GC
Goldman Academy
Academy › Level 2 · Intermediate

Backtesting a strategy honestly

Intermediate⏱ 10 min readUpdated 31 Oct 2026
GC
Written by
Goldman Academy
F
Reviewed by
Faisal
In this lesson you'll learn
  • What backtesting can and can't prove
  • A manual backtesting method anyone can use
  • The traps that make backtests lie

Why test the past?

A backtest applies your exact rules to past data to see what would have happened. It can't promise the future, but it can quickly show that a strategy has no edge, which saves real money.

A manual method

  1. Write the rules so precisely that there's no judgement left.
  2. Scroll the chart back months; move forward candle by candle (TradingView's replay mode helps).
  3. Record every valid setup in a spreadsheet: date, entry, stop, target, result in R.
  4. Include costs: subtract the spread and commission in R.
  5. Collect at least 100 trades across different conditions (trends, ranges, high and low volatility).
  6. Calculate win rate, average R, expectancy and the worst losing streak.

Try it

🎯 Expectancy and break-even calculator
Expectancy per trade
In money, per trade
Break-even win rate

The traps

  • Hindsight: seeing the result before deciding. Hide the future candles.
  • Cherry-picking: skipping setups that "look bad" in hindsight.
  • Curve fitting: tweaking rules until the past looks perfect. Keep rules simple; test on a separate period you didn't use to build them.
  • Ignoring costs and slippage, especially on short timeframes.
  • Too few trades: 20 trades prove almost nothing.

Then forward-test

A good backtest earns a forward test on demo or small size, not a full-size live account.

Common mistakes
  • Backtesting 15 trades and declaring a winner.
  • Changing rules after each losing test trade.
  • Leaving out costs.

Check yourself

1. What can a backtest reliably show?
2. Curve fitting means…
3. A reasonable minimum sample is about…

Key takeaways
  • Backtests are best at disproving bad ideas.
  • Test 100+ trades, with costs, across conditions.
  • Avoid hindsight, cherry-picking and curve fitting; then forward-test.
← Correlations: why three trades can be oneForward testing and journaling: the habit that improves everything →
Education, not advice. Goldman Academy teaches how markets and tools work. Nothing here is a recommendation to buy or sell, and no lesson promises profits. Trading leveraged products carries a high risk of losing money; only trade money you can afford to lose. · Written by Goldman Academy, reviewed by Faisal