Backtesting explained simply
A backtest checks how a fixed rule would have performed in the past. It can show a lot and deceive even more. For learning only, not investment advice.
To the trapsWhat is a backtest?
You define a rule precisely, for example: enter when a certain pattern appears at a support zone, loss limit below the low, exit at a fixed target. Then you go through old data and count how often it would have worked and how often not.
That shows the strengths and weaknesses of an idea. It does not show how things will go in the future.
How to go about it
- Write the rule down completely beforehand: entry, loss limit, exit, position size.
- Collect enough cases across different market phases.
- Include costs such as fees and price differences.
- Record the results: hit rate, average win and loss, longest losing streak.
- Then test the rule on other data you have not used before.
The most common traps
- Overfitting: the rule is tweaked until it fits old data perfectly, and then fails.
- Too few cases: ten hits say very little.
- Hindsight bias: you unconsciously use knowledge you did not have at the time.
- Forgetting costs: many small gains disappear with fees.
- Only one market phase: what works in an uptrend can fail in a sideways market.
Why chance plays a big role
Even a rule without any real edge can look good for a while. And a good rule can have long losing streaks. If you do not know this, you throw away good ideas too early or trust bad ones too much. In the Academy's Trader Lab you can try this yourself with the statistics simulator: same rule, very different paths.
Frequently asked questions
Do I need programming skills for a backtest?
No. You can also go through old charts by hand and count in a spreadsheet. It takes longer, but trains your eye.
How many cases are enough?
There is no fixed number. The more cases across different market phases, the more meaningful. A few dozen is rather the lower limit.
Does a good backtest mean the rule works?
No. It only shows how it would have performed in the past. The future can look very different.
You might also like
Trading journal
Why a trading journal helps, what goes into it, how to review it and which mistakes to avoid. With a template you can copy.
Practise without real money
Demo account, paper trading, chart replay or a learning program: how to practise trading without real money, what each way is good for and its limits.
Stop-loss
What a stop-loss is, how to set a sensible loss limit, why the exit does not always happen exactly there and which mistakes beginners make.
Test ideas instead of believing them
Start with the free mini lesson. For learning only, not investment advice.
Start the free mini lesson