Optiver cuts most of its trader applicants in eight minutes, before a single person reads a word they wrote. The gate is a timed arithmetic test the firm calls 80-in-8: eighty questions, eight minutes, no calculator. Miss the bar and the application stops there. Clear it and you move into rounds that feel less like a job interview and more like a trading desk checking whether you stay sharp while numbers fly past.
That front-loaded filter says exactly what the firm cares about. Optiver makes markets in options and other listed products, so its traders quote prices thousands of times a day and earn small edges computed fast. Every stage of the process is built to surface two traits that are hard to fake: arithmetic under time pressure, and sound decisions when the information in front of you is incomplete.
The 80-in-8 test, and why people underestimate it
The questions are not hard. Two-digit multiplication, decimals, a few fractions and percentages. Any single one you would get right every time with a pen. The whole difficulty is the clock. Eighty questions in eight minutes is six seconds each, and the test is tuned so that even strong candidates run out of time before they run out of questions.
People who clear it comfortably have drilled mental shortcuts until they fire without thinking. 48 times 25 is faster as 4800 divided by 4. Percentages flip, so 16% of 25 equals 25% of 16, which is 4. A long column of prices adds up cleaner if you round each one and correct at the end. Optiver is not looking for tricks nobody has heard of. It wants the obvious shortcuts to be reflexes. Roughly four in five applicants do not pass this stage, which is the reason it comes first instead of last.
If you have a few weeks, the single best use of them is a daily arithmetic drill. A speed game like Zetamac, set to the right number ranges, mirrors the format closely, and ten focused minutes a day moves your score more than any clever theory. Treat it like training reps. The goal is to stop computing and start recognizing.
What the rest of the online assessment screens for
The full online assessment runs about an hour and bundles several pieces beyond the arithmetic sprint. There is a number-sequences test where you see a row of numbers and give the next or the missing one. There is a probability speed round, which candidates often call Beat the Odds, with around thirty probability questions packed into roughly ninety seconds. And there is a set of short cognitive games, sometimes run as Zap-N style tasks, that measure attention, reaction time, and how well you hold accuracy as you tire.
The sequences catch people because they reward flexible searching, not a memorized rule. A run might be arithmetic, geometric, two interleaved patterns, the differences between the differences, or built on primes. The actual skill is generating guesses fast and dropping the ones that break on the next term. That is the same mental motion you use reading order flow that keeps changing.
| Stage | What it measures | Where candidates lose it |
|---|---|---|
| 80-in-8 | raw arithmetic speed | freezing on the clock, no rehearsed shortcuts |
| Sequences | flexible pattern detection | locking onto one rule and not switching |
| Beat the Odds | probability at speed | overthinking, then running out of time |
| Cognitive games | attention and reaction | careless slips as focus fades |
The market-making game
This is the round Optiver is known for, and the one worth real practice. Formats vary by office and year, but the shape is always the same. You are given some hidden quantity to price, often the sum of several face-down cards, and you make a two-sided market on it: a price you will buy at and a higher price you will sell at. The interviewer, or a competing group of candidates, then trades against your quote. They can buy at your offer or sell at your bid, and after each trade more cards get revealed and your estimate of fair value moves.
Say the value being traded is the total of five cards and you have seen enough to think it sits near 40. You might quote 38 bid, 42 offered. If the interviewer lifts your offer and buys at 42, that is a signal they think the real number is higher, so you raise your whole market. If they hit your bid, you lower it. You widen the spread when you barely know fair value and tighten it as cards come out and your uncertainty shrinks.
What they score is whether you center your market on a defensible expected value, size your spread to match how unsure you are, and react correctly when someone trades against you. Quoting a tight market when you have almost no information is the classic way to get picked off, and they will pick you off on purpose to see if you notice. Saying your reasoning out loud is not optional here. They want to hear you price the risk, not just land on a number.
The technical interview with a trader
Alongside the games, a working trader runs a round of probability, expected value, and brainteasers. The questions stay close to trading logic rather than abstract puzzles. A few in the spirit of what gets asked:
- You roll a fair die and get paid whatever shows, but you may reroll once and take the second result instead. What is the game worth, and on which numbers should you reroll?
- Two dice get rolled and you are paid their sum. What is fair value, and what market would you quote around it?
- A bag has more red chips than blue. You draw one red. How does that change your estimate, and how would you bet on the next draw?
- What is 13 times 17, out loud, right now.
The reroll question wants you to see that you keep the first roll only when it beats the expected value of rerolling, which is 3.5. So you keep a 4, 5, or 6 and reroll a 1, 2, or 3, which makes the game worth 4.25. They care less about the final figure than about whether you reasoned to it cleanly and could quote a market around your own answer. Being comfortable saying the true value sits somewhere near a number, then defending a bid and offer around it, reads as trader instinct.
The behavioral round nobody should wing
The fit interview looks soft next to the math, and candidates pay for treating it that way. Optiver is reading temperament. Why trading and not research or engineering. How you handle a loss. Whether you tilt after a mistake or correct and move on. Whether competitiveness shows up as focus or as ego. The strongest answers are specific: a time you were clearly wrong, what it cost, and what you changed, told without drama. Anyone who claims they rarely make mistakes has just failed the question.
The through-line across every stage is the trait the 80-in-8 was built to find on day one. Stay precise and unbothered while the clock runs, price what you do not fully know, and adjust the moment the evidence shifts. People who trade well already think this way before they walk in, and the rounds are arranged so the desk learns that about you faster than any resume could tell them.
