What to Expect from a Citadel Interview
Citadel interviews are among the most technically demanding in the finance industry. The process is designed to test mathematical reasoning, coding ability, and problem-solving under pressure - and it's deliberately harder than most bank interviews.
This guide covers the full process for both Citadel (the hedge fund) and Citadel Securities (the market maker), including real question types, preparation strategies, and what the interview panel actually looks for. If you're preparing for quant roles more broadly, our quant interview questions guide covers the wider landscape of firms.
Citadel vs Citadel Securities - What's the Difference?
They're separate companies. Citadel LLC is Ken Griffin's multi-strategy hedge fund, managing roughly $65 billion in assets. Citadel Securities is one of the world's largest market makers: its own 2026 market commentary calls it the "#1 US Retail Market Maker, executing ~35% of all US-listed retail volume", and CEO Jim Esposito has put its share at "close to 25 percent of daily U.S. equity volumes".
For quant research and engineering, though, the two recruit together. Citadel Securities' published interview guides say that when you interview you are talking "with two firms at once: Citadel and Citadel Securities", and hiring managers across both firms review the same interview feedback.
Citadel (the hedge fund) hires quant researchers, portfolio managers, and fundamental analysts. The work centres on alpha generation - finding signals that predict asset price movements across equities, fixed income, commodities, and macro strategies.
Citadel Securities (the market maker) hires quant traders, quant researchers, and software engineers. The work focuses on market making, execution optimisation, and trading infrastructure. The engineering challenges tend to involve ultra-low latency systems and real-time pricing.
For research and engineering, the published process is the same for both firms. The clearer difference is in trading: Citadel Securities' quant trading track is built around probability, mental maths and market-making games, while the hedge fund's investing seats are described as using stock pitches, modelling cases and portfolio manager rounds. Both are extremely competitive, and neither publishes acceptance rates.
The Interview Process - Timeline and Stages
Citadel Securities publishes step-by-step guides for its quant research and engineering processes. It has no equivalent guide for quant trading, so that track comes from candidate reports.
| Track | Official stages | Official timeline |
|---|---|---|
| Quant research | Remote first round of 45 to 60 minutes on CoderPad, then an onsite of three to five 60-minute interviews (four steps in total) | About four to five weeks, faster for competing deadlines |
| Engineering (interns and new grads) | 45-minute CoderPad interview, then three 45-minute interviews, a leadership interview for a specific team, and a final review | About eight weeks |
| Quant trading | Not published | Candidates report under a month to over two months |
Stage 1: Online Assessment (OA)
Neither official guide lists an online assessment among its four steps, but candidates on several tracks report one.
- Software engineering: a HackerRank test, which three or more sources agree on. One Reddit comment in January 2026 described "2 or 3 problems, around LeetCode medium".
- Quant trading: mostly probability. Two commenters in a 2023 thread described "15 easy prob questions in 30 minutes", and a Sydney candidate in March 2025 reported "many probability questions". A New York intern in October 2025 reported an "OA with coding and probability questions". One 2023 comment said US quant trading roles had no OA at all, so this has varied.
The platform for the quant trading OA is not confirmed by any candidate report we found. A figure of "50 questions in 12 minutes" circulates online, but it comes from a single source that does not distinguish Citadel from Citadel Securities.
Stage 2: First Round
Quant research (official): a "45- to 60-minute remote interview" on video, covering "programming skills, research ability, data structures/algorithms and problem-solving". Python and C++ are the core languages, though the firm says "we welcome any", and you get a CoderPad link beforehand.
Engineering (official): a 45-minute video interview on CoderPad covering data structures, algorithms and problem-solving, plus behavioural questions.
Quant trading (candidate-reported): a 45-minute first round after the OA. A July 2026 thread lists Fermi estimates, market making and dice or coin problems, with one reply warning "It's not standardized".
Stage 3: Superday
Quant research (official): the onsite "usually consists of three to five 60-minute interviews", covering technical and behavioural skills.
Engineering (official): "three 45-minute interviews", with next steps promised "within two weeks", then a leadership interview with a senior engineer for a specific team. Team fit is not assessed in the first two rounds.
Quant trading (candidate-reported): a New York intern in October 2025 described two 45-minute rounds followed by an in-person superday that was "more market-making". A Sydney candidate described two technical interviews on probability and game theory.
Citadel is known for competitive compensation - first-year total comp for quant researchers in London is reported at £150,000 to £250,000, with senior researchers earning significantly more. These are estimates, not Citadel figures.
Did you know?
Tap a tile for a quick interview snippet — then keep going.
Citadel Securities Roles and Programmes
The careers site groups roles into three families: Markets & Trading, Quantitative Research and Engineering. From the current postings:
- Quantitative Trading graduate (US): Chicago, Miami and New York. The posting asks for "advanced training in Applied Math, Engineering, Statistical Modeling, Calculus, Computer Science, Physics", the "ability to deal with uncertainty in a rigorous and statistical approach", and says programming experience is "strongly preferred".
- Quantitative Trader intern (US): one "single application for our summer trading and quantitative research program", with a Learning Series covering asset-class fundamentals, mock trading and game theory challenges. The application asks which coding languages you want to use in interviews, and says experience with "LLMs and agentic systems is highly desirable".
- Quantitative Trading intern (Europe): London, an "11 week program".
- Quantitative Research Analyst intern, BS/MS (Europe): London, Paris and Zurich, 11 weeks from June to August, requiring Python, R or C++. A separate PhD internship runs in London and Paris.
There are also earlier routes in. Discover Citadel Securities is an invitation-only two-day event in New York, London, Hong Kong or Singapore for first and second-year undergraduates. The firm runs datathons, which it describes as "a chance to interview with us", a Trading Invitational, and a three-day Quant Invitational in London with a $10,000 prize.
Reported Citadel Securities Questions
These come from candidate reports, with our worked answers.
The stones in a bag (quant trading superday, October 2025). A bag holds four stones, each black or white, with the split unknown. You draw two and both are black. What is the probability the next one is black, and would you bet on it?
The candidate called it an "extended Bayes" problem, and the point is that the answer depends on your prior, so say yours out loud. If every split from zero to four black is equally likely, the chance of drawing two blacks is k(k - 1)/12 when there are k black stones: 2/12, 6/12 and 12/12 for k = 2, 3 and 4. Normalising gives posteriors of 0.1, 0.3 and 0.6. The next stone is black with probability 0, 1/2 and 1 in those cases, so the answer is 0.3 × 0.5 + 0.6 × 1 = 0.75. But if each stone was independently black or white with probability 1/2, the first two draws tell you nothing about the rest and the answer is 0.5. A good candidate gives both and asks which world they are in before pricing the bet.
Two correlated series (quant research, London, November 2025). Given time series X and Y, if X moves up two basis points, how does Y move? The candidate had to state their own modelling assumptions. The natural start is a regression of Y's changes on X's changes, where the expected move is the slope times two basis points, followed by the obvious caveats: is the relationship stable, contemporaneous or lagged, and are you regressing levels (risking a spurious fit) or changes?
The meeting problem (quant research, reported by one prep site). Two people each arrive at a uniformly random time between 4pm and 5pm and wait 10 minutes. What is the chance they meet? They miss each other only if one arrives more than 10 minutes after the other, which covers two corner triangles of the unit square with total area (50/60)², or 25/36. So they meet with probability 11/36. Only one prep site reports this question, so treat it as a style guide rather than a leak.
Engineering. Candidates report implementing a trie and heap hybrid data structure (intern, March 2026), a harder variant of "parallel courses" and a sliding-window mean (intern, November 2025), C++ questions on iterators, memory management, std::variant and std::vector, and for experienced hires, a k-way merge of files read from disk using a min-heap (April 2026).
For more practice on the trading side, see our market making interview questions.
How Interviews Differ by Role
Quant Researcher
The heaviest emphasis on mathematics. Expect questions on probability theory, stochastic processes, statistical inference, and time series analysis. You'll likely be asked to derive results on the whiteboard and explain your reasoning step by step. Familiarity with the concepts in our probability fundamentals guide is essential.
Typical split: 50% probability/statistics, 25% coding (Python), 15% finance/markets, 10% behavioural.
Quant Trader
Similar mathematical rigour but with added emphasis on decision-making under uncertainty, market microstructure, and mental arithmetic speed. You'll face rapid-fire estimation questions and scenarios where you need to price instruments or make trading decisions on the spot.
Typical split: 35% probability/brainteasers, 25% market knowledge, 25% mental maths/estimation, 15% behavioural.
Software Engineer
Focused on algorithms, data structures, and systems design. Citadel Securities engineering interviews resemble FAANG-style tech interviews but with finance-specific system design questions - think order book architecture, real-time pricing engines, or low-latency message processing. C++ knowledge is a strong advantage.
Typical split: 50% coding/algorithms, 25% system design, 15% language-specific knowledge (C++/Python), 10% behavioural.
Types of Questions Asked
Probability and Statistics
These form the backbone of any Citadel quant interview. Questions range from classic problems to multi-step scenarios that test whether you can reason formally under pressure.
Example 1: Coin sequence problem You flip a fair coin repeatedly. What's the expected number of flips to get two heads in a row?
Approach: Set up states - S0 (start), S1 (one head), S2 (two heads, done). Write the system of equations: E0 = 1 + 0.5 * E1 + 0.5 * E0, and E1 = 1 + 0.5 * 0 + 0.5 * E0. Solving gives E0 = 6 flips.
Example 2: Conditional probability You have two children. Given that at least one is a boy born on a Tuesday, what's the probability both are boys?
Approach: This is a twist on the classic Boy or Girl paradox. The sample space of (child, day) pairs matters. The answer is 13/27, not 1/3 or 1/2. Interviewers want to see you carefully enumerate the sample space rather than jump to an intuitive answer.
Brain Teasers and Estimation
Citadel still uses these, particularly for trading roles. They're testing how you structure problems and communicate your reasoning.
Example 3: Estimation question How many piano tuners are in London?
Approach: Work top-down. London population ~9 million. Estimate households (~3.5 million), fraction with pianos (~5%, giving ~175,000 pianos), tuning frequency (once per year), time per tuning (~2 hours including travel), working hours per tuner (~1,500 productive hours/year). That gives roughly 175,000 * 2 / 1,500 = ~230 piano tuners. The exact number doesn't matter - the framework does.
Example 4: Game theory / betting I roll a fair die. I'll pay you the face value in pounds. How much would you pay to play this game? What if you can re-roll once?
Approach: Base game: E[X] = 3.5, so you'd pay up to £3.50. With one re-roll, you'd re-roll if the first result is 1, 2, or 3 (below the expected value). Expected value becomes: 0.5 * E[X | X >= 4] + 0.5 * 3.5 = 0.5 * 5 + 0.5 * 3.5 = £4.25.
Coding Questions
Citadel expects clean, efficient code. For Python roles, they'll look at your use of standard libraries and algorithmic thinking. For C++ roles, expect questions about memory management, templates, and performance.
Example 5: Algorithm question Given a stream of stock prices arriving in real time, design a data structure that supports: (a) adding a new price, (b) querying the maximum price in the last N seconds. Both operations should be faster than O(N).
Approach: Use a monotonic deque (double-ended queue) that maintains prices in decreasing order, with timestamps. When a new price arrives, pop all smaller elements from the back. When querying the max, pop expired elements from the front. Both operations are amortised O(1).
Example 6: Python coding Write a function that takes a list of trades (each with a timestamp and price) and returns the maximum profit from a single buy-sell pair, where the buy must occur before the sell.
Approach: Single pass - track the minimum price seen so far and the maximum profit at each step. O(n) time, O(1) space. Interviewers will ask you to analyse complexity and handle edge cases (empty list, all decreasing prices).
System Design (Engineering Roles)
Example 7: Order book Design an order book that supports limit orders, market orders, and cancellations. How would you optimise for latency?
Approach: Discuss the price-time priority model, sorted data structures for bid/ask sides (balanced BST or sorted map), and the tradeoffs between memory layout and cache performance. Mention lock-free structures and memory pools for the latency discussion.
Market Knowledge and Finance
Example 8: Options pricing An at-the-money call option on a stock is priced at £5. The stock pays no dividends. Without doing any calculations, what can you say about the price of the corresponding put?
Approach: Put-call parity: C - P = S - K * e^(-rT). For ATM options (S = K), C - P = K(1 - e^(-rT)), which is positive but small. So the put is slightly cheaper than the call. Being able to reason through parity relationships quickly is essential.
Behavioural and Fit Questions
Example 9: Teamwork under pressure Tell me about a time you disagreed with a team member on a technical approach. How did you resolve it?
Approach: Citadel values intellectual honesty and direct communication. Describe a specific situation, what the technical disagreement was, how you presented evidence, and what the outcome was. Avoid vague answers - they want concrete details.
Example 10: Motivation Why Citadel over Two Sigma, Jane Street, or DE Shaw?
Approach: Be specific. Reference Citadel's multi-strategy approach, the particular team you're applying to, or a specific aspect of their technology or culture that genuinely interests you. Generic "prestige" answers won't impress.
Quick quiz
Citadel-style mini quiz
A quick mix of coding intuition and probability.
Average-case time complexity of dict lookup in Python?
How to Prepare
3-Month Preparation Timeline
Months 1-2: Build foundations
- Work through probability and statistics from first principles. Sheldon Ross's A First Course in Probability and Frederick Mosteller's Fifty Challenging Problems in Probability are both excellent starting points. Our probability fundamentals resource covers the core concepts.
- For coding, solve 100+ LeetCode problems focusing on arrays, trees, dynamic programming, and graph algorithms. Prioritise medium and hard difficulty.
- Read at least the first 5 chapters of Mark Joshi's Quant Job Interview Questions and Answers.
Month 3: Citadel-specific preparation
- Do timed practice sessions simulating the OA format - 3 problems in 90 minutes.
- Practice mental maths daily (there are apps for this, or use the Trader Test questions online).
- Run through mock interview scenarios with a friend or study partner, particularly for probability questions where talking through your reasoning matters.
- Read Citadel's recent investor letters and public commentary. Understanding their strategy shows genuine interest.
Recommended Books
| Book | Best for |
|---|---|
| Heard on the Street by Timothy Crack | Brain teasers and quant interview questions |
| A Practical Guide to Quantitative Finance Interviews (the "Green Book") by Xinfeng Zhou | Probability, statistics, and mental maths |
| Quant Job Interview Questions and Answers by Mark Joshi | Derivatives pricing and stochastic calculus |
| Cracking the Coding Interview by Gayle McDowell | Algorithms and data structures |
| Trading and Exchanges by Larry Harris | Market microstructure (essential for trading roles) |
For a broader reading list, see our guide to becoming a quant which covers educational pathways and recommended resources in detail.
What Citadel Looks For Beyond Technical Ability
Strong technical skills get you through the door, but they're not sufficient on their own. Based on patterns from successful candidates and Citadel's own hiring materials, here's what separates offers from rejections at the final stage.
Speed of thought. Citadel interviews move fast. Candidates who take five minutes to start a probability problem - even if they eventually get the right answer - often don't pass. You need to recognise problem types quickly and begin structuring your approach within seconds.
Communication clarity. Interviewers want to follow your reasoning in real time. Mumbling through calculations or jumping to answers without explaining your logic is a red flag. Practice thinking out loud.
Intellectual curiosity. Do you ask good questions about the role, the team's research, or the market? Candidates who treat the interview as purely one-directional miss an opportunity to demonstrate genuine interest.
Resilience under pressure. Interviewers will push back on correct answers, introduce complications mid-problem, and test how you handle being wrong. Staying calm and adjusting your approach matters more than never making mistakes.
Collaboration instinct. Despite the competitive reputation, Citadel's quant teams are highly collaborative. They're looking for people who can debate ideas constructively without ego. If you're also considering the broader quant career path, our quantitative analyst career guide covers the skills and traits firms value across the industry.
Compensation & recruiting notes
Pay ranges in this guide are illustrative estimates from publicly discussed bands and anecdotal reports - not official figures from the employer. Packages vary widely by desk, office, performance, and year. Hiring processes change; nothing here guarantees an interview, assessment format, or offer.
Frequently Asked Questions
How hard is it to get an interview at Citadel?
Very competitive. Citadel receives tens of thousands of applications annually for a few hundred positions. The strongest route in is through campus recruiting at target universities (Cambridge, Oxford, Imperial, UCL, Warwick in the UK; MIT, Harvard, Princeton, Stanford, CMU in the US) or through a referral from a current employee. Experienced hires typically need a strong track record at another top firm.
Does Citadel hire from non-target universities?
Yes, but it's harder. Citadel's online assessment is open to applicants from any university, and strong OA performance will advance you regardless of school name. That said, the reality is that campus recruiting events and on-campus interviews at target schools create more direct pathways. If you're at a non-target, an exceptional GitHub profile, competitive programming results (Codeforces, ICPC), or published research can help your application stand out.
What programming languages should I know for a Citadel interview?
For quant research roles, Python is the primary language tested. Citadel Securities' official guides name Python and C++ as the core interview languages, while saying "we welcome any". For software engineering and quant developer roles, C++ is heavily emphasised - expect questions on memory management, move semantics, and template metaprogramming. Some teams also use Java or Rust. Regardless of role, being fluent in at least one language and able to write clean code under time pressure is non-negotiable.
How does Citadel's compensation compare to other quant firms?
Citadel is consistently among the highest-paying firms in the industry. Graduate quant researchers in London can expect a first-year package of £150,000 to £250,000 (base plus bonus). In the US, total first-year comp for similar roles is typically $200,000 to $400,000. Senior quant researchers and portfolio managers earn significantly more, with top performers reportedly earning seven or eight figures. Compensation is heavily performance-linked, with bonuses making up a large portion of total pay.
How many interview rounds does Citadel have?
It depends on the track. Citadel Securities' official quant research process has four steps, including a 45 to 60-minute remote first round and an onsite of three to five 60-minute interviews, and takes about four to five weeks. The official engineering process has four steps (a CoderPad interview, three 45-minute interviews, a leadership interview and a final review) and takes about eight weeks. Quant trading candidates report an OA, one or two 45-minute rounds and a superday with market-making games.
What is on the Citadel Securities quant trading OA?
Mostly probability, by candidate reports. A 2023 thread described 15 probability questions in 30 minutes, and a 2025 New York intern reported coding and probability questions. Citadel Securities does not publish the format, and one 2023 comment said US quant trading roles had no OA at all.
Can I reapply if I'm rejected?
Usually, though Citadel does not publish a waiting period, so check with the recruiter for your cycle. If you were rejected at the OA stage, focus on improving your timed problem-solving speed. If you made it to later rounds, ask for feedback (Citadel recruiters sometimes provide it) and address specific weaknesses. Many successful candidates were rejected on their first attempt.
Want to go deeper on Citadel & Citadel Securities Interview Process 2026?
This article covers the essentials — next, open your free Quantt prep workspace: a real course lesson, interview practice, and a preview of your personalised plan.
Free lesson + interview practice · Or code in the browser with no signup