Monte Carlo Simulation for Strategy Testing
Backtesting tells you what happened in one historical sequence. Monte Carlo shows you what could happen across thousands of possible sequences — including the worst case.
Why Backtesting Is Not Enough
A backtest runs your strategy through one specific historical sequence. The future will not repeat the same sequence. Different orderings of the same win/loss outcomes produce dramatically different equity curves. A strategy showing 15% max drawdown in a backtest might produce 35% drawdown if the same trades occurred in a different order.
What Monte Carlo Does
It takes your historical trade data (win/loss/size) and randomly reorders the sequence 1,000–10,000 times. For each iteration it calculates the resulting equity curve, maximum drawdown, and final return. The distribution of these simulations shows the realistic range of outcomes your strategy could produce.
Key Outputs to Analyze
- 95th percentile maximum drawdown: The worst drawdown experienced in 95% of simulations — your realistic worst case
- 5th percentile final return: The return achieved in 95% of simulations — your conservative expected return
- Ruin probability: The percentage of simulations hitting your defined ruin threshold (e.g., 30% drawdown). Above 5% means your sizing is too aggressive
Practical Application
Run your strategy at 1% risk through 1,000 simulations. Then run it at 2%. Compare the distribution of maximum drawdowns. The difference between 1% and 2% risk often means 20% vs 40% max drawdown at the 95th percentile — a critical distinction for psychological and financial survival.
Backtesting tells you what your strategy did. Monte Carlo tells you what it might do. Trade the strategy that survives the Monte Carlo worst case — not just the backtested best case.