The trap everyone falls for
“My strategy wins 87% of the time.” It sounds unbeatable. It is not. A high win rate is the single most misleading number in trading, and we can prove it with our own data.
We took one strategy entry — a Bollinger Band mean-reversion signal on altcoins — and ran it with seven different exits, changing nothing else. Same coins, same fees, same entry. Only the exit changed.
The data
| Exit method | Win rate | Avg win | Avg loss | Expectancy (after fees) |
|---|---|---|---|---|
| Tight TP 1% / SL 10% | 87% | +1.0% | −9.9% | −0.93% |
| TP 2% / SL 8% | 75% | +2.0% | −7.9% | −1.04% |
| Symmetric TP 5% / SL 5% | 46% | +5.0% | −5.0% | −0.97% |
| Wide TP 10% / SL 3% | 21% | +9.6% | −3.0% | −0.91% |
| Trailing stop 3% | 32% | +2.4% | −1.6% | −0.90% |
| Trailing stop 8% | 32% | +7.2% | −4.9% | −1.59% |
| Trailing stop 15% | 37% | +12.5% | −9.1% | −1.62% |
Look at the first row. An 87% win rate that loses money. You win 1% eighty-seven times, then lose 10% thirteen times. The math: 0.87 × 1.0% − 0.13 × 9.9% = −0.42%, and fees make it worse.
The conservation law
Notice that the win rate ranges from 21% to 87% — you can set it almost anywhere just by choosing the exit. But expectancy stays stubbornly negative in every single case.
This is the law:
The exit method only redistributes a fixed expectancy between win rate and win size. It cannot create expectancy.
- Want a high win rate? Take small profits fast (tight TP). But your losses grow large — picking up pennies in front of a steamroller.
- Want big winners? Use a wide target or trailing stop. But your win rate collapses, because normal noise stops you out.
Win rate and reward-to-risk are two ends of a seesaw. Push one up, the other goes down. The product — your expectancy — is fixed by the entry, not the exit.
”But I’ll just add a trailing stop”
This is the most common reply, so we tested it specifically (rows 5–7). A trailing stop flips the distribution — low win rate (32%), bigger winners — but expectancy is still −0.90% or worse. The trailing stop is just another point on the same seesaw, and it adds whipsaw cost on top.
What this actually means
If the entry has no predictive edge, no exit can save it. And for retail directional trading on public data, the entry edge — after fees — is essentially zero. We have tested this exhaustively across indicators, timeframes, coins, and combinations. The honest result: direction is not predictable enough to overcome cost.
So where does a real edge come from?
- The entry must have genuine predictive power — which, for retail direction-betting, it does not.
- Or the edge is non-directional — harvesting a risk premium (like selling volatility, where the seller is paid a premium that exists regardless of direction). That is being the house instead of the gambler.
The takeaway
Stop optimizing exits to chase a pretty win rate. Expectancy is set at the entry. Test your own strategy on PRUVIQ’s simulator and look at the expectancy after fees — not the win rate. If the entry has no edge, the prettiest 87% win rate in the world still bleeds.