How to Backtest a Forex Strategy (Manual Method)
Your strategy has rules. Now you need to know if the rules work.
Researched and written by Laurent Why Backtest?
You have a strategy with rules. But you have no idea if it's profitable or not. You could trade it live with real money and find out, but that's an expensive, slow, and emotionally destructive way to learn. Backtesting lets you run your strategy through years of historical data in days, giving you statistical evidence about whether it has an edge before you risk a penny.
Backtesting is not perfect. Past performance doesn't guarantee future results. Markets change. But a strategy that was consistently profitable over 1,000 historical trades across different market conditions is far more likely to be viable than one you made up yesterday with no data behind it.
Manual Backtesting: How to Do It
Manual backtesting is available to anyone with a charting platform that allows you to scroll through history:
- Set up your chart: Load the pair and timeframe you trade. Add your indicators.
- Scroll back in time: Go back at least 1 year. Further is better.
- Cover the right side of the chart: You need to see price as if it's unfolding in real time. Only look at what you would have seen at each moment.
- Apply your rules strictly: Every time your entry conditions are met, take the trade at the correct price. Record it. Every time conditions are NOT met, skip it.
- Move forward candle by candle: Manage the trade according to your rules (stop-loss, take-profit, trailing stop).
- Record every trade in a spreadsheet: Date, pair, direction, entry, stop, target, outcome (win/loss), pips, R-multiple.
What to Record
| Data Point | Why It Matters |
|---|---|
| Win/loss result | Basic win rate |
| Pips won/lost | Average win and average loss in pips |
| R-multiple | Win in terms of units of risk (1R = exactly your planned risk; 2R = twice your risk) |
| Date and market condition | Identifies if strategy worked better in trending vs ranging periods |
| Maximum adverse excursion (MAE) | How far against you each trade went before recovering (or not) |
Sample Size: Why 50 Trades Is the Minimum
Trading has variance. Even a good strategy can have 7 losses in a row through pure bad luck. Ten trades tells you almost nothing. 30 trades are suggestive. 50 trades start to show a pattern. 100+ trades give you meaningful statistical confidence.
If your backtest has only 20 trades, you don't have a sample, you have an anecdote.
Key Takeaways
- • Backtesting means applying your strategy to historical data and recording every trade it would have taken.
- • Minimum sample: 50 trades. Better: 100+. Under 50 is not statistically meaningful.
- • Manual backtesting: scroll through historical charts, trade bar by bar.
- • Track: win/loss, average win pips, average loss pips, maximum drawdown, risk-reward.