Goldman Sachs Interview Questions

Inside the Goldman Sachs interview

Goldman Sachs runs one of the deepest strats and quantitative research benches on the Street, where desk-facing pricing intuition meets rigorous mathematics. Its interviews lean on probability, expectation, and the stochastic calculus behind derivatives, with brain-teasers and clean coding to round out the picture.

What they test

The core is probability and expectation — the largest blocks here — spanning conditional expectation, hitting and stopping times, card and dice setups, and permutation fixed points. Around it sit a stochastic-calculus and pricing layer (Brownian motion, integrated-BM variance, a full Black-Scholes walkthrough) plus a quant-stats core (CLT, filtrations, regression and PCA). Combinatorics brain-teasers and clean coding fill out the breadth strats and QR candidates are expected to show.

The recurring shapes

Expect Brownian-motion algebra — covariance of sums, autocorrelation, variance of the integrated process — and hitting/stopping-time arguments solved by a one-step or martingale recursion. On the pricing side, you are asked to reason about Black-Scholes, not just quote it: where its assumptions break and how desks patch them. The combinatorics and card problems reward a clean counting frame over brute force.

How to approach

Set up the right state variable first — condition on a sum, a first step, or a stopping rule — before grinding algebra. For the stochastic problems, fall back on the defining properties (independent increments, E[W_sW_t] = min(s,t)) rather than memorized results. Treat the pricing and CLT questions as conversations: state assumptions, name where they fail, and give the fix. Keep the coding crisp and correct.

The mix leans medium (19 of 30), with a handful of easy warm-ups and two genuinely hard problems — closest-to-center geometry and integrated-Brownian-motion variance — to separate the field.