Brownian Motion Joint Probability at Two Times

This quant interview question is about understanding the joint behavior of Brownian motion at different times and how its continuous-time structure translates into finite-dimensional distributions. It lives at the intersection of stochastic calculus and multivariate normal theory, a core part of serious quant prep for trading and quantitative research roles. You are tested on how well you connect the abstract definition of Brownian motion to concrete probabilistic statements about simultaneous events.

It trains your intuition for conditional probability in Gaussian processes, especially how to handle multiple time points for a Markov process. You must be comfortable moving between the pathwise description and the joint law of selected times, and with using symmetry and scaling properties of Brownian motion in a precise, quantitative way.

This matters in quant interviews because pricing, hedging, and risk models rely on multi-time distributions of stochastic drivers, not just marginal behavior. Understanding these joint probabilities is crucial for barrier options, path-dependent payoffs, and exposure profiles across time. Interviewers use this style of question to check that your quant prep includes genuine mastery of Brownian motion, not just memorized formulas, and that you can reason rigorously about correlated values in continuous-time models.

What it tests

For problems involving the joint behavior of a Markov process like Brownian motion at multiple times, the key structure is the independence of increments and the Gaussian symmetry. The joint distribution at different times can be decomposed into the value at an earlier time and the independent increment to the later time. This allows joint probabilities to be expressed as integrals over independent normal variables, often reducing geometric regions in the plane to angular or area fractions due to rotational symmetry. The underlying reason is that the normal distribution is stable under linear combinations, and the Markov property ensures that future increments depend only on the present, not the past. This structure turns what looks like a complex dependency into a tractable calculation using geometry and symmetry in the space of independent normals.