Monte Carlo reshuffling shows how lucky the equity curve was

[Amanhasnoname]

A single backtest is one particular order of trades. Shuffle them randomly, and the equity curve changes shape, along with the drawdown and the final result.

Running a thousand shuffles gives a range of outcomes. The median shows the typical path. The worst five percent show the bad luck scenarios, and the best five percent show the lucky ones.

If the original curve sits near the top of the distribution, the test was fortunate, and live results will probably be worse. If it sits near the middle, the result is more representative.

Another variation resamples trades with replacement, which simulates different sequences including repeated bad streaks. The resulting drawdown distribution is a more realistic planning tool than the single historical figure.

Free tools and simple scripts can do this with a list of trade results, and it takes minutes.

Has a Monte Carlo analysis been run on a trade history, and did it change the position sizing?