Solve stress testing and probability
Company: Morgan Stanley
Role: Data Scientist
Category: Statistics & Math
Difficulty: medium
Interview Round: HR Screen
A quantitative interviewer asks several statistics and probability questions.
1. **Commodity stress testing.** You manage risk for positions exposed to WTI and Brent crude. Describe a framework to stress test the portfolio. Assume you have historical daily returns for both series and want to generate joint downside scenarios over a 10-day horizon. Explain how you would model the marginals, dependence structure, and tail risk. If you use a t-copula, explain the workflow and why it may be preferable to a Gaussian copula.
2. **Variance of a sum.** Let `X` and `Y` be normally distributed random variables with means `mu_X` and `mu_Y`, standard deviations `sigma_X` and `sigma_Y`, and correlation `rho`.
- What is `Var(X + Y)` when `X` and `Y` are independent?
- What is `Var(X + Y)` when they are dependent?
- Under what condition is `X + Y` itself normally distributed?
3. **Monty Hall.** There are three doors, one hides a car and two hide goats. You pick one door. The host, who knows the location of the car, opens one of the remaining doors and always reveals a goat. You may stay or switch. What strategy maximizes the chance of winning, and what is that probability?
4. **Eight balls puzzle.** You have eight visually identical balls, and exactly one is heavier than the other seven. Using a balance scale at most two times, describe a strategy that always identifies the heavier ball.
Quick Answer: This question evaluates proficiency in statistical modeling and probabilistic reasoning, spanning risk-stress testing, marginal and dependence modeling (including copulas), variance and distributional properties, conditional probability paradoxes, and combinatorial logic puzzles.