Backtest
How each strategy performed on historical data, and how to read those numbers without fooling yourself.
What it shows
A backtest replays a strategy's rules over past market data. The page shows our recorded runs of the strategies as they are shipped, with fees and slippage deducted. These are the same numbers as on the home page.
- Strategy
- All, Day Trading, Swing or Spot Accumulate
- Starting balance
- Converts results to dollars. Every other figure does not depend on it.
- Period
- 1, 2 or 3 years. Swing and spot use their full history, as a short window is noise for them.
You do not run backtests yourself: the page shows runs we record each time a strategy changes.
The numbers
| Number | What it tells you |
|---|---|
| Growth | How many times the balance multiplied over the period |
| Max drawdown | The deepest fall from a high. What you would have had to sit through. |
| Win rate | The share of trades closed in profit. Means little by itself. |
| Profit factor | Gross profit divided by gross loss. Over 1 makes money. |
| Average win and loss | The typical size of each, as a share of the balance |
| Trades | The sample size. More trades mean a more reliable number. |
- 1.The high before the fall
- 2.The lowest point. The gap from the high is the drawdown.
- 3.The balance recovers to a new high
Reading it honestly
- Look at the drawdown first. Growth is the number that gets the attention, but the drawdown decides whether you would have stayed in.
- Win rate needs the sizes beside it. A 95% win rate with small wins and larger losses can earn as much as 45% with the reverse.
- Do not expect the total return. Compounded backtests assume every order fills at any size. Real depth and fills limit this as an account grows.
Backtest and live
To see how real accounts fare, look at the public Leaderboard, when it is open, and the live trade feed on the home page. To see how a strategy behaves in today's market with no risk, use paper trading.