Monte Carlo trade resampling
Did the observed trade sequence hide paths that would have been much harder to survive?
Generate many plausible paths from the submitted outcomes and inspect their distribution. One historical order is only one path through the same trade set.
THE PRESSURE TEST
Move the convenient assumptions.
Trade-order reshuffling exposes sequence risk; replacement resampling exposes sensitivity to which outcomes happened to appear in the sample. They answer related but different questions.
- Compare the observed drawdown with a high-percentile resampled drawdown.
- Inspect the distribution instead of reporting only the median path.
- Run replacement and no-replacement experiments separately.
- Record the random seed and method version so the result is reproducible.
The published history may alternate winners and losers neatly. The same outcomes in a different order can cluster losses and create a drawdown far larger than the one shown. That path was not observed, but it was compatible with the submitted trade set.
The check stops here.
Resampling commonly treats trades as more independent and stationary than markets are. Serial dependence, changing volatility, capacity and new regimes are not repaired by running more simulations.
SEE THE CALCULATION, NOT A CLAIM