Beta's Stopping Time Expectation

This quant interview question is about a stopping time defined by repeated draws from a Beta distribution and a random threshold drawn from the same law. It lives at the intersection of expected value, iid continuous distributions, and symmetry arguments that are standard in advanced quant prep. The setup looks simple but hides a subtle dependence structure between the sequence and the threshold.

It trains your understanding of stopping times, conditional expectation, and how to handle probabilities involving one random variable exceeding another when everything is iid and continuous. It also reinforces comfort with distributional properties of Beta variables and how these tie into more general facts about cumulative distribution functions and expectations.

This matters for quant interviews because it tests whether you can connect measure-theoretic probability ideas to practical calculations. Top firms use such questions to distinguish candidates who truly understand random times, distributions, and expectations from those who only know cookbook formulas.

What it tests

When dealing with the probability that one random variable exceeds another independent and identically distributed random variable (i.e., $\mathbb{P}(X > T)$ for iid $X, T$), the key structure is symmetry and the law of total probability. For continuous distributions, $\mathbb{P}(X > T) + \mathbb{P}(X < T) = 1$ because $\mathbb{P}(X = T) = 0$. The expectation $\mathbb{E}[F_X(T)]$ leverages the fact that the CDF evaluated at an iid variable is uniformly distributed, and for Beta distributions, this expectation is $a/(a+b)$ due to the mean of the distribution. This symmetry allows us to reduce a seemingly complex double integral to a simple function of the parameters. The principle generalizes: for iid continuous random variables, the probability one exceeds the other is determined by the mean of the CDF evaluated at a random draw from the same distribution.