Maven Securities Interview Guide: OA, Process & Questions

The short answer

For Maven Securities trading internships and graduate programmes, prepare for timed numerical and probability assessments, followed by technical and motivational interviews. Some recent routes also include Arctic Shores games, an on-demand video interview and a final assessment with traders.

Maven also recruits quantitative researchers and software engineers, whose assessments can include HackerRank coding and more specialised technical questions. The 2027 recruiting cycle has used different probability-test timings in London and the US, so prepare for the format in your own invitation.

How the Maven Securities interview process works

Maven is an options and derivatives market maker with early-career opportunities in trading, quantitative research and technology. Its published Emerging Talent process covers applications, role-dependent assessments, interviews and a possible assessment centre or final stage.

Apply to one programme. Submit your application through Maven's careers portal and select the role and office that fit your interests.
Complete the assigned online assessments. Trading applicants commonly encounter numerical and probability questions; some programmes add Arctic Shores games or a HackerRank coding assessment.
Complete a recorded interview, if invited. Recent trading and research recruitment paths have included a short on-demand interview on motivation, communication and sometimes technical reasoning.
Speak with traders, researchers or engineers. Technical interviews test how you solve problems, explain decisions and respond to follow-up questions; a recruiter or hiring manager can also assess motivation and fit.
Attend final interviews or an assessment centre. Later rounds focus on deeper problem-solving, collaboration and role-related tasks, including market-making exercises on some trading paths.

Those stages describe the early-career picture, not a compulsory five-round sequence for every candidate. Maven publishes a separate three-part framework for experienced hires: screening and introductions, practical technical assessments, then conversations about personality and potential.

Application and initial screening

You can apply to one Maven Emerging Talent role per annual recruiting cycle across all locations. If your application to a graduate or internship programme is unsuccessful, Maven permits you to apply again in the next year's cycle.

The trading programmes target spring-week students, penultimate-year interns and graduates, while quantitative research and technology have their own entry routes. Read the programme description before selecting an office, because work eligibility and assessment content differ by track.

For a trading application, highlight numerical strength, probability, decision-making and genuine interest in market making. Research applicants should foreground statistical analysis, Python and research projects; engineering applicants should point to substantial code, debugging and performance work.

Programme-specific rule

Maven offers visa sponsorship for graduate roles, discussed later in recruitment. Internship applicants need the right to work in the region when applying.

Maven Securities online assessments: formats and timings

Maven's official recruiting guidance names four assessment types: numerical, probability, coding and Arctic Shores. Numerical and probability assessment support is provided through VidCruiter, coding through HackerRank, and the cognitive games through Arctic Shores.

Trading applicants should prioritise numerical speed and probability before treating coding as a compulsory stage. The employer assigns different combinations by programme, and 2027 applicants have encountered materially different question counts and rules.

Numerical test: mental maths and sequences

The numerical assessment measures rapid, accurate mental arithmetic. Across several recent trading accounts, a timed segment used 50 arithmetic questions in five minutes, often followed by number or letter sequences; one graduate-trader account also described 30 sequences in six minutes.

Train percentages, two-digit multiplication, fractions, decimals, ratios and four-digit addition or subtraction without a calculator. For sequences, recognise arithmetic and geometric patterns, alternating rules, differences of differences and alphabet-position changes.

Example pace

Fifty questions in five minutes leaves six seconds per question, including navigation and typing. This is a documented historical test pace, not a universal Maven score target.

Maven explicitly recommends Zetamac for practice. EverythingQuant's Optiver 80 in 8 tool trains a similar six-second arithmetic pace, although it is not a replica of Maven's test.

Probability assessment: what the 2027 cycle has used

The official Maven sample is a multiple-choice expected-value problem involving repeated rolls of an eight-sided die. Beyond that example, recent trading tests have covered dice and cards, conditional probability, expected values, combinatorics and variance-based decisions.

The table distinguishes specific accounts rather than combining incompatible assessments into a single supposed standard.

Role and recruiting account	Probability section	Other observed details
London trading summer internship, November 2025 interview	18 questions in 30 minutes	Incorrect probability answers carried negative marks; subsequent mental-maths and sequence tests ran as short timed sections.
US trading application for 2027, August 2026 account	20 questions in 10 minutes	Its instructions prohibited pen, paper, calculators and other aids; arithmetic answers were typed.
London six-month trading internship for 2027, September 2026 account	15 questions in 40 minutes	Pen and paper were permitted for probability; its 50-question, five-minute speed-maths section did not allow them.

The 18-in-30 format also appears in a separate graduate-trader account published in October 2026, with a 50-in-5 arithmetic round and a 30-in-6 sequences round. For training, run both a deliberate 15-question, 40-minute probability set and a rapid 20-question, 10-minute recall set; these are EverythingQuant practice drills modelled on observed variants.

Do not transfer the US no-paper rule to the London version. Read the exact assessment instructions for permitted materials, marking and whether answers can be revisited.

Arctic Shores: task-based cognitive games

Maven uses Arctic Shores to assess a range of cognitive skills and traits. Recent trading recruitment accounts describe short activities involving attention, reactions, pattern recognition and decisions under uncertainty, with game combinations differing across applicants.

Use the practice task offered in your invitation, complete the assessment in a quiet environment and take the breaks Maven recommends between tasks. Memorising another person's responses to a behavioural game is less useful than understanding the controls and responding consistently.

HackerRank coding assessment

Maven links an official HackerRank practice assessment for candidates asked to code. In May 2026, a London C++ software-engineering assessment included template metaprogramming and a simplified matching-engine implementation.

For technical preparation, practise data structures, event processing, algorithmic complexity and boundary-case testing in the language relevant to your application. C++ candidates should also be comfortable discussing types, templates, ownership and performance rather than relying on Python-only solutions.

Video interviews and live technical rounds

The next trading-stage conversation focuses on explaining mathematical answers clearly while remaining calm under time pressure. Recent London trader interviews have included mental percentages, probability problems, estimation and combinatorics, with one November 2025 first-round technical interview lasting 15 minutes.

A later London graduate-trader account described a technical probability interview, a separate HR conversation and a final day with market-making games. Recorded interviews have also appeared in 2027 trading and quantitative-research applications, even though Maven's current general process page does not prescribe one for every candidate.

What to practise for a trader's technical screen

Explain a probability model. For repeated dice or card draws, define the states, make assumptions explicit and calculate an expectation without jumping to an unexamined formula.
Calculate aloud. Work through percentage changes, approximate products and comparisons of large quantities while narrating a short, defensible method.
Reason under follow-ups. If the interviewer changes a payoff or reveals new information, update the answer instead of restarting a memorised solution.
Make a market. Estimate fair value, give a bid and ask, and explain how inventory or asymmetric information affects the spread.

For an EverythingQuant practice interview, give yourself 20 minutes for two probability problems, two rapid arithmetic calculations and one short market-pricing question. Speak throughout and reserve the last two minutes for the interviewer to challenge one assumption; this is a preparation benchmark, not a fixed Maven round.

Motivation and behavioural questions

Prepare a specific answer to why Maven's options market-making business interests you. Its graduate training includes options theory, trading simulations, early responsibility and rotations across desks, which gives you concrete points to connect to your own goals.

Expect to discuss a difficult decision, an error you corrected, a time you learned quickly and how you handled conflicting information. Focus on the facts of what you did, the trade-off you considered and what changed after feedback.

What happens in Maven's final interviews?

Maven's published Emerging Talent guidance describes assessment centres and final interviews that evaluate problem-solving, communication and collaboration. In a November 2026 London graduate-trader account, the final stage included market-making games with traders.

A useful simulation is to quote a price for a random payoff, respond to a customer buying or selling at your quote, and adjust your next price after your inventory changes. The exercise reveals whether you know the fair value, can manage risk and can explain the reason for each adjustment.

For quantitative research or engineering, prepare to discuss what your code or model does when the data are wrong, the workload grows or assumptions fail. Experienced trading applicants should be ready for detailed discussion of actual strategies and track records, as described in Maven's separate experienced-hire process.

Interview preparation by Maven Securities role

The most useful preparation depends on which Maven team you have applied to. Trading, research and technology all value problem-solving, but their technical exercises test different forms of it.

Graduate Trader and Trading Intern

Prioritise the numerical and probability assessments, then practise speaking through expected-value questions and market-making decisions. Maven's options desk works with pricing and risk, so an answer that changes sensibly after new information matters more than reciting an isolated result.

Work through a dice or card payoff, quote a two-sided market and explain when you would widen the spread. Review arithmetic and geometric sequences separately from your probability practice, because both have appeared in timed trading screens.

Quantitative Research Graduate

Maven describes its research work as fundamental pricing, algorithm design, signal generation and analysis of real-world data. A 2027 graduate-research recruitment path progressed through numerical and coding tests and a recorded interview before a specialist technical discussion.

For preparation, use Python to simulate a stochastic process and compare the sampled result with an analytical expectation. Review statistical estimation, time-series leakage, option payoffs, delta and gamma, and be ready to explain why a result might disappear out of sample.

A practical EverythingQuant drill is to fit a simple predictive model, validate it using a chronological split and explain the effect of changing transaction-cost assumptions. The technical detail is a research-preparation exercise rather than a disclosed Maven interview question.

C++, Python and other technology roles

Maven's technology teams work on low-latency trading applications, pricing systems and data infrastructure. The firm's recruiting pages identify online coding tasks or pair programming as common practical assessments for technology roles.

For C++, a May 2026 London screening example used template metaprogramming and an order-matching task. Practise implementing a small limit-order book, handling partial fills and cancellations, and testing time-priority rules; then explain memory allocation and runtime trade-offs.

For Python and data engineering, prioritise clean transformations, streaming events and error handling. An older C# candidate account also describes a one-hour pair-programming exercise involving asynchronous functions, illustrating the difference between coding for a job and preparing only generic puzzles.

Experienced trading and engineering hires

Maven describes screening and introductions, technical assessments, then final conversations about fit and potential. Its recruiter screening call is generally 20–30 minutes, while technical tests depend on the seniority and discipline.

Experienced traders should be prepared to discuss their strategies, P&L, risk limits and decisions under adverse market conditions. Engineers can expect practical tasks such as online coding, pair programming or debugging and should be able to justify architecture and performance decisions with examples from their own systems.

Maven-style interview questions with worked solutions

The first question is Maven's own public probability sample, paraphrased from its Emerging Talent page. The market-making and coding exercises are original EverythingQuant practice problems, not purported leaked questions.

1. Maven's published probability sample: two consecutive eights

Keep rolling a fair eight-sided die until the result is eight on two consecutive rolls. What is the expected total number of rolls?

Show the solution

Use two states: E0 is the expected rolls when the previous roll was not an eight, and E1 is the expectation just after rolling one eight. A fresh roll always costs one step.

From the first state, an eight moves you to E1; any other result leaves you in E0. Thus E0 = 1 + (7/8)E0 + (1/8)E1, so E0 = 8 + E1.

From the second state, another eight finishes the game, while any other face returns you to E0. Therefore E1 = 1 + (7/8)E0.

Substituting gives E0 = 9 + (7/8)E0, which yields E0 = 72. The correct choice in Maven's published sample is 72 rolls.

2. Original trading practice: price the maximum of two dice

A contract pays the larger result from two independent fair six-sided dice. What is its fair price, and how might you quote a market around that value?

Show the solution

For a positive integer payoff, add the probabilities that the maximum reaches each level. Here P(max >= k) = 1 - ((k - 1)/6)^2 for k = 1, ..., 6.

Adding the six probabilities gives 6 - (0² + 1² + 2² + 3² + 4² + 5²)/36 = 6 - 55/36 = 161/36. The fair expected payout is approximately 4.47 units.

One illustrative quote is 4.25 bid and 4.70 ask, leaving room for spread and uncertainty. If you accumulate too much inventory, move your quotes to reduce the exposure rather than treating 4.47 as an immutable trading price.

3. Original coding practice: match an incoming buy order

An ask book is sorted by increasing price, with arrival order preserved at equal prices. Fill an incoming buy order against asks at or below its limit, allowing partial fills, and return the remaining book and unfilled quantity.

Show a Python reference solution

Walk through the price-ordered asks. At each acceptable price, fill the smaller of the available size and the remaining buy quantity, then retain any unmatched quantity.

def match_buy(asks, limit_price, quantity):
    fills = []
    remaining_book = []

    for price, size in asks:
        if quantity > 0 and price <= limit_price:
            matched = min(quantity, size)
            fills.append((price, matched))
            quantity -= matched
            size -= matched
        if size > 0:
            remaining_book.append((price, size))

    return fills, remaining_book, quantity

# Assumes positive ask sizes and quantity, sorted by price/time.
# match_buy([(100, 3), (101, 2), (103, 4)], 101, 4)
# -> ([(100, 3), (101, 1)], [(101, 1), (103, 4)], 0)

The implementation runs in linear time in the number of ask entries because it scans the supplied list once. In a production matching engine, discuss how you would avoid scanning unaffected price levels while maintaining price-time priority and consistent order state.

Test zero fill when the limit is too low, an exact fill, a partial fill at one price and the empty book. If the book contains distinct orders at the same price, preserving their original order enforces the assumed time priority.

How to prepare for Maven Securities interviews

Start by matching your practice to the next stage rather than revising every quant subject equally. For the trader track, the fastest improvements usually come from arithmetic fluency, classic probability structures and concise verbal explanations.

Build six-second arithmetic habits. Use Zetamac or a comparable drill, then attempt a 50-question, five-minute sprint. Review slow calculations separately and repeat with fresh questions.
Practise both probability paces. Solve a 15-question set in 40 minutes for deeper modelling and a 20-question set in 10 minutes for rapid recognition. These are practice sets based on specific 2027 applicant formats, not Maven-mandated training scores.
Include sequences and estimations. Work on alternating letter and number patterns, then estimate a large quantity and explain the assumptions behind your range.
Prepare for a live trader conversation. Give fair values, quote two-sided markets, react to inventory changes and answer why Maven without reading a script.
Tailor specialist practice. Quant researchers should validate noisy data and explain model assumptions; engineers should code, test and analyse real stateful systems under time pressure.

Numerical screening

Train arithmetic speed with the 80 in 8 simulator. Use the six-second pace as transferable practice, not as a Maven score predictor.

Probability foundations

Review conditional probability, recursive expectations, variance and distributions in Probability for Quants.

Technical coding

Use EverythingQuant's programming questions for data structures, matching logic and performance trade-offs.

A structured roadmap

Build a role- and deadline-specific plan using the EverythingQuant preparation roadmap.

If you have an interview in the next few days, spend more time explaining a handful of problems clearly than reading dozens of solutions passively. Keep a brief record of mistakes and retry them until you can solve each without the earlier prompt.

Maven Securities interview FAQs

Does Maven Securities use VidCruiter, HackerRank or Arctic Shores?

Yes. Maven identifies VidCruiter support for numerical and probability assessments, HackerRank for coding and Arctic Shores for cognitive games; the assigned combination depends on the programme.

How many questions are on the Maven probability test?

A 2027 US trading assessment used 20 questions in 10 minutes, while a London six-month internship account used 15 in 40 minutes. The older 18-in-30 format also appears in recent graduate and internship accounts, so use the count in your actual invitation.

Can I use pen and paper or a calculator?

An August 2026 US assessment explicitly prohibited these aids, while a September 2026 London internship account allowed pen and paper for probability. Follow the assessment-specific instructions rather than borrowing another cohort's rules.

Does the Maven interview involve market-making games?

At least one November 2026 London graduate-trader interview account included market-making games at the final stage. Practise fair-value estimation, inventory adjustment and explaining changes to your quote.

What score do I need to pass the online assessment?

Maven does not publish a general pass percentage for its different assessment formats. Track accuracy and speed across your own timed sets, and use the marking rules in the invitation when deciding whether to skip or guess.

How long does Maven take to reply after an online assessment?

Maven aims to respond within two weeks after completion of an online assessment. That is an assessment response target, not the duration of the full recruitment process.

Can I apply to more than one Maven internship or graduate role?

No. Maven allows one Emerging Talent application per annual cycle across its regions, so choose the most relevant track and location.

Is the trader interview the same as the quant research or software interview?

No. Trader preparation centres on numerical speed, probability and trading judgement, whereas research includes analysis, modelling and code, and engineering focuses on implementation and systems.

Sources and editorial notes

Evidence checked 4 November 2026. Official Maven pages establish assessment providers and recruiting policies; the dated firsthand sources below illustrate specific applicant experiences rather than guarantees for every office or year.

The eight-sided-die question is paraphrased from Maven's published example, while the market-making and coding exercises are original EverythingQuant practice. No unpublished questions or score thresholds are claimed.

Ready to take your quant preparation further?