Quant interviews are less like software engineering interviews and more like a modified math olympiad. Firms rarely care which framework you know. They care how you think under pressure when the problem is ambiguous, the clock is running, and the arithmetic has to be right.
This plan assumes roughly 10 to 15 hours of preparation per week over 12 weeks. If you are further out, spread the same material over more weeks; if you are closer, compress the early probability review and spend the extra time on mock interviews.
Weeks 1 to 3: Probability foundations
Nearly every quant interview reduces to probability in some form. Get genuinely fluent before touching anything exotic.
- Combinatorics: permutations, combinations, and the discipline of asking "does order matter, and is replacement allowed?" before computing anything.
- Conditional probability and Bayes' theorem, done with real numbers rather than formulas.
- Expectation and variance, including linearity of expectation — the single most useful trick for cutting short problems that look simulation-only.
- Classic distributions: Bernoulli, binomial, geometric, Poisson, uniform, and normal. Know the mean, variance, and a typical use case for each.
Drill standard set pieces until they are reflexive: coin flipping until a pattern appears, dice sums, expected waiting times, and the inversion trick for symmetry problems.
Weeks 4 to 5: Mental math and estimation
You will be asked to compute out loud. That is a skill separate from knowing the math, and it is trained separately.
- Two-digit multiplication and precise division, daily.
- Fractions to decimals cold: 1/6, 1/7, 1/8, 1/9, 1/12, 1/16, and their percentages.
- Squares up to 30 and cubes up to 15.
- Fermi estimation: annual US gasoline consumption, number of piano tuners in a city, revenue of a stadium on game day. The point is structured assumptions and clean arithmetic, not a right answer.
Use the mental math and estimation drills on the resources page — they are built for exactly this.
Weeks 6 to 8: Brain teasers and problem solving
Work problems by category rather than one by one:
- Games: optimal strategy puzzles where you think backwards from the final move.
- Betting and probability games: fair pricing, expected value of option-like payoffs, martingale-style arguments.
- Logic and sequencing puzzles: weighing problems, bridge crossing, handshake counting.
- Geometry under time pressure: areas, angles, and the trick of adding an auxiliary line.
When you miss a problem, write down the key insight in one sentence. Revisit that notebook weekly — patterns repeat far more than candidates expect.
Weeks 9 to 10: Market knowledge and fit questions
For trading-adjacent roles, this is where well-prepared candidates separate themselves:
- Know what has happened in markets over the past few months and have a view on why.
- Understand the basics of what your target firm does — market making, systematic strategies, risk arbitrage, execution services.
- Prepare honest, specific stories: a hard technical problem you solved, a failure and what changed after it, and why this firm specifically.
Weeks 11 to 12: Coding and mock interviews
Most quant interview loops include at least one coding round, usually in Python. Practice array and string manipulation, hash map patterns, and basic probability simulations. You rarely need advanced algorithms — but you do need to write clean code while narrating your thinking.
Finish with at least four timed mock interviews. Talking through a problem out loud is a different skill from solving it alone, and it is the skill that gets evaluated.
The candidates who do best are not the ones who know the most obscure formulas. They are the ones who are fluent in the standard toolkit and unshakeable when a problem goes somewhere unexpected.
What to do the day before
- Light review of flubbed problems only. No new topics.
- Do one short mental math set to warm up, then stop.
- Sleep. Serenity in an interview is a trainable state, and it is trained by being well-rested and over-prepared.