Monte Carlo for an EA review: what reshuffling trades can tell you

A sequence experiment asks how outcomes change under a defined perturbation of an existing result. It is useful for exploring dependence on trade order, but it does not create new independent market history.

Choose the experiment explicitly

Reordering fixed cash trade results, resampling trades and perturbing execution costs answer different questions. State which one you use. Do not combine outputs from different methods under one unexplained confidence label.

Respect position sizing

For automatic sizing, cash profit from a trade depends on the account path. Simply shuffling realised cash profits may ignore that dependency. A better experiment models the sizing rule or clearly states that it only studies the recorded cash sequence.

Report a range

Keep the seed, number of scenarios and measured outputs. Show a distribution or several percentiles rather than selecting the most attractive path. Preserve the original report as the baseline.

Use it as a complement

MQLume’s historical reviews provide a starting dataset and method. Additional sequence experiments can deepen your understanding, but they do not replace independent periods or establish a probability of future profit.

Reference and further reading

Start with a free editable EA project · Explore MQLume source code

Back to top