← Optiver Interview Insights

Optiver·Software Engineer·Onsite - Multi Round·Intermediate

Intermediate
Jun 2026

Summary

Third round at Optiver for a Quant Engineer role. It was a live trading simulation where prices shift in real time and you get a few seconds to accept or reject pop-up offers. Genuinely stressful in a way that's hard to prepare for on paper.

Questions Asked (2)

Q1

In the live trading simulation, how do you continuously calculate expected value under tight time pressure while managing your inventory and avoiding getting picked off by bots?

Algorithms & Data StructuresTechnical Trade-offsAdaptability & Ambiguity
Author's notes

The first few minutes I was fine, then a couple bad fills stacked up and I started second-guessing every quote.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clarifying the simulation's constraints (time pressure, bot behavior, inventory limits) to show you understand the problem. Then describe a systematic approach: precompute fair value using a lightweight model, update it in real-time with streaming data, and use inventory-aware quoting with dynamic spreads to avoid adverse selection. Emphasize trade-offs between speed and accuracy, and how you'd validate and iterate under pressure.

Pro tip: In live trading simulations, latency is critical—focus on optimizing the critical path (e.g., using efficient data structures and avoiding unnecessary computations) rather than building a perfect model. Also, explicitly mention how you'd monitor for bot patterns (e.g., order flow toxicity) and adjust your quotes accordingly.

1. Clarify the problem and constraints

Ask about the simulation setup: what data is available, how fast it updates, what inventory limits exist, and how bots behave. This shows you can quickly scope the problem and identify key variables.

2. Define expected value calculation

Explain how you'd compute EV in real-time: use a fair value model (e.g., based on order book imbalance, recent trades) and incorporate inventory risk. Mention using incremental updates and caching to keep calculations fast.

3. Manage inventory and adverse selection

Describe how you'd adjust quotes based on inventory: skew prices to mean-revert inventory, widen spreads when inventory is high, and use signals (e.g., trade intensity) to detect and avoid bots picking you off.

4. Optimize for speed and adaptability

Discuss implementation choices: use efficient data structures (e.g., heaps for order book), precompute as much as possible, and have fallback strategies if latency spikes. Emphasize monitoring and quick iteration.

5. Validate and iterate

Explain how you'd test your approach in the simulation: start with a simple model, measure P&L and inventory metrics, then refine based on feedback. Highlight the importance of learning from failures under time pressure.

Key Points to Mention

  • Fair value estimation using order book imbalance and trade flow
  • Inventory risk management: skewing quotes and position limits
  • Adverse selection detection: identifying bot patterns and toxic flow
  • Latency optimization: incremental updates, efficient data structures
  • Dynamic spread adjustment based on volatility and inventory
  • Trade-offs between model complexity and speed under time pressure

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.

Q2

How do you stay composed and decide which trades to prioritize when you're on a losing streak mid-simulation?

Adaptability & AmbiguityTechnical Trade-offs
Author's notes

Honestly the hardest part wasn't the math, it was not tilting after a few bad ticks in a row.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Acknowledge the emotional challenge of a losing streak, then emphasize a systematic, data-driven approach to decision-making. Highlight your ability to detach from outcomes, focus on process, and prioritize trades based on expected value and risk management. Show that you can adapt by reassessing market conditions and your strategy without letting emotions dictate your actions.

Pro tip: Interviewers at Optiver value intellectual honesty and emotional control. Admit that losing streaks are inevitable and that the key is to stick to your strategy while being flexible enough to adjust when the data suggests it. Demonstrate that you use losses as feedback to refine your approach, not as a reason to deviate impulsively.

1. Acknowledge and Normalize

Recognize that losing streaks are a normal part of trading and can trigger emotional responses. Briefly mention how you manage stress, such as taking a step back or using breathing techniques to regain composure.

2. Reassess the Situation

Analyze the losing streak objectively: Is it due to market conditions, a flaw in your strategy, or simply variance? Use data to determine whether to stay the course or adjust.

3. Prioritize Based on Expected Value

Rank potential trades by their expected value and risk-adjusted return. Focus on opportunities with the highest probability of success and favorable risk-reward ratios, even if they are less frequent.

4. Implement Risk Controls

Apply strict risk management rules, such as reducing position sizes or setting tighter stop-losses, to prevent further losses and preserve capital.

5. Review and Adapt

After the session, review your decisions and outcomes to learn from the experience. Continuously refine your strategy based on feedback and changing market dynamics.

Key Points to Mention

  • Emotional regulation techniques (e.g., mindfulness, taking breaks) to maintain composure.
  • Data-driven decision-making: using metrics like expected value, win rate, and risk-reward ratio.
  • Prioritization based on opportunity cost and market conditions.
  • Risk management: position sizing, stop-loss orders, and diversification.
  • Adaptability: willingness to change strategy when evidence suggests it's not working.
  • Learning from losses: post-mortem analysis to improve future performance.

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.