What Is a Realistic Win Rate for Trading Signals?

Sustained win rates for systematic strategies sit between roughly 40% and 70%, and advertised figures typically run 10–20 percentage points above independently measured performance. What matters is break-even, which depends on reward-to-risk: 50% at 1:1, 33% at 2:1, 25% at 3:1. A 90% win rate is usually manufactured by placing targets very close and stops very far, dropping trades that went nowhere, or resolving ambiguous outcomes favourably.

The honest answer is that there is no single good win rate, because the number is meaningless without the reward-to-risk ratio beside it. But people asking this question usually want to know one specific thing: is the 90% I keep seeing advertised real?

No. Here is what the numbers actually look like.

The realistic range

For a systematic strategy holding a sensible reward-to-risk, sustained win rates in the 40% to 70% band are normal, and anywhere in that range can be highly profitable depending on the ratio. Where independent trackers have compared advertised rates against measured performance, the advertised figures have run roughly 10 to 20 percentage points too high.

So a service advertising 92% is very likely operating somewhere in the seventies at best — and that is assuming the underlying record is complete.

What win rate you actually need

Break-even depends entirely on the ratio between your average winner and your average loser:

At the fixed 2:1 we use, break-even is a 33% hit rate. That is why a 40% win rate is a genuinely good result here, and why quoting a win rate without the ratio tells you nothing at all. The arithmetic is worked through in why a 40% win rate can beat a 90% one.

How a 90% win rate gets manufactured

None of these require inventing a single number:

• Move the target very close and the stop very far. A target 0.3% away and a stop 6% away will win the overwhelming majority of the time — until one loss erases twenty wins. The win rate is real. The strategy still loses money.
• Drop the trades that went nowhere. Signals that reached neither target nor stop are quietly excluded, removing every mediocre outcome from the denominator.
• Resolve ambiguity favourably. When target and stop both fall inside one candle, the order is unknowable — scoring those as wins adds several points for free.
• Close losers “manually”. If a losing position is exited by judgement rather than a predetermined stop, it never has to be recorded as a stop-out.

What to look at instead

• Expectancy — the average signed return per signal across every resolved outcome. Positive means the average signal gains. This is the number that decides whether a strategy makes money.
• Profit factor — gross profit divided by gross loss. Above 1.0 means winners outweighed losers in total. Harder to flatter than a win rate because it is driven by size, not count.
• Average winner and average loser — the two numbers that make a win rate interpretable.
• Sample size and window — 20 decided trades is a floor for quoting anything; a few hundred before you weight it heavily.

What ours looks like

Our own numbers are published live, computed rather than claimed, and include every loss. At a fixed 2:1 reward-to-risk, we would rather show a mid-thirties or forties win rate that is real than a ninety that is not. You can read the current figures — and check the arithmetic yourself against the raw JSON — on the track record, and see exactly how each outcome is decided in the methodology.

If you want to apply the same scrutiny to anyone else, the checklist is in how to verify a signal provider's win rate.

Educational content, not financial advice. Trading carries risk, including loss of capital.