Flow Traders Online Assessment
Practice it first: Free Flow Traders numerical + sequences replica → Free, no signup, generated questions with an instant score.
Flow Traders is a low-latency ETF market maker, and its OA reflects that profile: it tests language fluency far more than algorithms. The screen is roughly 7 multiple-choice questions on uncommon C++ and Python quirks (output prediction, undefined behavior, evaluation order) plus 3 coding problems - 2 easy and 1 harder (a reported 'Job Scheduling' balanced-processors question). Know your language corner cases cold.
Practice every firm's OA: timed replicas, free game replicas and guides in one place →
Trader / Junior Dev — Language MCQ + coding
A short, fast screen: about 7 multiple-choice 'what does this print' questions on less-common C++ / Python syntax (operator precedence, integer division, pre/post-increment, list aliasing, default mutable arguments, short-circuit evaluation, char arithmetic), then 3 coding problems (2 straightforward, 1 harder). The timed MCQ section below reproduces the language-quirk half on the clock.
Trader / Junior Dev — Coding (2 easy + 1 harder)
After the language-quirk MCQs, the OA has 3 coding problems (2 easy, 1 harder). Auto-graded below.
- Second Largest Distinct Value Coding · easy
- Balanced Brackets Coding · easy
- Balanced Job Scheduling: Max Jobs on the Fast Processor Coding · medium
Trader / Junior Dev — Probability & statistics
Candidate-reported probability and statistics questions from a Flow Traders screen, alongside the language quirks and coding.
Candidate-reported probability and statistics questions appear alongside the language-quirk MCQs and coding: a hypothesis test for whether a coin or sample is fair (a one-proportion z-test) and a density transform such as a uniformly random point in a disk expressed in polar coordinates. The closest problems in the bank are below.
- Radial Distance Distribution in a Uniform Disk medium Probability
- Type I and Type II Error Trade-off in a Gaussian Z-Test medium Statistics
Flow Traders OA — FAQ
What is the Flow Traders online assessment?
Flow Traders is a low-latency ETF market maker, and its OA reflects that profile: it tests language fluency far more than algorithms. The screen is roughly 7 multiple-choice questions on uncommon C++ and Python quirks (output prediction, undefined behavior, evaluation order) plus 3 coding problems - 2 easy and 1 harder (a reported 'Job Scheduling' balanced-processors question). Know your language corner cases cold.
What is the format of the Flow Traders OA?
It runs as 3 tracks (Language MCQ + coding, Coding (2 easy + 1 harder), Probability & statistics). Trader / Junior Dev, Trader / Junior Dev, Trader / Junior Dev — each timed, auto-graded where applicable.
How do I practice for the Flow Traders online assessment?
Work the sample questions above (each with a full worked solution), then take the interactive Flow Traders OA practice on QuantVault. For the full interview, see the Flow Traders interview-questions set.
Quant Research track: the Fast Fire assessment
Junior Quant Researcher applicants report a different screen from the trader numerical test: an online assessment called Quant Research Fast Fire that mixes multiple choice and coding on data science, algorithms, statistics and machine learning. Prepare it with the statistics and coding problems in the Flow Traders set plus the timed replica for the MCQ and coding halves; the numerical sequence test on this page is the trader track screen. A September 2026 trader-track candidate reports 26 sequences in 25 minutes, harder than the public practice site that hosts it, with interleaved doubling and Fibonacci strands, subtractive and three-term recurrences, index-dependent rules, high-starting primes, and fraction sequences with alternating differences shown in lowest terms; the free replica lists these families.
September 2026 report: it comes as two take-homes. The second link arrives right after you submit the first. Part one is 30 questions in 15 minutes: 25 multiple-choice (mostly single answer, a few choose-all-that-apply) on statistics concepts, machine learning (lasso and ridge, feature selection, sources of error), 2 to 3 Brownian motion questions and covariance matrices, plus 5 standalone questions that lean heavily on integrals. No Python or C++ syntax questions. Part two, labelled Fast Fire 2.2, is 30 minutes for 6 short-answer probability questions and 1 Python coding task: expected number of colour runs through a shuffled deck, a plant that dies after two consecutive missed waterings, the expected distance between two random points in the unit square, a bivariate-normal region probability written with the Q function, the standard deviation of a two-regime normal mixture, an expected fine that needs the third moment of a binomial, and a Monte Carlo estimate of a sample-variance tail for a triangular distribution (at most 100,000 iterations, answer to 2 significant figures). Every one of those is in the Flow Traders problem set below.