Home › Comparisons

Backtesting vs forward testing: test on the past, then prove it live

A backtest asks, "Would my rules have worked on old charts?" A forward test asks, "Do they still work when I don't know what happens next?" You need both answers before real money is on the line.

Comparison · 6-minute read · Updated 2026-10-21

The short answer

Backtesting means testing your rules on past price data. You step through old charts and record what your rules would have done. It's fast: you can test months in an afternoon.

Forward testing means testing the same rules in real time, as new bars form. Usually that's Paper trading first, then very small real size.

A backtest shows whether an idea is worth more time. A forward test shows whether you can follow it when the future is truly unknown.

Side by side

Backtesting, in plain words

Pick a market and a Timeframe. Scroll back. Hide the future. Step forward bar by bar and take every trade your rules allow. Log each one.

On TradingView, the bar replay tool does this: it hides everything to the right and shows one new bar at a time. See TradingView bar replay.

The good parts:

• Lots of practice, fast. Many trades in one sitting.
• No money at risk.
• Finds broken ideas early. If a rule fails badly on old charts, you saved yourself a lot.

The limits:

• Look-ahead. If you can see what happened next, your "test" isn't a test. Never peek.
• Overfitting. Tweak the rules until they fit the past perfectly, and they often stop fitting the future.
• Easy fills. A backtest often assumes you got your price. Real fills can slip. See Slippage.
• No pressure. Losing on an old chart feels nothing like losing now.

Forward testing, in plain words

Same rules, but now the next bar really is unknown. You take trades as they set up, on paper or at very small size, and log them the same way.

The good parts:

• Honest. No way to peek.
• Real conditions. Spreads, news spikes and slow days all show up.
• Tests you. Will you take the setup after two losses? Will you skip a trade that isn't there?

The limits:

• Slow. It can take weeks to collect enough trades.
• Paper isn't real money. Feelings are still lighter than live. See paper trading vs live trading.
• Few trades can mislead. Ten trades say very little. Judge over many.

The order we suggest

1. Write the rules first. Entry, stop, target, size and a daily limit. See how to write a trading plan.
2. Backtest about 100 trades on past charts with replay. Log every one, including losses and skipped trades.
3. Read the results. Win rate, average win and loss in units of risk, the longest losing run. The expectancy calculator puts these together.
4. Forward test on paper with the same rules. Keep logging.
5. Compare. Are the forward results roughly in line with the backtest? If they're far worse, find out why before going further.
6. Go live at small size. Size from the stop with the position size calculator.

Results in any test are hypothetical. Past results, real or tested, don't promise future ones. Losses are always possible.

Which fits you?

You don't choose one. You do them in order.

• Have a new idea? Backtest first. It's the cheapest filter.
• Have a backtest you like? Forward test before real money.
• Already trading live? Keep logging. Live trading is a forward test that never ends. See how to keep a trading journal.

Common mistakes

• Peeking during replay. The most common, and the most damaging.
• Changing the rules after every loss. That's fitting the past, not testing a plan.
• Testing 15 trades and calling it proven. Aim for many more.
• Trusting strategy-tester results blindly. Coded backtests can be fooled by look-ahead, unrealistic fills and costs left out. See Pine Script strategy backtest.
• Skipping the forward test. A great backtest that you can't follow live isn't much use.

Questions