Every profitable trader has one thing in common: they tested their strategy before risking real money on it.
Most retail traders skip this step. They watch a YouTube video, see a "strategy" with a few cherry-picked wins, and load up real capital. Then they blow their account and blame the market.
Here's how to actually backtest — properly, honestly, and in a way that tells you if your edge is real.
What "backtesting" actually means
Backtesting means manually (or automatically) running your strategy through historical price data to see how it would have performed.
Done right, it tells you:
- Whether your strategy is profitable at all
- Your average win, loss, and R:R
- Your win rate over a meaningful sample
- Your maximum drawdown (worst losing streak)
- What market conditions help or hurt your strategy
Done wrong, backtesting just convinces you a losing strategy will work and gets you into bigger trouble.
The 7-step backtesting process
Step 1 — Write your strategy down. ALL of it.
If you can't write it down, you don't have a strategy — you have a feeling.
Be specific. "Buy when it looks bullish" is not a strategy. "Buy when the 50 EMA crosses above the 200 EMA on the 4H chart, with RSI above 50, with stop loss at the swing low and TP at 2R" is a strategy.
Your written strategy needs:
- Entry condition — exactly what triggers a buy/sell
- Stop loss rule — exactly where the stop goes
- Take profit rule — exactly where you exit (or how you trail)
- Position size rule — how much you risk (usually 1% of account)
- Filter conditions — what setups you SKIP (time of day, news days, trend filter, etc.)
Step 2 — Pick a market and timeframe
Pick ONE market (e.g., EURUSD, BTC, SPY) and ONE timeframe (e.g., 1H, 4H, daily). Don't mix.
Why? Strategies that work on EURUSD might fail on Gold. What works on 5-minute charts might fail on daily. You need to know what your edge actually is, not blend together five different things.
Step 3 — Choose a long sample period
Minimum: 6 months of price action. Better: 1-2 years. Best: include both trending and ranging conditions.
Why? A strategy that "worked" through May-July 2025 might have been pure luck. Test through volatile periods, ranging periods, news events, and quiet periods. Real strategies hold up across regimes.
Step 4 — Find setups bar by bar
Scroll through your chart from the start of your sample period. When your strategy says "enter," mark that bar. Use TradingView's replay feature so you don't see future bars while testing.
Step 5 — Log every trade
For each setup, log:
- Date
- Entry price
- Stop price
- Target price
- Final outcome (win at TP, loss at SL, or breakeven if you got stopped out)
- R:R achieved
- Notes — what worked, what surprised you
This is exactly what TradeLens does — you can enter the R:R outcome (like +2 or -1) and it tracks everything for you.
Step 6 — Aim for at least 30 trades (50+ is better)
Fewer than 30 trades and your sample is too small. You could have a 90% win rate over 10 trades and still be unprofitable in the long run because you got lucky.
50-100 trades gives you statistically meaningful data. Anything below 20 is just noise.
Step 7 — Calculate your metrics
After logging your sample, calculate:
| Metric | What to look for |
|---|---|
| Win Rate | Realistic for your R:R (see our win rate guide) |
| Profit Factor | 1.5+ minimum to trade live |
| Expectancy | Positive, ideally +0.3R or higher |
| Max Drawdown | Could you handle this with real money? |
| Average R:R | 1.5+ minimum |
The brutal honesty check
After backtesting, ask yourself:
- Would I take EVERY one of these setups live? If not, you're cherry-picking and your live performance will be worse.
- Could I survive the max drawdown emotionally? If max DD was 25% of your test account, you WILL panic when real money hits that level.
- Did I see at least 30 trades? If not, keep going.
- Does it work in different conditions? Strategies that only work in trending markets WILL fail when markets range.
Common backtesting mistakes
- Hindsight bias — Looking at known outcomes and rationalizing what you "would" have done
- Cherry-picking — Counting only the trades you wish you'd taken
- Ignoring slippage and spreads — Real trading has costs. Subtract 0.5-1R from each winner to account for it
- Curve-fitting — Tweaking your strategy until it perfectly fits past data, but won't work on future data
- Too small a sample — 10 trades is meaningless
- Only one market regime — Test through both trending and ranging periods
What to do AFTER backtesting
If your strategy passed:
- Forward-test it on a demo account for 30+ trades
- Then go live with the SMALLEST possible position size
- After 50 live trades, compare your live stats to your backtest stats. If they match within 20%, scale up
If they don't match — your backtest had a flaw. Find it before risking more.
TradeLens makes backtesting easy: log each trade as you go through historical data, see your live stats update, export your full backtest as a PDF report. No spreadsheet needed.