, by uncertainty. You need to make confident decisions with incomplete information, be comfortable being wrong, and recover quickly. The feedback loop is brutal: you know by end of day whether you were right. Some people love that cadence. Others find it unsustainable after a few years.
Relevant skills: probability and expected value, mental arithmetic under pressure, options pricing and Greeks, market microstructure, game theory, risk management intuition.
Compensation
Entry level at top-tier firms: 400K total comp. Mid-level (three to seven years): 1.5M. Senior traders and partners at top firms can reach 10M or more in strong years. Trading has the highest variance of all three paths. Extraordinary in good years, potentially painful in bad ones.
Quant Researcher
If the trader is the person making decisions, the researcher is the person who figured out what decisions to make. Researchers sit further from live markets but their work determines whether the firm makes money over the long run.
A typical day is a mix of literature review, data analysis, model development, and backtesting. They are applied scientists: they form hypotheses, design experiments with rigorous controls for overfitting, and draw conclusions that inform the firm's strategies. At a market-making firm, they develop the pricing models and risk frameworks that traders use. At a systematic hedge fund, they build the strategies that drive the book.
Most research ideas do not work. A researcher who generates a hundred hypotheses and finds three that hold up out-of-sample is having a productive year. The skill is knowing which failures are informative, which directions still have potential, and when to walk away from a line of inquiry. Intellectual honesty matters as much as intellectual horsepower.
What the interview actually looks like
Researcher interviews are the most academically rigorous of the three paths. Expect deep probability, not just puzzles but distributions, conditional expectation, and stochastic processes. Some firms assign take-home projects where you analyze a dataset and present findings. The bar for mathematical precision is high and the interviewer wants to see your process, not just your answer.
Coding is expected (usually Python) but the emphasis is on clean, correct analytical code rather than low-latency systems work.
Who thrives in this role
The researcher role suits people who are comfortable with long feedback loops. You might spend three months developing a strategy only to find it doesn't hold up out-of-sample. That requires patience and the ability to stay rigorous when confirmation bias is pulling you toward a result you want to see.
If you were the kind of student who genuinely enjoyed statistics or machine learning courses for the intellectual substance rather than just the grade, that's a meaningful signal toward research.
Relevant skills: probability and statistics at depth, stochastic calculus, machine learning, Python for data analysis and backtesting, time series analysis, rigorous experimental design.
Compensation
Entry level: 350K. Mid-level: 900K. Senior researcher comp is closely tied to the performance of the strategies they build. A researcher whose model generates significant alpha can earn 3M or more, often structured as a share of P&L.
Quant Developer
The quant developer is the most underrated of the three roles. Without the systems they build, neither traders nor researchers can do their jobs.
Quant developers write the execution engines, order management systems, data pipelines, risk frameworks, and backtesting infrastructure that run the firm. At high-frequency trading shops, they work on microsecond-level optimizations where every nanosecond of latency is a competitive factor.
The strongest quant developers are not just good engineers who happen to work in finance. They understand the mathematical models their code implements well enough to find bugs that a pure software engineer would miss. That intersection of strong systems engineering and genuine quantitative knowledge is rare, and the market pays accordingly.
What the interview actually looks like
Developer interviews are closest to traditional software engineering interviews, but with meaningful differences. Expect LeetCode-hard algorithmic questions, and expect to be strong at them. You'll also get systems design questions, low-latency programming concepts, and often some financial mathematics. The coding bar is higher than at most tech companies because the stakes of production bugs are higher.
Who thrives in this role
The developer path suits people who take genuine satisfaction in building robust, correct systems. If debugging a performance issue gives you the same pleasure as solving a math problem, you'll do well here. The work is deep, the problems are interesting, and the feedback loop is cleaner than in trading or research.
There is also a stability argument. Developer comp is not directly tied to a desk's P&L, which means bad years on the trading side don't hit developers as hard. The extraordinary upside of a trader's great year doesn't apply, but neither does the downside.
Relevant skills: production-grade C++, Python for scripting and tooling, data structures and algorithms, systems design and architecture, Linux, networking and concurrency, mathematical finance fundamentals.
Compensation
Entry level: 400K. Mid-level: 700K. Senior: 1.5M. Developer comp has the highest base salary of the three paths and the lowest variance. The tradeoff is that the $5M+ bonus years belong to traders and researchers.
How to choose
Ignore the compensation tables for a moment. Senior comp across all three paths is exceptional, and optimizing for peak upside is a poor framework for a decision you are making at the start of your career.
Ask yourself four questions instead.
What kind of problem do you actually enjoy? Not what you think sounds impressive. Probability puzzles and fast decisions point toward trading. Statistical modelling and long research projects point toward research. Building systems and making code fast and correct points toward development.
What is your relationship with uncertainty? Trading asks you to make confident decisions quickly and repeatedly with incomplete information. Research asks you to tolerate months of ambiguity. Development asks for meticulous correctness, where "probably works" is not acceptable. Which of those environments sounds like somewhere you'd do your best work?
Where does your background take you naturally? A PhD in statistics or applied math is a strong signal toward research. A competitive programmer with strong C++ is a strong signal toward development. A math or physics undergrad who solves puzzles quickly and thinks well under pressure is a strong signal toward trading. These aren't absolute, but they are worth taking seriously.
What risk profile can you actually live with? Trading comp is the most volatile. Research comp follows the performance of what you build. Developer comp is the most predictable. There is no right answer, but be honest rather than optimistic.
The most common mistake is candidates who default to "researcher" because it sounds prestigious, or to "developer" because their coding is strong, without asking whether the daily reality of that role is actually how they want to spend their time. The role you want is the one whose day-to-day you find genuinely engaging.
Preparing for each path
Once you have chosen a direction, the prep changes significantly.
Trader prep is about pace and composure. Practice probability questions fast. Get comfortable with market-making exercises. Think out loud, update your reasoning when given new information, and don't freeze. Speed and clarity of thought under pressure is what the interviewer is watching for.
Researcher prep is about depth and rigor. Go deep on probability and statistics. Work through stochastic processes carefully. If you get a take-home project, show your methodology, not just your conclusions. The interviewer wants to see how you think, especially when the result is ambiguous.
Developer prep leans on code and systems. LeetCode-hard questions in C++, systems design, concurrency, and performance. Also spend time on basic financial concepts because understanding what your code actually does matters at these firms.
One last note: at smaller funds and quant startups, these roles blur considerably. A quant at a twenty-person shop might build models, write infrastructure, and manage positions in the same week. That breadth can be enormously valuable early in a career. But for interview prep purposes, identify the primary emphasis of the role you are targeting and prepare for that. At Jane Street or Citadel, you are being evaluated for a specific function, and your prep should reflect that precision.
The first quant job is the hardest to get. Pick a path that fits how you actually think, prepare specifically for it, and optimize for learning over the first few years rather than chasing the highest initial number.