Company-Style Tests

The online screens are where most candidacies end, usually for lack of format practice, not ability. These simulators mirror the real formats and scoring rules so the first time you feel the clock isn't the time it counts.

Formats are faithful approximations based on what candidates publicly report; QuantPit is independent of every firm named. Every paper is generated fresh, no two runs repeat.

Optiver-stylePRIVATE

Arithmetic, 80 in 8

+1 correct · −2 wrong · skip 0

Eighty arithmetic questions, eight minutes, with the brutal −2 penalty that makes guessing a losing strategy. The folklore bar at the toughest desks is a score in the high 40s, most first attempts land under 20.

Take the test →
Flow-stylePRIVATE

Arithmetic, 10-minute endurance

+1 correct · −2 wrong · 100 questions

A longer numeracy grind in the spirit of the Flow Traders screen: the test isn't whether you can compute, it's whether your accuracy survives minute nine. Pacing and skip discipline win it.

Take the test →
IMC-stylePRIVATE

Number Sequences

+1 correct · 20 patterns · 5 minutes

Five terms, find the sixth. Arithmetic and geometric warm-ups give way to second differences, interleaved strands, and recursive rules. It's the screen most candidates have never practiced.

Take the test →
SIG / Jane Street-stylePRIVATE

Probability Quick-Fire

12 questions · 20 seconds each

Rapid-fire probability with no time to second-guess: classify the structure (complement, conditioning, linearity), compute, type. Draws live from the question bank, so every run is fresh, and every miss links to its full solution.

Take the test →
Fermi roundPRIVATE

Estimation Screen

8 quantities · 30s each · 2 pts within ±10%

Point-estimate real quantities under the clock, scored on accuracy bands. Trains the decompose-multiply-sanity-check loop that anchors every trading discussion that starts with 'roughly how big is…'.

Take the test →
Optiver-stylePRIVATE

Reflex Screen

hit +1 · false alarm −1 · 60 seconds

Numbers stream past at a fixed beat; tap only when the rule fires. Measures the filter, not the math, hesitation and trigger-happiness lose at exactly the same rate. New rule every run.

Take the test →
Akuna-stylePRIVATE

Options Quick-Fire

10 MCQs · 18s each · generated fresh

Parity, breakevens, hedge ratios and payoff arithmetic at screen speed, and every wrong choice is a real mistake pattern, so the plausible-looking number is the trap.

Take the test →
Warm-upPRIVATE

Arithmetic, 2-minute sprint

+1 correct · −2 wrong · 30 questions

The daily calibration run. Two focused minutes with real scoring keeps the arithmetic muscle warm without eating your study session.

Take the test →
The flagship · final-round simulationPRIVATE

Assessment Day

Priced bets, a winner's-curse jar auction, and calibration intervals, every decision graded against the computable optimal play, ending in the hiring committee's verdict. Outcomes are luck; only your decisions are scored.

Walk into the room →

How to use these

Run one timed test per study session, log the score, and only then do untimed technique work (the Mental Math question set teaches the actual tricks, difference of squares, fraction wheels, percentage flips). Score trends matter more than any single run: most candidates add 15–25 points to their 8-minute score in three weeks of dailies.