How to backtest a strategy without fooling yourself

Backtesting means testing your strategy on past data before risking money. Done well, it saves you months of losses. Done badly, it gives you false confidence that the live market makes you pay for.

1. Write the rules before looking at the chart

Entry, stop, target, management and when NOT to trade. If a rule depends on "it looks like", it can't be tested. The more concrete they are, the more reliable the result.

2. Trade candle by candle, without seeing the future

The most common mistake is looking at the whole chart and "seeing" the good entries in hindsight. Move forward candle by candle and decide with the same information you'd have live.

Katanith's backtesting replays the market candle by candle with one-minute data, so you can't see what's coming.

3. Enough data, in different conditions

• At least 100 trades before drawing conclusions; with 20, any result can be luck.
• Several years and different markets: trend, range, high and low volatility.
• Keep part of the data untouched to check at the end (out of sample).

4. Include costs and news

Spread, commissions and slippage turn many "winning" strategies into losing ones, especially those with tight stops. And high-impact news moves price in ways your strategy may not account for.

In Katanith you can see historical news on the chart and pause the replay right before high-impact events.

5. Look at the right metrics

• Expectancy in R: how much you make per trade on average.
• Profit factor and max drawdown: if you win, how much you suffer along the way.
• Losing streaks: could you handle the worst one mentally?

6. Then, on demo or with minimal risk

Backtesting doesn't replace live trading: confirm it for a few weeks with minimal risk and compare the results. If they match, you have a real edge.