← BNP Paribas Interview Insights

BNP Paribas·Software Engineer·Technical Phone Screen·Intermediate

Intermediate
Apr 2026

Summary

Technical phone screen for a Quant Engineer role at BNP Paribas, about 30 minutes, structured around the resume but with real technical depth mixed in. They bounce between your background and actual quant fundamentals pretty quickly, so you can't just coast on storytelling.

Questions Asked (3)

Q1

Define the options Greeks, with a focus on delta, and compute the delta exposure for a sample portfolio.

Technical Trade-offsProduct Analytics & Metrics
Author's notes

This is where I felt most pressure.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by clearly defining the Greeks and their role in measuring risk sensitivities, then focus on delta as the first-order price sensitivity to the underlying. For the portfolio delta, explain the aggregation method (sum of delta-weighted positions) and walk through a concrete example with a few instruments, showing the calculation step by step.

Pro tip: Emphasize that delta is not static—it changes with the underlying price (gamma) and time (charm), so portfolio delta must be dynamically managed. Also, mention that for software engineers, implementing Greeks calculations requires careful handling of numerical precision and market data updates.

1. Define the Greeks

Briefly list the main Greeks (delta, gamma, vega, theta, rho) and state that they measure sensitivity of option prices to various factors. Highlight that delta is the most fundamental.

2. Explain Delta in Detail

Define delta as the rate of change of option price with respect to the underlying price. Mention that call deltas are positive (0 to 1) and put deltas are negative (-1 to 0), and that delta approximates the hedge ratio.

3. Describe Portfolio Delta

Explain that portfolio delta is the sum of the deltas of individual positions, each weighted by the quantity and contract multiplier. For example, delta_portfolio = Σ (quantity_i × delta_i × multiplier_i).

4. Compute a Sample Portfolio Delta

Choose a simple portfolio (e.g., 100 shares of stock, 10 call options with delta 0.6, 5 put options with delta -0.4) and calculate the total delta. Show the arithmetic clearly.

5. Interpret and Discuss Implications

Interpret the result: e.g., a delta of 50 means the portfolio gains $50 for a $1 increase in the underlying. Mention that delta hedging involves offsetting this exposure and that delta changes over time.

Key Points to Mention

  • Delta as the first derivative of option price with respect to underlying price.
  • Call delta ranges from 0 to 1; put delta ranges from -1 to 0.
  • Portfolio delta is the sum of individual position deltas, adjusted for quantity and contract size.
  • Delta hedging and its role in risk management.
  • Delta is dynamic: it changes with underlying price (gamma) and time (charm).
  • For software engineers: implementing Greeks requires efficient data structures and real-time updates.

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

Q2

State the standard assumptions behind linear regression and explain how you would diagnose violations of each.

Data ModelingTechnical Trade-offs
Author's notes

Went through linearity, independence of errors, homoscedasticity, normality, no perfect multicollinearity.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Start by listing the standard assumptions of linear regression (linearity, independence, homoscedasticity, normality of errors, no multicollinearity, and no influential outliers). Then, for each assumption, describe a diagnostic method (e.g., residual plots, statistical tests, VIF) and briefly explain how you would address violations if found.

Pro tip: Emphasize that in practice, some assumptions (like normality) are less critical for large samples due to the Central Limit Theorem, while others (like independence) are non-negotiable. This shows you understand the practical implications and can prioritize.

1. List the assumptions

Clearly state the six key assumptions: linearity, independence of errors, homoscedasticity, normality of errors, no multicollinearity, and no influential outliers.

2. Diagnose linearity and independence

For linearity, use residual vs. fitted plots or component+residual plots. For independence, use Durbin-Watson test or plot residuals vs. order.

3. Diagnose homoscedasticity and normality

For homoscedasticity, use residual vs. fitted plots or Breusch-Pagan test. For normality, use Q-Q plots or Shapiro-Wilk test.

4. Diagnose multicollinearity and outliers

For multicollinearity, compute VIF or condition index. For outliers, use leverage (hat values), Cook's distance, or studentized residuals.

5. Explain remedies

Briefly mention potential fixes: transformations, robust standard errors, removing outliers, or using regularized regression.

Key Points to Mention

  • Linearity: residual plots should show no pattern; consider polynomial terms or transformations if violated.
  • Independence: Durbin-Watson statistic near 2 indicates no autocorrelation; for time series, consider ARIMA or GLS.
  • Homoscedasticity: Breusch-Pagan test or White test; use robust standard errors or weighted least squares if violated.
  • Normality: Q-Q plot and Shapiro-Wilk test; with large samples, normality is less critical due to CLT.
  • Multicollinearity: VIF > 5 or 10 indicates problematic multicollinearity; consider removing variables or using PCA.
  • Outliers/influential points: Cook's distance > 1 or leverage > 2p/n; investigate and possibly remove or use robust regression.

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

Q3

Short algorithm and math questions woven into the resume discussion.

Algorithms & Data Structures
Author's notes

These came fast and without much warning between resume questions.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Treat the resume discussion as a two-way conversation: when the interviewer asks about a project or skill, be ready to explain the underlying algorithm or mathematical reasoning. Clarify the problem, state assumptions, and walk through your thought process step by step, even for short questions. If you don't know, reason aloud and connect to known concepts rather than guessing.

Pro tip: Interviewers at BNP Paribas often care more about how you think than the final answer. Verbalize your reasoning, check edge cases, and ask clarifying questions—this shows the analytical rigor expected in a regulated financial environment.

1. Listen and Clarify

Repeat the question in your own words and ask clarifying questions about input size, constraints, or expected output. This ensures you understand the problem and buys time to think.

2. Relate to Your Experience

Connect the question to a similar problem you solved in a past project or coursework. Briefly mention the context and the approach you used, then adapt it to the current question.

3. Outline the Approach

State the algorithm or mathematical method you plan to use, including time and space complexity if relevant. Keep it concise but show you have a structured plan.

4. Walk Through an Example

Use a small example to demonstrate your approach step by step. This makes your reasoning concrete and helps catch mistakes early.

5. Discuss Edge Cases and Optimizations

Mention potential edge cases (e.g., empty input, large numbers) and how you would handle them. If time permits, suggest possible optimizations or alternative approaches.

Key Points to Mention

  • Time and space complexity analysis (Big O notation)
  • Edge cases and input validation
  • Trade-offs between different algorithms or data structures
  • Mathematical reasoning (e.g., probability, combinatorics, modular arithmetic)
  • Real-world application or relevance to financial systems
  • Testing and verification of the solution

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