D.E. Shaw is one of the oldest and most storied systematic funds in the world, founded in 1988 by David E. Shaw, a former computer science professor, and famous both for its pioneering role in computational finance and for its alumni (Two Sigma's founders came from here, and so did Jeff Bezos). The firm runs both systematic and discretionary strategies from New York, and its quant interview has a reputation to match the brand: selective, rigorous, and heavier on statistical judgment than on speed.
This is a depth interview, not a speed interview. Expect probability and statistics done carefully, regression taken apart pitfall by pitfall, and a sustained discussion of your own research. Mental-math sprints barely feature.
The process varies by group and by seniority, so expect your own version to deviate from the outline below.
- 1Recruiter screen. Background, motivation, and logistics.
- 2Technical phone rounds. One or two calls with quantitative analysts: behavioural questions leading into a research discussion, then one or two technical problems.
- 3Case study. A take-home requiring creative quantitative analysis, reported at 24 to 48 hours.
- 4Virtual onsite. Multiple interviewers, deeper technicals, and a defence of your case-study work and your own research.
Stage 1: Recruiter screen
Standard: your background, why quantitative research, why D.E. Shaw. For PhD candidates, be ready to describe your research compactly; the real interrogation of it comes later, but the screen decides whether it sounds interesting enough to interrogate.
Stage 2: Technical phone rounds
Candidates report rounds that open behaviourally, move into a discussion of your research, and finish with one or two problems in probability or statistics. The statistics questions lean toward judgment rather than computation, and one reported theme captures the flavour:
“What is bootstrapping, and what are the pitfalls of running regressions?”
What it testswhether your statistics survives contact with real data.
The strong answer covers the mechanism (resample with replacement, recompute the statistic, read the distribution) and then volunteers its limits: the bootstrap cannot manufacture information a small sample lacks, and naive i.i.d. resampling is invalid on autocorrelated data, where block methods take over. On regression, the expected pitfalls are collinearity, non-stationarity and spurious fits, outliers, and look-ahead in the data itself. Our bootstrap problem implements the tool, and the validation course covers the pitfalls as a system.
Classic logic puzzles also appear at the margin; the reported example is an information-encoding card puzzle, the kind where the trick is realising how much information an arrangement can carry.
How to prepare
Drill conceptual statistics rather than formulas: the brainteaser bank covers the probability layer, and the validation course's early sections cover exactly the regression-pitfall territory these rounds probe.
Stage 3: The case study
A take-home reported at 24 to 48 hours, requiring creative thinking and quantitative analysis rather than a routine model fit. It is heavily weighted for research roles, and it is graded twice: once on the work, and once on how you defend it.
How to prepare
Treat it as a research exercise with a written-up validation story: what you checked for leakage, how you split the data, what you did not conclude and why. The course's twenty-question checklist is built for exactly this, and the problems below rehearse the mechanics. One is a directly reported task here, and a second was built from the backtest-interpretation territory this firm's rounds probe:
Stage 4: The onsite and the defence
Multiple interviewers, deeper technical rounds, and sustained probing of your case study and your own research: why this method, what would break it, what you would do with more time. Go in ready to name the weak points of your own work before an interviewer finds them; the project-defence lesson shows what that sounds like in the room.
Key tips for success
- Know your own research cold, including its flaws. The interview returns to it repeatedly, and the follow-ups escalate until they find something you cannot defend.
- Bring the validation vocabulary: leakage, survivorship, honest sample sizes, multiple testing. It is the clearest differentiator at this firm type, per our researcher process guide.
- Do not over-invest in speed arithmetic; this process rewards depth.
Closing remarks
D.E. Shaw's process is the project-driven researcher interview in its purest form: statistics with judgment, a demanding take-home, and a defence of everything you claim. Prepare the research discipline first and the puzzle drill second, and go in knowing exactly how much of your own work you can stand behind.