Why a 95% Win Rate Is a Warning Sign, Not a Selling Point

A 95%+ win rate is the single most effective number in EA marketing, and it is also the most reliable predictor of a blown account. That is not a coincidence. This article explains the mechanism, so you can recognise it before you pay for it.

The recipe has three ingredients. First, a tiny take-profit — a few pips or dollars per position — so almost every trade closes green quickly. Second, no hard stop loss: a losing position is never realised, it is held, averaged into, or hedged until price comes back. Third, lot sizes that grow with the balance, so the position pile gets heavier exactly as the account grows. Each closed trade is a small real win; the risk lives entirely in the open positions, which the win-rate statistic does not count.

With this structure the balance curve is a smooth diagonal line, because balance only records closed trades. The equity curve tells the truth: it dips hard every time price trends away from the position pile, and one day a trend runs far enough that the floating loss exceeds the account. There is no gradual warning. The system does not degrade — it works, works, works, and then removes the account in one event. Ninety-five small wins do not survive one uncontrolled loss.

You will rarely see this failure in a published backtest, because the window is chosen so it cannot appear. A test that starts after a crash and ends before the next one shows an unbroken curve — the number is real, the impression is false. Whenever a report covers a suspiciously specific period, ask what happened just outside it. A strategy with no hard stop must be tested through the worst regimes available, precisely because its failure mode is rare and total.

A trend-following system with a hard stop on every position typically wins well under 50% of its trades. Many small controlled losses, a few large wins. The balance curve is ugly — flat stretches, visible losing streaks — and the equity curve looks almost the same as the balance curve, because nothing is hidden in floating positions. That is the trade-off: a worse-looking statistic for a bounded worst case. We publish sub-50% win rates on our own trend systems and consider that a feature, not an embarrassment.

Ignore the win rate. Ask instead: does every position carry a hard stop loss? Is maximum drawdown reported on equity, not balance? Does the backtest include crisis periods? If the answers are no, no, and no, the 95% is not evidence of skill — it is the visible half of a structure whose other half is your account.

An example of the opposite design — hard stops, sub-50% win rate, multi-market diversification: https://www.mql5.com/en/market/product/182751

All products and free utilities: https://www.mql5.com/en/users/app.develop.sk/seller