Poisson Limit for Fixed Points

This quant interview question is about the asymptotic distribution of the count of special positions in a random mapping. It sits at the intersection of discrete random variables, combinatorics, and probability limits, and is a classic example where a complicated finite model simplifies beautifully as the system size grows. For candidates doing serious quant prep, it's a clean benchmark of comfort with random functions and their limiting behavior.

It trains your intuition for rare events and counting arguments in large systems, as well as your ability to formalize that intuition in probabilistic terms. You need to connect a discrete combinatorial structure with a limiting random variable and recognize when a count of events stabilizes in distribution. This is central to building a strong quant interviews toolkit.

This matters for quant interviews because many models in trading and risk involve large universes of instruments, states, or paths with small individual probabilities. Interviewers use such questions to test whether you can see when a complex dependence structure behaves like a simpler limiting distribution. Strong performance here signals that you can move from exact finite descriptions to workable approximations, a crucial skill in real quant research and front-office roles.

What it tests

When analyzing the number of occurrences of a rare event across many independent trials, the binomial distribution with parameters $n$ and $p$ (where $p$ is small and $n$ is large, such that $np$ is constant) converges to a Poisson distribution with mean $\lambda = np$. This is known as the Poisson Limit Theorem. The reason this works is that as $n$ increases and $p$ decreases, the probability of multiple events overlapping in the same trial becomes negligible, so the distribution of the total count is well-approximated by the Poisson. This principle applies broadly to problems where you count the number of times a specific outcome occurs independently in a large population, especially when individual probabilities are small.