Markov Chain Interview Basics
Markov chain basics for quant interviews, including states, transition probabilities, absorbing states, stationary intuition, and when simple recursion is enough.
Candidates encountering state-transition probability questions in quant interviews.
States summarize the future
A Markov-style setup works when the current state contains all information needed to model the next step. The future depends on the current state, not the full history.
Transitions drive the model
For each state, list the possible next states and their probabilities. The transition probabilities should sum to one from each state. This check catches many setup errors.
Absorbing states
An absorbing state ends or traps the process. Random walk barriers, completed patterns, and game-ending states are common absorbing states in interview questions.
Concrete example
For a coin pattern problem waiting for HH, states can be no progress, have H, and done. From have H, another H goes to done while T returns to no progress.
Keep formalism minimal
Most interviews do not need matrix-heavy Markov chain theory. They need clear states, transitions, boundaries, and a recurrence for probability or expectation.
Common mistakes
Candidates often define states that omit important progress or include unnecessary history. The best state is just detailed enough to make the next transition valid.
Practice the pattern
Use the LeetQuidity curriculum and calibration to turn this topic into a focused practice plan.