What is Monte Carlo simulation in backtesting?

Monte Carlo simulation re-runs your backtest's trade sequence thousands of times in random order (or with random resampling) to map the range of outcomes your strategy could plausibly produce — not just the single path history happened to take. The output is a distribution: median outcome, percentile bands, and worst-case drawdowns.

Its most practical use is drawdown expectation: if the historical backtest shows a 20% max drawdown but the Monte Carlo 95th percentile shows 45%, you should size positions for the 45% — the historical path was one draw, not the boundary.