What RiskPercent Actually Controls in an EA — and Why the Same Number Means Different Drawdown on Every Market

RiskPercent looks like the simplest input in any EA: a single number that says how much you risk. It is also the most misunderstood one. Buyers routinely set it once, apply it everywhere, and are then surprised when the same "2%" feels calm on one market and violent on another. Neither the EA nor the number is lying — the number just does not mean what most people think it means.

In a properly built EA, RiskPercent defines the loss you accept if this one trade hits its stop loss. The EA takes that percentage of your balance, divides it by the stop distance converted into money per lot, and gets the position size. That is all. It is a per-trade sizing rule, computed from the stop — which is why an EA with no hard stop loss cannot honestly offer a RiskPercent input at all, whatever its settings page claims.

The same percentage produces very different behaviour across markets because everything around it changes: how wide the stop is relative to typical daily movement, how often the strategy trades, how frequently it loses in a row, and whether several positions can be open at once. A market that gaps or spikes can also slip beyond the stop, so the realised loss on one trade can exceed the configured percentage. RiskPercent caps the intended loss per trade; it caps nothing else.

This is the core confusion. Drawdown is what a sequence of losses does to the account, and losing streaks are a statistical certainty for every strategy, including profitable ones. Ten consecutive 2% losses are roughly an 18% hole. Add correlated positions — two gold-based trades open at once are close to one double-sized trade — and the same input produces double the streak damage. This is why identical RiskPercent values yield mild drawdown on one system and frightening drawdown on another: the streak profile and the position overlap differ, not the arithmetic.

Work backwards from the drawdown you can genuinely tolerate, not forwards from the profit you want. Look at the worst losing streak in a long backtest of the specific EA on the specific market, assume the future will be somewhat worse, and pick the RiskPercent that keeps that streak inside your limit. Then consider halving it for the first months of live trading — live fills are never better than tested ones, and the goal of the first months is to survive them, not to maximise them. Raising risk later is always available; recovering a halved account is not.

Our defaults are set from measured worst streaks, not from what makes a marketing screenshot grow fastest, and multi-market products size each leg independently so one instrument's volatility does not silently scale another's risk. Example of a portfolio EA built around this sizing logic: https://www.mql5.com/en/market/product/182751

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