Why a 40% Win Rate Can Beat a 90% One
A win rate is uninterpretable on its own. At a fixed 2:1 reward-to-risk, break-even sits at a 33% hit rate, so a sustained 40% is a real edge. Very high advertised win rates usually conceal a poor reward-to-risk, where one loss erases dozens of wins.
Almost every signal service leads with its win rate, because it is the number people ask for. It is also, on its own, close to meaningless. A service winning 80% of the time can lose money steadily, and one winning 38% of the time can grow an account. What decides the outcome is how much you make when you are right versus how much you lose when you are wrong.
The arithmetic
Expectancy per trade is the only number that matters. It is the average win multiplied by the win rate, minus the average loss multiplied by the loss rate:
Expectancy = (Win% × Average win) − (Loss% × Average loss)
The same win rate means completely different things at different reward-to-risk ratios. Risking one unit each time, here is what a few hit rates produce at 1:1 (target as far from entry as the stop) and at 2:1 (target twice as far):
| Win rate | Expectancy at 1:1 | Expectancy at 2:1 |
|---|---|---|
| 30% | −0.40 | −0.10 |
| 33.3% | −0.33 | 0.00 (break-even) |
| 40% | −0.20 | +0.20 |
| 50% | 0.00 (break-even) | +0.50 |
| 55% | +0.10 | +0.65 |
| 60% | +0.20 | +0.80 |
Break-even is a third at 2:1 and a half at 1:1. A 40% win rate is a genuinely good result at 2:1 and a losing one at 1:1 — which is why a service quoting any win rate without telling you its reward-to-risk has told you nothing at all.
Every signal we publish is set at a fixed 1:1 reward-to-risk — the target sits the same distance from entry as the stop, so it needs to win more than half the time. We moved from 2:1 to 1:1 in October 2026: at 2:1 our engine lost roughly two trades for every one it won, and on two years of out-of-sample testing the 1:1 target was the only one that stayed profitable in both halves of the data. The trade-off is a smaller average winner.
Why very high win rates are a warning sign
You can manufacture almost any win rate by moving the target closer and the stop further away. Set a target 0.2% from entry and a stop 5% away and you will win the overwhelming majority of trades — until the one loss erases forty wins. High advertised win rates usually mean a poor reward-to-risk hiding behind them.
- Ask for the reward-to-risk alongside the win rate. Neither number means anything alone.
- Ask for the average loser, not just the average winner. That is where a bad strategy hides.
- Ask whether losses are published at all. A track record showing only winners is marketing, not evidence.
How we report it
Our published track record shows the win rate, the average winner, the average loser and the resulting reward-to-risk together, because the first number cannot be interpreted without the other three. Losses and expired signals are included in the denominator.
We also refuse to publish a win rate until at least 20 signals have resolved. Below that a single outcome swings the figure by five percentage points or more, which makes it a decoration rather than a measurement. Until then we show the raw counts instead. The full scoring rules are in the methodology.
The short version: a win rate is only interpretable next to a reward-to-risk. At the fixed 1:1 we use, break-even is a 50% hit rate (a 33% one at 2:1), so the win rate has to be read against that line — and a service advertising 90% without publishing its average loss is telling you less than it appears to.
Frequently asked questions
- Is a 40% win rate good for trading signals?
- It depends on the reward-to-risk. At 2:1 it is genuinely good: break-even sits at a 33% hit rate because each win pays twice what each loss costs, so a sustained 40% produces about +0.20 per unit risked before costs. At 1:1 — what our signals use — break-even is 50%, so 40% would be a losing system.
- What is expectancy in trading?
- Expectancy is the average result per trade: (win rate × average win) − (loss rate × average loss). It is the only number that determines whether a strategy makes money over time, which is why a win rate quoted without an average loss tells you almost nothing.
- Why is a 90% win rate a warning sign?
- A very high advertised win rate usually conceals a poor reward-to-risk — a wide stop and a tiny target, so that a single loss erases dozens of wins. A service quoting 90% without publishing its average loss has told you nothing about whether it is profitable.
- What reward-to-risk do SmartInvest AI signals use?
- A fixed 1:1 — the target is always the same distance from the entry as the stop, so break-even is a 50% win rate. It was 2:1 until October 2026; we lowered it because at 2:1 the engine lost about two trades for every win and its edge did not hold up across the full test period. Holding the ratio constant is what makes the published win rate interpretable.
Related reading
Educational content, not financial advice. Trading carries risk, including loss of capital.