Lab Monkey Test
- Indicadores
-
Gilberto Cardona Velez
Laboratorio de Estrategias - transformamos ideias de trading em sistemas automatizados e testados, com veredictos honestos baseados em dados.
Serviços:
- Expert Advisors para MT4 / MT5 feitos com as suas regras (entradas, saídas, risco, filtros de horário), com auditoria de risco por escrito - Versão: 1.11
- Atualizado: 19 novembro 2026
- Ativações: 5
Most strategies that look good in a backtest would have done just as well by entering at random. Lab Monkey Test tells you whether yours is different.
How it works
1. It reads the closed trades of your strategy: from the account history (filter by symbol, magic number and dates), from a CSV file, or from a manual summary (trade count and net result).
2. It builds hundreds of "monkey" versions of your strategy: the same number of trades, the same stop and target, entries at the same hours of the day and the same long/short ratio, but with random entry bars.
3. It applies the bar spread and your average commission to every random trade, so the comparison is fair.
4. It shows where your real result falls inside the random distribution, with a histogram on the chart.
The verdict
- PASSES: fewer than 5% of the random versions (configurable) did as well as your strategy.
- DOES NOT PASS: your result is within what random entries usually achieve. The verdict is never softened.
What you get
- Histogram panel on the chart: random distribution, your result (red line) and the 95th percentile.
- Median, 5th and 95th percentile of the random results, % of random versions equal or better, and p-value.
- CSV file with every random result and a BMP image of the panel, saved to MQL5\Files.
- Reproducible results with a fixed random seed.
Calibration
We checked the method with 20 strategies that have no edge at all (pure random signals). None of them received a false PASS.
Inputs
- Source of the strategy trades: Account history / CSV file / Manual summary
- Period start and end, magic number filter, CSV file name
- Number of random versions (default 500)
- Stop and target in points (0 = use each trade's own)
- Same entry hours, same long/short ratio, bar spread, extra cost per trade
- Pass level (default 5%), random seed
- Panel position and size, save CSV/BMP report
Notes
- Analysis tool only: it never opens, modifies or closes orders.
- Use at least 100 trades for a meaningful result; with fewer trades the panel still works but the test is weak.
- Attach it to the same symbol and timeframe your strategy traded, with enough history loaded.
- In the free demo (Strategy Tester) there is no account history: use the "Manual summary" source.
- Passing the test does not mean a strategy will make money. It only means its past result is hard to explain by luck.
Not financial advice. Past results do not guarantee future results.
