A backtest replays an exact rule set against historical candles to see what would have happened. Done honestly, it tells you whether a strategy has a measurable edge before you risk a cent. Done carelessly, it tells you a flattering lie. This guide is the honest version, the same execution assumptions Backtestify enforces on every run.
How to backtest
a trading strategy
6Steps
5Traps to avoid
- 1
Define exact rules, no ambiguity
Write down every rule before looking at any result: the entry conditions, the stop loss, the take profit or exit signal, position size, and the market and timeframe. If a rule is vague ("buy on momentum"), the backtest cannot test it. Concrete: "Buy when RSI(14) crosses above 30 on the daily close, risk 1% per trade, stop at 2× ATR(14)."
- 2
Choose the market, timeframe and budget
Pick the instrument (for example BTC/USDT spot), the candle timeframe (1 minute to 1 week), the date window, and the starting budget. The timeframe changes everything, the same rules can win on 4h and lose on 15m.
- 3
Replay candles with honest fills
Replay the rules bar by bar on historical candles. The execution assumptions matter more than the rules: a signal only counts on a completed candle, the fill happens at the next candle's open, and if one candle touches both the stop and the target, count it as the stop. Anything else invents fills the market never offered.
- 4
Charge real costs on both sides
Deduct commission and slippage on entry and on exit. A strategy with a 0.2% average edge per trade is not a strategy once it pays 0.15% per side. Skipping costs is the most common way backtests lie.
- 5
Validate on unseen data
If you searched or tuned parameters on part of the history, re-measure the final rules on a withheld window that was never used to pick them. A result only on the data that produced it is a description of the past, not evidence of an edge.
- 6
Read the risk report, not just the return
Judge the test on win rate, profit factor, max drawdown, average result per trade, and the number of trades. Compare the return against simply buying and holding the same asset over the same window. Small samples prove nothing, a handful of lucky trades is not a strategy.
Why most backtests lie
Lookahead bias
Using information from the same candle to trade that candle. Signal confirms on the close; the fill is the next open.
Survivorship of exits
Assuming the target fills when the stop was nearer the open on the same bar. Take the nearer level first, as TradingView does, and count the trade as ambiguous.
Zero-cost fantasy
Ignoring commission and slippage. Charge both sides, every trade.
Overfitting
Tuning parameters until the curve looks good, then reporting that same data. Always keep an unseen window.
Small samples
Drawing conclusions from a dozen trades. Edge needs a sample large enough to be meaningful.
Do it in one sentence
Backtestify does this whole workflow for you. Describe the strategy in plain English, it is turned into an exact rule set, replayed on real candles with signal confirmation on the close, fills at the next open, and costs on both sides. You get a verdict with win rate, drawdown, profit factor, buy-and-hold beside your return, and a TradingView strategy you can verify.
Backtestify is a research tool, not financial advice. Every backtest is a hypothetical simulation of past data, past performance never guarantees future results.