FARO · Calculator
How many trades do you need to know if your strategy works?
- It assumes each trade is independent of the others.
- It assumes the same reward/risk ratio on every trade: the average you enter.
- It does not include commissions, costs or slippage.
- A backtest is not a forward test: if your data comes from a backtest, the result describes a simulated past, not real trades.
- If you tried many variants and kept the best one, this calculation does not correct for it: some will come out well by chance.
- Past performance does not guarantee future results.
Enter a whole number of trades, 1 or more.
Winning trades cannot be more than closed trades.
The reward/risk ratio must be a number greater than 0.
Your win rate: ( of )
95% Wilson interval: to
Break-even win rate at R = :
With this data, your win rate is distinguishable from the break-even point.
The lower bound of the interval () is above the break-even win rate (). This does not say the strategy works or will be profitable: it says that, with the data you entered and the assumptions above, this win rate is hard to explain by chance alone.
With this data, your result could be chance.
The lower bound of the interval () does not clear the break-even win rate (): with the data you entered, a result like this is within what chance produces.
At the same win rate, you would need about trades in total — more — for the lower bound to clear break-even.
At this win rate, no number of trades would be enough: your win rate is not above break-even.
At the same win rate, you would need more than a million trades.
Calculated with the data you entered. Nothing leaves your browser.
A track record’s win rate is an estimate: other trades from the same strategy would give a different one. The 95% Wilson interval gives the range where the real win rate can reasonably be, and it narrows as the number of trades grows. It is the same interval FARO’s methodology uses for every analyst and every automated system.
With a reward/risk ratio R, a strategy breaks even — neither wins nor loses, before costs — when it wins one in every 1 + R trades: 50% at R = 1, 33% at R = 2. If the lower bound of the interval is above that point, chance is a poor explanation for the result; if not, it can explain it.
With 20 trades and a 60% win rate, the interval runs from roughly 39% to 78%: it fits a losing strategy and a very good one. With 200 trades and the same 60%, it runs from 53% to 67%. The percentage is the same; what changes is how much it can tell you.
And the figure only counts if the trades are all of them, winners and losers, without choosing which ones to show. For that, see how to tell if a trader is reliable and, if you design systems, how to seal your parameters before a forward test.
If you want your track record to count, seal it on FARO: as an analyst or with your automated system.