A backtest does not prove a strategy will work in the future. It helps answer a more modest but essential question: what would these exact rules have done across past examples? That evidence is more useful than choosing a method because its chart illustration looked convincing.
Freeze the rules first
Write the pair, timeframe, market condition, setup, entry, stop, target, risk model, and no-trade filters. If you change rules while seeing the outcome, hindsight will shape the test.
Use the first-strategy guide to select a testable method.
Choose a sample without cherry-picking
Use a continuous historical period that includes different conditions rather than searching only for perfect examples. Move candle by candle with chart replay when possible so future price is hidden.
Log every valid trade
Record date, pair, direction, entry, stop, target, result in risk units, spread or estimated costs, screenshot, and notes. Include every setup that meets the rules, not only wins or visually clean trades.
Use a consistent risk unit so trades with different stop distances remain comparable. This prevents a large position from making one result look more important simply because more money was exposed.
Review more than win rate
Move from backtest to forward test
After the historical test, apply unchanged rules in a demo account. Live market pacing, spreads, and your attention can expose issues the backtest missed. Compare results before considering small live risk.
Follow the progression in the safe-practice guide.
Frequently asked questions
Can I backtest forex without coding?
Yes. You can manually use historical charts or replay and record every rule-based trade in a spreadsheet.
How many trades do I need for a backtest?
There is no universal number. Use enough examples across varied conditions to observe normal losses, drawdowns, costs, and rule ambiguity.