How to check any EA's real risk per trade (a 6-step Strategy Tester routine)
Almost every EA page talks about profit. Far fewer tell you, in money, how much a single trade can lose. Yet that number decides whether an account survives a bad month. Here is a simple routine you can run on any Expert Advisor, including mine, using only the MetaTrader 5 Strategy Tester.
1. Find out whether every trade has a stop loss
Run a backtest with the default settings, open the Backtest report and switch to the list of orders and deals. Check that each opening order carries a stop loss price from the first moment. If positions open without a stop and only get one later — or never — the risk per trade is not defined at all. It depends on how far price happens to move before the EA decides to act.
2. Count how many positions can be open at once
Risk per trade only means something if there is one trade. A grid or martingale EA may open five, ten or twenty positions in the same direction. In that case the real risk is the sum of all of them at the moment the last one is added. In the tester report, sort the deals by time and look for several "in" entries on the same symbol with no "out" deal in between — and watch how high the deposit load climbs on the graph.
3. Compare the worst losing trade with the balance
The report shows Largest loss trade. Divide it by the balance at the time of that trade. If an EA advertises "1% risk" but its largest loss is 4–5% of the balance, something is different from what the description says: missing stops, slippage through gaps, or position sizing that does not account for the instrument's contract size.
4. Check the position sizing on gold specifically
Gold is where sizing mistakes show up most. Contract sizes differ between brokers (100 oz per lot is common, but not universal), and the tick value reported by some symbols is not always reliable for converting a stop distance into money. A robust EA calculates the money at risk directly — in MQL5, for example, with OrderCalcProfit() between the entry and the stop price — and then scales the lot. Simple test: run the same EA on two brokers' gold symbols, or change the deposit currency, and see whether the loss per trade stays close to the stated percentage.
5. Look at losing streaks, not just single losses
Even with a fixed 1% risk, ten losses in a row is roughly a 10% drawdown. The report shows Maximal consecutive losses. Multiply that count by the risk per trade and you get a realistic "bad run" to prepare for. If you would not be comfortable living through that number, lower the risk input before going live.
6. Watch equity, not only balance
A smooth balance line can hide deep floating losses. Turn on the equity line in the tester graph. Large gaps between balance and equity usually mean positions are held through big adverse moves — the typical signature of averaging down.
How Glassbox Gold Breakout handles this
I built my EA around these checks: one position at a time, a fixed stop loss and take profit on every order, lots sized from the money at risk with OrderCalcProfit() , and a daily loss limit. In my own tests at 1% risk, the 2015–2021 period had a maximum drawdown of 12.0% (profit factor 1.47, +33% net), and the unseen 2022–2023 period had a profit factor of 1.26. To be clear: simply holding gold did better than the EA over 2015–2021. The point of the design is controlled, predictable risk — not beating gold. Past results do not guarantee future results.
Whatever EA you use, run these six checks yourself before trusting it with real money.
If you want to try the checklist on a real example, you can download the free demo of Glassbox Gold Breakout and run it in your own Strategy Tester.
