Forex glossary

Monte Carlo simulation

Monte Carlo simulation repeatedly samples scenarios from a chosen model. Strategy analysis can shuffle trades, resample them with replacement or model price changes to produce distributions of outcomes and drawdowns. Estimates are conditional on data and assumptions; they do not automatically cover every possible loss.

Illustrative example

How does it work in practice?

Assume three trade outcomes of +100, −80 and −80 USD. Reordering them preserves the total loss of 60 USD but changes the equity path; sampling with replacement can also change the final result.

A case to work through

Same set, different path

Hypothetical fixed cash outcomes are +60, −30 and −20.

  1. Every reordering of those three values totals +10.
  2. Sampling with replacement can select +60 three times, producing +180 in that sample.

What follows? Name the sampling method alongside the result distribution.

Example boundary: With equity-dependent sizing, reordering can change final results; the model must specify this.

Method changes sample properties

Method changes sample propertiesConclusions are conditional on the data and method.1ShufflingReorders the same fixed outcomes without changing their sum.2Sampling with replacementCan repeat an outcome and change sample composition.3Model assumptionsPosition sizing and dependencies need specifying.
Conclusions are conditional on the data and method.
Text description of the diagram
  1. Shuffling: Reorders the same fixed outcomes without changing their sum.
  2. Sampling with replacement: Can repeat an outcome and change sample composition.
  3. Model assumptions: Position sizing and dependencies need specifying.

A useful distinction

Monte Carlo simulation / Backtesting

A backtest generates outcomes from historical rules; Monte Carlo examines alternatives conditional on the selected model and data.

Check your understanding

Identify what shuffling preserves when trade amounts are fixed.

Fixed results +40, −10 and −20 are only reordered. What total remains?

Show the answer and explanation

+10

Shuffling unchanged values preserves their sum.

From a plan to automation tests

Step 11 / 12 · 0 / 12 reviewed

View the full sequence →

Progress is saved only in this browser.

Sources and documentation

  • SciPy – bootstrap

    Resampling with replacement and sample-statistic distributions conditional on the data.

    Document read:

The example illustrates the mechanism. Check current account conditions and instrument specifications with the provider.

Read this term in Polish →

This definition is unclear — send feedback

Feedback goes to the editorial team. Do not include account details, personal data or contact information.

← Back to the glossary