Replay Market Data Interview Guide
Replay market data interview guide for recorded events, ordering, timestamps, deterministic tests, gaps, simulations, and limitations.
Candidates using historical event streams for testing, backtesting, and incident investigation.
Replay reuses recorded event streams
Market data replay feeds past events into code for testing, simulation, debugging, or research. The replay should preserve enough ordering and timing to answer the test question.
Determinism is useful but limited
Replay can make a bug reproducible, but it may not reproduce live network timing, queue pressure, or venue behavior exactly. State what the replay does and does not prove.
Concrete example
A strategy bug found during a volatile minute can be investigated by replaying the same feed sequence and checking order book state after each event.
Gaps and corrections matter
Recorded data may have missing packets, corrections, or timestamp issues. A replay system should surface those conditions rather than pretending the stream is perfect.
Common mistakes
Candidates often say replay validates a strategy. Replay validates behavior under a recorded scenario; it does not guarantee live robustness or profitability.
Practice the pattern
Use the LeetQuidity curriculum and calibration to turn this topic into a focused practice plan.