Citadel Interview Prep: Quant, Trading, and Research Roles
How to prepare for Citadel's notoriously rigorous interview process — covering the math tests, probability questions, brainteasers, and trading simulations that screen for elite candidates.
Citadel Interview Prep: Quant, Trading, and Research Roles
Citadel and Citadel Securities are among the most selective employers in finance. Citadel's quantitative researcher roles are essentially software engineering plus mathematics PhD-level work; Citadel Securities' trader roles require elite mental math and probabilistic reasoning.
The interview process is long, the math is hard, and the bar is real. This guide covers what to expect and how to prepare.
What Citadel Is Looking For
Citadel hires two types of people: quantitative researchers (PhDs who build models) and traders/market makers (who execute and manage risk in real time). The skills overlap but the emphasis differs.
For trading roles at Citadel Securities:
- Mental arithmetic under time pressure
- Probability and expected value in novel scenarios
- Understanding of market microstructure
- Ability to make decisions with incomplete information
For quantitative research roles at Citadel:
- Statistics, probability theory, stochastic processes
- Machine learning and signal generation
- Coding (usually Python/C++)
- Financial mathematics (options pricing, risk models)
Stage 1: Online Assessments
Citadel typically administers HireVue video interviews and online math/probability assessments early in the process.
The Math Assessment
Expect questions spanning:
- Arithmetic speed tests (similar to Optiver's 80-question test but less extreme)
- Probability calculations under time pressure
- Basic statistics questions
Practice target: Comfortable mental arithmetic on 2-digit multiplication, percentages, and fractions in under 5 seconds.
The HireVue Interview
Video interview with recorded questions. Common topics:
- "Walk me through your best quantitative project"
- "Describe a time you made a decision with incomplete information"
- Rapid-fire math/probability problems
Preparation tip: Practice answering probability questions out loud, talking through your reasoning as you go. Interviewers want to see your thought process, not just the answer.
Stage 2: Phone/Virtual Screens
Technical screens with quant researchers or traders on the team you're applying to.
Probability Warm-Ups You'll Face
Q: You roll two fair dice. What's the probability that the sum is 7?
6 ways to sum to 7: (1,6), (2,5), (3,4), (4,3), (5,2), (6,1). Total outcomes: 36. P(sum = 7) = 6/36 = 1/6 ≈ 16.7%
Q: A bag has 4 red balls and 6 blue balls. You draw 2 without replacement. What's P(both red)?
P = (4/10) × (3/9) = 12/90 = 2/15 ≈ 13.3%
Q: Expected value of a roll of a fair die?
EV = (1+2+3+4+5+6)/6 = 21/6 = 3.5
Trickier Probability Questions
Q: You flip a coin until you get heads. What's the expected number of flips?
Geometric distribution with p = 0.5. E[flips] = 1/p = 2 flips
Q: You flip 3 coins. Given that at least 2 are heads, what's the probability all 3 are heads?
P(at least 2 heads) = P(exactly 2) + P(exactly 3) = 3/8 + 1/8 = 4/8 = 1/2. P(all 3 heads | at least 2 heads) = (1/8) / (4/8) = 1/4 = 25%
Q: A stock doubles with probability 0.6 and halves with probability 0.4. What's the expected value after 2 periods?
After 1 period: E[S₁] = 0.6 × (2S₀) + 0.4 × (0.5S₀) = 1.2S₀ + 0.2S₀ = 1.4S₀
After 2 periods: E[S₂] = 1.4 × E[S₁] = 1.96S₀
Note: the geometric mean is (2^0.6 × 0.5^0.4)^1 = 1.19^0.6 × ... actually let's compute: 2^0.6 × 0.5^0.4 = 2^0.6 / 2^0.4 = 2^0.2 ≈ 1.15. So the portfolio grows on average but the geometric mean is lower — this is the volatility drag problem.
Market Structure Questions
Citadel Securities is one of the largest US equity market makers. Expect questions about:
Q: What is bid-ask spread and where does it come from?
The spread compensates market makers for:
- Adverse selection risk: the counterparty might have information you don't
- Inventory risk: holding a position has directional risk
- Processing costs: the cost of facilitating the transaction
Q: If a market maker quotes a $0.05 bid-ask spread and trades 100,000 shares per day, what is their gross daily revenue?
Revenue per share = $0.05/2 = $0.025 (half-spread on each side). On 100,000 shares: $2,500/day gross before costs.
Q: What happens to bid-ask spreads during high-volatility periods?
Spreads widen. Market makers face higher adverse selection risk (informed traders are more active) and higher inventory risk (prices can move more while they're holding). They demand more compensation.
Stage 3: The Superday
The superday is typically 4–8 interviews covering different aspects: technical math, trading games, behavioral, and case studies.
Trading Simulations
Citadel uses trading game simulations. Common formats:
The Estimation Game: You're shown cards or given information and must quote a bid/ask on some outcome. A classic is "I'll show you 5 cards from a shuffled deck. Quote a bid/ask on the number of red cards."
With no information: fair value is 2.5. You'd quote something like 2.3/2.7.
After seeing 3 red cards in the first 5: now you know 3 are red, so out of 47 remaining cards, 23 are red. The remaining 10 cards you'll see (hypothetically) have EV = 10 × (23/47) ≈ 4.9. Updated total: 3 + 4.9 = 7.9 out of 15 ≈ more red than expected.
The Dice Game: Make markets on dice rolls or combinations. The key skill is updating your estimate as information arrives and maintaining consistent bid/ask spreads.
Options and Derivatives Questions
Q: What is delta and what does it tell you?
Delta measures how much an option's price changes for a $1 change in the underlying.
- Deep in-the-money call: delta ≈ 1
- At-the-money call: delta ≈ 0.5
- Deep out-of-the-money call: delta ≈ 0
Q: Why does implied volatility smile/smirk exist?
The Black-Scholes model assumes constant volatility, but markets observe that:
- Out-of-the-money puts on equities trade at higher implied vol than at-the-money options
- This creates a "volatility smirk" (not a full smile in equities)
- It reflects the market's fear of large downside moves (tail risk) and the premium paid for downside protection
Q: A call option has a delta of 0.6 and a gamma of 0.05. If the stock price increases by $1, what's the new delta?
New delta ≈ 0.6 + 0.05 × $1 = 0.65
(Gamma measures how much delta changes per $1 move in the stock.)
Stochastic Processes
For research roles especially:
Q: What is a Martingale?
A stochastic process where E[X_ | X_1, ..., X_t] = X_t. In plain English: your best forecast of tomorrow's value is today's value. A fair game where your expected wealth stays constant is a martingale.
Q: What is geometric Brownian motion and why is it used for stock prices?
GBM: dS = μS dt + σS dW, where W is a Brownian motion.
It's used because:
- Stock prices can't go negative (unlike arithmetic BM)
- Returns (not prices) are approximately normally distributed
- It's analytically tractable (leads to Black-Scholes)
The limitation: real returns have fat tails and volatility clustering, which GBM misses.
Q: What's the difference between a random walk and mean reversion?
- Random walk: future changes are independent of current level. Stock prices in efficient markets.
- Mean reversion: when above the mean, expected to drift down; when below, expected to drift up. Interest rates, volatility (VIX), commodity prices often show mean reversion.
Stage 4: Team-Specific Technical Interviews
If you've reached this stage, you're talking to the team that might hire you. Questions become highly specific to their strategy.
For equity market making at Citadel Securities:
- Adverse selection modeling
- How to handle a sudden large order flow
- How you'd adjust quotes given recent prints
For quantitative research:
- Describe how you'd backtest a strategy rigorously
- What are the pitfalls of overfitting?
- Walk through a time you found a signal that turned out to be spurious
The Behavioral Component
Citadel takes fit seriously. Expect questions like:
- "Describe your most intellectually challenging project."
- "Tell me about a time you were wrong and how you handled it."
- "Why trading/quantitative research over other options?"
The right answers share: intellectual honesty (admitting uncertainty), comfort with ambiguity, and genuine interest in markets and math.
Preparation Timeline
8 weeks out:
- Rebuild probability foundations from scratch. Work through all the expected value, conditional probability, and combinatorics problems you know.
- Aim for 200+ estimation problems via Fermiq's daily drill.
4 weeks out:
- Switch to more advanced topics: stochastic processes (for research roles), options Greeks, market microstructure.
- Practice making markets out loud: quote bid/ask on problems with a study partner.
2 weeks out:
- Focus on weak spots. Which problem types take you longest?
- Read Citadel's recent news (they occasionally publish research, and being familiar with their strategies shows genuine interest).
1 week out:
- Only light review. Interview performance tracks rest and clarity of thinking more than last-minute cramming.
Key Differentiators
Candidates who get Citadel offers typically share:
- Quantitative fluency: they can go from problem to equation to answer without hesitation
- Communication: they explain their reasoning out loud, even when uncertain
- Intellectual honesty: they flag when they don't know something rather than guessing badly
- Market interest: they've actually engaged with markets, not just studied math
The most common failure mode: technically capable candidates who freeze when asked to reason about uncertainty, or who can't explain their thinking while computing.
Sharpen your mental math and quantitative estimation at Fermiq's daily drill — 5 calibrated questions every day in the categories Citadel tests.
Practice the probability category and markets category specifically if you're targeting Citadel Securities.