DRW Interview Guide

How the DRW interview process works

DRW publishes a four-step recruiting sequence for students and new graduates. The main technical difference is the job you applied for: trading and research candidates begin with a quantitative challenge, while software candidates take a coding challenge.

1. Application. Submit a current CV. DRW reviews student applications on a rolling basis and recommends applying when the role opens.
2. At-home technical challenge. Complete a timed probability/logic assessment for trading or research, or a coding assessment for software development.
3. Zoom or phone interview. Spend roughly 45 minutes with a trader, researcher or developer discussing technical problems, your projects and your motivation.
4. Final-round interviews. DRW describes a two-day student/graduate visit: introductions and a challenge on the first afternoon, followed by technical and behavioural interviews on the second day.

The quantitative trading analyst (QTA) path is particularly well documented. DRW names a probability-focused Zoom interview followed by an in-person Superday with a Python quantitative coding exercise and a trading simulation. Software developers have a separate coding-focused final. For experienced hires, the interviews are tailored to the team rather than this student timetable.

DRW online assessment: format, questions and timing

The DRW online assessment is a reasoning test for quant applicants and a Codility-style programming test for developers. The most useful timing data comes from recent candidates, alongside DRW's own assessment guidance.

The DRW quantitative challenge

For research internships, the six-question, 45-minute format is supported by separate 2026 Glassdoor reports and a Wall Street Oasis account. One October 2026 New York applicant described open-response probability and statistics questions. A September 2026 research-intern report specifically mentions Bayes' theorem, birth-death chains, random walks, chi-squared distributions and combinatorics. These are longer problems where setting up the model usually matters more than rapid arithmetic.

A useful rehearsal is to solve six mixed probability questions in 45 minutes, writing a short justification for each result. Spend time on conditioning, transition probabilities and expectation recurrences; simply memorising distribution formulas will leave you exposed to multi-step questions. DRW has also published a sample geometric counting question, which we solve below.

The DRW Codility assessment for software engineering

DRW itself points developer candidates to Codility and says the tasks are evaluated for both correctness and efficiency. Its examples include writing functions for a scenario or solving the rules of a small game. In August 2026, SWE applicants on Reddit described three algorithm questions in 115 minutes; a separate January 2026 Montréal candidate described an assessment lasting about two hours. One 2026 report mentions array changes, balanced substrings and keeping a running balance non-negative. These are coding-test examples, not the quant research exam.

Practise implementing complete solutions, then challenge them with repeated values, empty or minimal inputs, and the largest allowed input sizes. In a typical running-balance problem, for example, you must update state correctly at every step and avoid an algorithm that rescans the whole prefix after every change.

DRW first technical interview: what gets asked?

The first live technical interview is normally around 45 minutes. DRW's own QTA interview description explicitly names probability and statistics, and its general recruiting advice adds detailed discussion of your CV, coursework and research. For software roles, DRW recommends having a computer available for coding.

Trading candidates should be able to calculate an expectation, explain conditional odds and make a defensible market on an uncertain outcome. In a September 2026 London trading-intern report, the conversation included probability and market-making games alongside finance and CV questions. A useful practice prompt is: “Quote a two-sided market on the number of heads in 100 coin flips, then update your fair value when some flips are revealed.”

Research candidates should expect questions that require a model, not just a numerical answer. Practise defining a random variable, choosing a distribution, and explaining how you would check a result with simulation or a hypothesis test. DRW's research-intern job descriptions put particular weight on statistical modelling, Python, numerical methods and backtesting.

Software candidates can be questioned on implementation details after the OA. A candidate interviewing for a London internship in October 2025 described a 45-minute conversation centred on C++ vectors and iterators. Another 2025 account mentions algorithms, data structures, operating systems, networking and system design. In preparation, explain how a vector grows, why iterators can become invalid after reallocation, and how you would diagnose a slow or incorrect program.

What happens at the DRW Superday?

DRW's published student and graduate final round spans two days. The first afternoon starts with introductions and a technical exercise; the second day is devoted to interviews with potential teammates. The firm recommends business-casual clothing and says travel and the detailed schedule are arranged beforehand.

For quantitative trading analysts, DRW explicitly identifies a Python quantitative coding challenge, a trading simulation and multiple technical interviews. In a September 2026 trading-intern final, one candidate reported fitting a statistical model and then completing two technical interviews involving probability, expected value and Fermi estimation. A July 2026 trading candidate reported a group mock-trading activity.

For research applicants, prepare to defend the assumptions in your research: why a model could work, how you split time-series data, what might leak information into the training set, and whether apparent performance survives realistic costs. For software applicants, DRW describes another coding challenge and deeper computer science discussions; expect to reason through a design or implementation in detail rather than recite definitions.

A practical trading-simulation drill is to price a quantity, set a bid and ask, track your inventory after several trades, then revise the quote when new information arrives. Make yourself explain the change in both expected value and risk. That is a preparation exercise, not a claim about DRW's simulation rules.

DRW interview questions by role

Choose the role you actually applied for. DRW's job descriptions distinguish trading analysts who investigate and manage trading ideas, researchers who build statistical models, and engineers who make the systems work under real performance constraints.

DRW-style questions with worked solutions

These examples show the reasoning that matters in DRW's documented assessments and interviews. The first solves a sample problem DRW published; the other two are original EverythingQuant practice questions, not reported past interview questions.

def peak_exposure(events):
    active = {}
    net = peak = 0
    for event in events:
        if event[0] == "ADD":
            _, trade_id, qty = event
            active[trade_id] = qty
            net += qty
        else:
            _, trade_id = event
            net -= active.pop(trade_id)
        peak = max(peak, abs(net))
    return peak

For more worked practice, EverythingQuant's DRW interview question bank contains 40 selected probability, estimation and reasoning problems. It is a preparation bank, not a list of verified DRW interview questions.

How difficult is the DRW interview?

The hardest part of DRW's quantitative assessment is the depth of each problem. Six questions in 45 minutes give approximately 7½ minutes per answer, but an unfamiliar Markov chain or conditional-probability problem can consume that quickly. October 2026 research-intern applicants described the test as difficult, while trading candidates reported increasingly demanding technical rounds.

The pressure changes by role. Trading applicants must calculate correctly and defend quotes under new information; researchers must set up statistical reasoning and explain model assumptions; developers must deliver correct code and then answer detailed questions about performance and systems. A large mental-arithmetic drill is not a substitute for the six-question quantitative challenge.

How to prepare for DRW interviews

Prepare in the same order as the process. The first aim is to pass your actual online assessment; the second is to explain solutions aloud and handle changing constraints in live interviews.

For trading analysts, spend a separate session writing a small Python simulation and practising market-making with inventory changes; DRW explicitly includes both in the QTA final round. For research, implement a simulation or backtest and explain what would make its result unreliable. For software and research technology, practise reading unfamiliar code and explaining how the implementation behaves on large inputs.

Frequently asked DRW interview questions