Quant and trading interviews do not test the same things a software loop does, and they do not all test the same things as each other. A pure market-making shop leans on mental-math speed and trading games; a systematic fund leans on probability and research; an HFT engineering seat leans on C++ and low-latency systems. This index rates all four for every quant and HFT firm we have a full interview guide for, so you can see where a firm is genuinely hard and which kind of hard it is.
The index
Ranked hardest to easiest by overall difficulty. Higher is harder: 1 is a light bar, 5 is among the hardest anywhere. Overall is the plain average of the four score columns; role focus and comp shape are descriptive, not scored.
| # | Firm | Overall | Mental math | Probability | Coding / systems | Trading games | Role focus | Comp shape |
|---|---|---|---|---|---|---|---|---|
| 1 | Jane Street | 4.2 | 5 | 5 | 2 | 5 | mixed | base plus large discretionary bonus |
| 2 | Flow Traders | 4.2 | 5 | 4 | 4 | 4 | mixed | base plus volatility-linked bonus |
| 3 | Optiver | 4.0 | 5 | 4 | 2 | 5 | trader | base plus PnL-linked bonus |
| 4 | IMC Trading | 4.0 | 4 | 4 | 4 | 4 | mixed | base, sign-on, performance bonus |
| 5 | XTX Markets | 3.8 | 3 | 5 | 5 | 2 | mixed | base plus performance-driven bonus |
| 6 | Citadel Securities | 3.8 | 3 | 4 | 4 | 4 | mixed | bonus-weighted, discretionary, partly deferred |
| 7 | Akuna Capital | 3.5 | 3 | 4 | 3 | 4 | mixed | base plus performance-linked bonus |
| 8 | DRW | 3.5 | 3 | 4 | 3 | 4 | mixed | base, sign-on, performance-linked bonus |
| 9 | SIG (Susquehanna) | 3.5 | 3 | 4 | 2 | 5 | trader | base plus performance-tied bonus |
| 10 | Jump Trading | 3.5 | 3 | 3 | 5 | 3 | mixed | base plus performance bonus, partially deferred |
| 11 | Two Sigma | 3.2 | 2 | 5 | 5 | 1 | mixed | base plus bonus, partly deferred |
| 12 | Citadel (Hedge Fund) | 3.2 | 3 | 4 | 4 | 2 | mixed | pass-through, performance-linked bonuses |
| 13 | Tower Research Capital | 3.2 | 3 | 4 | 4 | 2 | mixed | base plus team-PnL bonus, upside-heavy |
| 14 | D. E. Shaw | 3.0 | 2 | 5 | 4 | 1 | mixed | bonus-weighted, performance-tied, partly deferred |
| 15 | Millennium Management | 3.0 | 2 | 4 | 4 | 2 | mixed | base plus pod-PnL-linked bonus |
| 16 | Point72 | 3.0 | 2 | 4 | 4 | 2 | mixed | pod-performance-linked bonuses, PnL-driven |
| 17 | NK Securities | 3.0 | 2 | 4 | 5 | 1 | dev/engineer | base plus large performance bonus |
| 18 | Hudson River Trading | 3.0 | 2 | 3 | 5 | 2 | dev/engineer | base plus performance bonus, partly deferred |
| 19 | AQR Capital Management | 2.8 | 2 | 4 | 3 | 2 | researcher | base plus bonus, partially deferred |
| 20 | Virtu Financial | 2.8 | 2 | 3 | 4 | 2 | mixed | cash bonus plus RSU equity |
| 21 | Renaissance Technologies | 2.5 | 2 | 4 | 3 | 1 | researcher | base plus fund-profit participation |
| 22 | Bridgewater Associates | 2.0 | 2 | 2 | 3 | 1 | mixed | steady, low spike, not PnL-linked |
How we scored it
For each firm we read its full techinterview.org interview guide and rated the four dimensions from 1 to 5, using only what the guide documents. The overall score is the equal-weight average of the four, so a firm that is brutal on mental math but light on coding can land near one that is moderate across the board. Ties break by probability, then mental math. Role focus and comp shape are read straight from the guide and are descriptive only.
| Dimension | What it measures | A 1 looks like | A 5 looks like |
|---|---|---|---|
| Mental math | Rapid mental-arithmetic and speed tests. | No timed arithmetic component. | A brutal timed test (e.g. 80 questions in 8 minutes). |
| Probability | Probability, expected value, combinatorics, and classic brainteasers. | Little to no probability. | Deep, multi-step probability and EV under pressure. |
| Coding / systems | Coding bar and low-latency / systems depth. | Pure-trader loop, little code. | Hard C++ and real low-latency systems design. |
| Trading games | Market-making games, betting/EV games, and live market intuition. | No trading games. | Multiple market-making and betting rounds that decide the offer. |
What the index shows
The hardest loops belong to the elite market-making and prop shops. Jane Street, Flow Traders, Optiver and IMC Trading sit at the top because they stack fast mental math, deep probability, and live trading games, and they run several elimination rounds before an offer. If you are aiming there, arithmetic speed and clean probability reasoning are the floor, not the ceiling.
The more approachable loops, Bridgewater Associates, Renaissance Technologies and Virtu Financial among them, tend to be the systematic funds and the engineering-heavy seats where the bar shifts toward coding, research, or systems and away from timed trading games. They are still real bars, just a different shape.
The pattern worth internalizing: quant difficulty is not one axis. Sort by the column that matches the seat you want — a trader should look hardest at mental math and trading games, an HFT engineer at coding and systems — because a firm that is a 5 for one role can be a 3 for another.
Firm by firm
The one-line rationale behind each rating, with role focus and how pay is structured.
- Jane Street 4.2 — Decisive 60Q/8min math test + hard probability/brainteaser rounds + market-making games; coding applied, non-leetcode, weak coders can pass.
- Flow Traders 4.2 — Optiver-style 80-in-8 math gates traders; dice/cards/conditional probability; real C++/latency depth for eng track; quote-a-market game round.
- Optiver 4.0 — Trader-track: famous 80-in-8 math filter and live market-making games dominate; solid probability brainteasers; explicitly no standard coding rounds.
- IMC Trading 4.0 — Fast mental-math/sequence OA, multi-round probability+EV, high SWE coding/systems bar (not IOI-extreme), and market-making sim/Prosperity games across trader+SW
- XTX Markets 3.8 — Deep probability/stats/ML plus C++ low-latency systems depth; mental-math and brainteaser rounds noted; no market-making/trading games described.
- Citadel Securities 3.8 — Timed prob/brainteasers escalate; JS-style market-making game critical even for SWE; deep C++/HFT systems; mental math noted as less central than Jane Street.
- Akuna Capital 3.5 — Mental-math explicitly lighter than Optiver's 80-in-8; solid probability/EV & poker puzzles; Python SWE depth but not C++ HFT; strong trader market-making games
- DRW 3.5 — Probability-heavy (3-5 brainteasers), market-making games for traders; mental math lighter than Optiver; coding high but below HRT/Jump.
- SIG (Susquehanna) 3.5 — Poker/EV & market-making games are core; probability-heavy; mental math lighter than Optiver's 80-in-8; coding only light sims for traders.
- Jump Trading 3.5 — Brutal CP-level C++ and deep low-latency systems bar; probability/mental-math present but less central than Jane Street; futures market-making for trader track.
- Two Sigma 3.2 — PhD-level stats/probability + FAANG-grade coding & low-latency systems; guide explicitly downplays prop-firm mental-math and trading-game gamesmanship.
- Citadel (Hedge Fund) 3.2 — Mixed hedge-fund tracks: QR stats/Bayes/EV, SWE DS&A high but below HRT, estimation-focused mental math; no market-making games.
- Tower Research Capital 3.2 — Probability/stats and modeling emphasized; high C++/low-latency SWE bar but below HRT; research-flavored, not market-making games; mixed roles.
- D. E. Shaw 3.0 — Probability/stats-heavy, FAANG-hard coding for SWE; guide explicitly downplays mental-math tests and trading gamesmanship.
- Millennium Management 3.0 — QR probability/EV depth and real SWE DS&A + systems-design loop; strategy/risk case-driven, no arithmetic-speed drills or market-making games mentioned.
- Point72 3.0 — Cubist tests prob/stats brainteasers (Green Book) and C++/systems SWE; discretionary pitch focus. No speed-math or market-making games mentioned.
- NK Securities 3.0 — C++/CP-heavy HFT dev loop with Brainstellar probability; no market-making games; mental math only inside puzzles; base+fixed+perf bonus up to 100%.
- Hudson River Trading 3.0 — Engineering-led SWE loop; highest prop-firm coding bar (CP + low-latency systems); probability real but intermediate; minimal games, no mental-math drills.
- AQR Capital Management 2.8 — Heavy Bayesian probability/stats and regression, practical Python (numpy/pandas) coding; research-driven long-horizon culture; only brief mental math, no tradin
- Virtu Financial 2.8 — SWE/systems-heavy, bar below HRT/Jump; clean EV/Bayes/regression, less mental-math than Optiver; no trading games; public-co cash+RSU comp.
- Renaissance Technologies 2.5 — PhD-level probability/stats rigor, research-depth focus; less brainteaser, no trading games; some C++ for SWE; comp via fund participation.
- Bridgewater Associates 2.0 — Macro reasoning + cultural fit dominate; realistic quant/Java coding, not trick puzzles; no speed math or trading games; comp steady, not spike-heavy.
Keep going
Every firm above links to its full interview guide. For the complete set — plus the investment banks and prop desks — see the quant, HFT & prop-trading firm guides hub.
