Quant Interview Question Types
A guide to the main quant interview question types: probability, expected value, statistics, market making, mental math, coding, and communication.
Candidates trying to understand what to practice for quant trading, research, developer, or internship interviews.
Probability and counting
Probability questions test whether you can define outcomes, condition correctly, and count without double-counting. Coins, dice, cards, urns, Bayes rule, and random walks all sit in this bucket, and the surface story matters far less than the method: sample space, complement, symmetry, recursion, or conditioning. A representative prompt is "draw two cards from a shuffled deck - what is the chance both are aces," where the answer is (4/52) x (3/51) = 1/221, and the entire question is whether you noticed that the deck changes between draws.
Expected value and decision questions
Expected value questions ask what an uncertain payoff is worth, and in trading interviews they usually become a bet, a quote, or a sizing decision. The standard opener is a fair die that pays its face value in dollars: the expectation is (1+2+3+4+5+6)/6 = 3.5, so 3.50 is the fair price to play once. The follow-up carries the real signal. If you may reroll once, you should reroll on 1, 2, or 3, which gives (1/2) x 3.5 + (4+5+6)/6 = 1.75 + 2.5 = 4.25. Interviewers often care as much about the risk discussion as the number, especially once variance, bankroll, or repeatability enters.
Statistics and research judgment
Statistics questions cover distributions, variance, covariance, regression intuition, sampling error, and experiment design. A common check is whether you carry the cross term: Var(X+Y) = Var(X) + Var(Y) + 2Cov(X,Y), so two unit-variance signals with covariance 0.5 sum to variance 3 rather than 2 - the distinction behind most portfolio and signal-combination follow-ups. Research-style interviews also test whether you can critique evidence, so a strong answer names its assumptions and says what data would make the conclusion more or less credible.
Market making, games, and estimation
Trading games test decisions under uncertainty. You may estimate a fair value, set a bid and ask, update after trades, or manage inventory, and the interviewer is watching whether your width tracks your actual uncertainty and whether you learn from trade flow instead of defending your first number. Estimation prompts test the same discipline in a different costume: a question like "how many piano tuners work in Chicago" wants a stated chain of assumptions and a plausible range, not a lucky number.
Coding and implementation
Quant developer and many research roles include coding prompts: data structures, algorithms, debugging, small simulations, or data manipulation. The signal is not only reaching an answer but reasoning about complexity out loud, testing on a small example you construct yourself, and handling the edge cases a working answer usually skips - empty input, a single element, ties, and integer overflow in a running sum.
Common mistakes
The dominant mistake is preparing only the type that feels comfortable. A puzzle-heavy plan misses statistics, coding, and communication; a coding-only plan misses probability and expected value; a plan built entirely from famous brainteasers misses the adaptive follow-ups that decide most interviews. Use the question types as a coverage map, mark which ones you have never practiced under time, and spend the next block on the weakest rather than the most enjoyable.
Practice the pattern
Use the LeetQuidity curriculum and calibration to turn this topic into a focused practice plan.