You can backtest any trading strategy for free, without writing code, using chart replay: play historical charts forward candle by candle with the future hidden, take your strategy's trades, and log the results. Here is the exact process.

Step 1: Write the strategy as rules

If the strategy isn't written down, you can't test it — you'll unconsciously bend it trade by trade. Define, on paper:

• Setup: what must be true before you're interested (e.g. price above the 200 EMA, in an uptrend)
• Entry trigger: the exact event that gets you in (e.g. bullish engulfing candle at support)
• Stop-loss: where the idea is proven wrong (e.g. below the signal candle's low)
• Exit: target, trailing rule, or condition (e.g. 2R, or opposite signal)
• Filters: when you don't trade (e.g. no entries in chop, first 15 minutes, major news)

Need a starting point? Every strategy in our strategies library is already written in this format.

Step 2: Choose market and timeframe

Test on the timeframe you actually intend to trade. A breakout strategy behaves completely differently on 5-minute vs. daily candles. Pick one combination — e.g. BTC/USDT, 15-minute — and keep it constant for the whole test.

Step 3: Replay with a hidden future

Open the free chart replay tool, which starts you at a random, hidden point in real market history. This matters more than any other detail: if you can see what comes next (or recognize the date), your results are fiction. Advance candle by candle and treat it as live.

Step 4: Take every valid trade — and only valid trades

The discipline rules of a good backtest:

1. If the setup appears, you must take it (no skipping trades that "feel" bad — that's a different, untested strategy).
2. If the setup isn't there, you must not trade (boredom trades poison the data).
3. Place the stop and target when you enter, then let the replay decide.

Step 5: Log every trade

Record per trade: date/session, direction, entry, stop, exit, result in R (risk multiples), and one line about the context. The Replay Trader web app tracks your P&L and win rate per session automatically; the mobile apps include a full trade journal.

Step 6: Read the results at 50 trades

Now compute the numbers that matter:

• Expectancy = (win% × average win) − (loss% × average loss). Positive after costs = viable.
• Max losing streak — if the strategy hits 7 straight losses in a backtest, expect it live. Could you keep executing?
• When it fails — look at your losers as a group. Most strategies lose in a specific regime (e.g. trend strategies lose in ranges). That insight becomes a new filter → new backtest.

A realistic outcome: your first backtest shows the strategy roughly breaks even, you spot that most losses share a context, you add one filter, and the second 50-trade backtest is profitable. That iteration loop — not finding a magic strategy — is what backtesting is for.

Common backtesting mistakes

• Testing with visible future (scrolling a normal chart) — always use replay.
• Changing rules mid-test — finish the sample, then revise.
• Testing 10 strategies shallowly instead of one properly.
• Ignoring fees — subtract realistic costs per trade, especially intraday.
• Declaring victory at 10 trades — variance dominates small samples.

FAQ

How can I backtest a strategy for free?

Use manual backtesting with a free chart replay tool: Replay Trader replays real Binance history candle by candle in the browser with paper trading built in — no account or download required.

How long does backtesting take?

With replay, a 50-trade sample on an intraday strategy typically takes 3–5 evenings. The same sample on live demo trading would take months.

Can I backtest without coding?

Yes — that's exactly what manual/replay backtesting is. Coding is only needed for automated backtests of fully mechanical strategies. See manual backtesting for when each approach fits.

What win rate should a strategy have?

There's no minimum — it depends on the win/loss size ratio. 40% winning is great when winners average 2–3R. Judge strategies on expectancy, never win rate alone.