An expert advisor follows its rules on every trade, but it doesn't decide how much of your account those rules are allowed to lose. That part is your job. This guide sets out a practical risk plan for automated trading: how much to risk per trade, how to size positions, which loss limits to set, how to run several EAs on one account and when to switch one off.
Start with risk per trade
Risk per trade is the amount you lose if a trade hits its stop loss, written as a percentage of your account balance. It is the one setting that decides how fast an EA can grow or damage an account, and it matters more than the entry logic or the win rate.
Many retail traders keep risk between 0.5% and 2% of balance per trade. On a $10,000 account, 1% means a losing trade costs about $100. That sounds small, but EAs take many trades, and losses arrive in streaks. The useful question isn't “what can I lose on one trade?” but “what will a run of losses do to the account?”
The table shows how far an account falls after consecutive losses when each trade risks a fixed share of the current balance.
| Risk per trade | 5 losses | 10 losses | 15 losses | 20 losses |
|---|---|---|---|---|
| 0.5% | 2.5% | 4.9% | 7.2% | 9.5% |
| 1% | 4.9% | 9.6% | 14.0% | 18.2% |
| 2% | 9.6% | 18.3% | 26.1% | 33.2% |
| 3% | 14.1% | 26.3% | 36.7% | 45.6% |
Long streaks are more ordinary than they feel. For a strategy that wins half its trades, a run of nine or ten losses somewhere in 1,000 trades is a normal result of chance alone. Find the longest losing streak in your EA's backtest, then allow for a longer one live. The drawdown calculator works out the same numbers for any risk and streak length.
Size every trade from its stop
A risk percentage only means something if the lot size follows the stop distance. A fixed 0.50 lots risks very different amounts on a 15-pip stop and a 60-pip stop. The formula is:
lots = risk amount ÷ (stop loss in pips × pip value per lot)
Worked example: EURUSD
- Balance $10,000, risk 1%, so the risk amount is $100.
- Stop loss 25 pips. One standard lot of EURUSD on a USD account is worth $10 per pip.
- Lots = 100 ÷ (25 × 10) = 100 ÷ 250 = 0.40 lots.
Worked example: gold
Gold (XAUUSD) usually trades in 100-ounce lots, and most traders count one pip as $0.10 of price, so one pip is worth $10 per standard lot. With the same $100 risk and a 50-pip stop, which is a $5.00 price move:
- Lots = 100 ÷ (50 × 10) = 100 ÷ 500 = 0.20 lots.
- Check: 0.20 lots is 20 ounces, and a $5.00 move on 20 ounces is $100.
Brokers trade in steps of 0.01 lots, so round down, never up: a result of 0.437 becomes 0.43. The position size calculator does this for any currency pair or gold in your account currency. Our gold trading guide explains why gold needs its own settings.
Most EAs offer either a fixed lot or a risk-percent input. Risk percent is usually the better choice when the EA places a real stop loss on every trade. If it doesn't, a “risk percent” setting may not mean what the label says, so read how the EA calculates it. Grid and martingale systems break the link between stop and size by design, which is why they need separate treatment; our grid and martingale guide covers it.
Drawdown and the math of recovery
Drawdown is the fall from an account's highest point to its lowest point before a new high. Maximum drawdown is the deepest of those falls. It tells you how painful an EA is to hold, and losses are harder to recover than they look:
| Drawdown | Gain needed to recover |
|---|---|
| 10% | 11.1% |
| 20% | 25.0% |
| 30% | 42.9% |
| 40% | 66.7% |
| 50% | 100% |
| 60% | 150% |
| 75% | 300% |
The arithmetic is simple. After a 50% loss, $10,000 becomes $5,000, and $5,000 has to double to reach $10,000 again. Past about 30%, the gain needed to recover climbs steeply.
Decide your maximum drawdown before you start. Write it down as a number, such as 15% below the starting balance, and treat it as the point where the EA stops trading, not a point to rethink. The risk of ruin calculator estimates how likely a strategy is to reach that limit from its win rate, reward to risk and risk per trade.
Daily and weekly loss limits
A maximum drawdown protects the account over months. Daily and weekly limits protect it from a single bad session, a broken price feed or an EA that misbehaves after an update.
- Daily loss limit. One simple approach is two to three times your risk per trade. With 1% per trade and a 3% daily limit, three full losses end the day.
- Weekly loss limit. One option is a little under twice the daily limit, for example 5% a week alongside a 3% daily limit, so two bad days in a row stop the week.
- Count open trades. Measure the limit on equity, which includes floating losses, not only on closed trades. An EA can sit on an open loss larger than your limit without closing anything.
- Fix the reset time. Use your broker's server time, the clock MetaTrader shows in Market Watch, and stick to it.
Hitting a limit isn't a signal to change settings. It means you stop, look at what happened and resume the next day or week if nothing is broken.
Running several EAs at once
Several EAs can smooth results when their strategies are genuinely different. They can also stack risk without you noticing, because each EA only sees its own trades.
Add up the combined risk
Three EAs each risking 1% can lose 3% together if they all hit their stops on the same day. Count the total risk of every open trade, and set your daily limit against that total, not per EA.
Watch for correlated pairs
Say EA A buys EURUSD, EA B buys GBPUSD and EA C sells USDCHF. That's three EAs and three symbols, but all three trades are bets against the US dollar. One strong dollar move hits all of them, so the account carries about 3% of risk on a single idea. The same overlap happens when two EAs trade gold, or when a trend EA and a breakout EA trade the same pair.
Practical rules help: cap total open risk across all EAs, for example at 3% to 4%, limit how many EAs can hold the same currency at once, and backtest the EAs over the same period to see how their losing stretches line up.
Give each EA its own magic number
The magic number is an ID an EA stamps on its orders so it manages only its own trades. Two copies of an EA with the same magic number on one account can close or modify each other's positions. Give every chart a unique number and note it in your journal.
Margin is shared as well. Several EAs in trades at the same time can use more margin than any one of them would alone, so check free margin at the moments when every EA is in the market.
Equity stops and kill switches
Limits only help if something enforces them while you're away. Build three layers:
- A stop on every trade. A stop loss held at the broker protects a trade even if MetaTrader or your server goes offline. An EA that closes losers only through its own code, often called a hidden or virtual stop, leaves the trade unprotected whenever the terminal disconnects.
- An equity stop. An equity protector is a small EA or utility that watches account equity. When equity falls a set amount below the day's start or the account peak, it closes every trade and blocks new ones. Set it slightly inside your written limits so it acts before they are reached.
- A manual kill switch. The Algo Trading button (AutoTrading on older MT4 builds) stops all EAs from placing new orders. It doesn't close open trades, so close or protect those separately.
If MetaTrader runs on a VPS, make sure you can reach it quickly from your phone or another computer. A kill switch you can't reach for an hour isn't much of a switch.
From demo to a small live account
A backtest shows how the rules handled the past. It can't show how your broker's spreads, slippage and execution treat the EA today. Move into live trading in stages:
- Backtest with realistic spreads and record the maximum drawdown, the longest losing streak and the average trade.
- Run the EA on a demo account with your broker for at least a few weeks, or until it has taken a few dozen trades.
- Go live at the minimum lot size, or at a fraction of your target risk, such as a quarter of it.
- Compare live results with the demo and the backtest: number of trades, average win and loss, and the spread and slippage on entries. Move up to your normal risk only when they line up.
Keep a journal
Record each EA, its version, the .set file, the symbol and timeframe, the magic number, the risk setting and the date of every change, with one line on why you made it. Add a weekly note of balance, equity, drawdown from the peak and any limit you hit. A journal turns “it feels worse lately” into numbers you can check, and it stops you changing three settings at once without knowing which one mattered.
When to pause or retire an EA
Every EA eventually meets conditions it wasn't built for. Set your exit rules in advance, while you're calm:
- Drawdown beyond the tested maximum. Live drawdown can exceed the backtest's worst, but a wide overshoot suggests something has changed. Some traders pause at about one and a half times the tested maximum: with a tested worst of 12%, that's a review at 18%.
- Behavior changes. Trade frequency, holding time, win rate or average loss drift well away from the test. An EA that used to trade twice a week and now trades ten times a day needs a look, even if it's making money.
- Broker conditions change. Wider spreads, new commissions, slower execution, a changed symbol specification, lower leverage or a new server time zone can all break an EA's assumptions. Scalpers feel it first.
- The market changes character. A range-trading EA in a long trend, or a trend EA in a flat market, can keep losing for reasons its rules can't fix.
Pausing isn't failure. Switch the EA off, run it on demo and let it earn its way back with evidence. Retire it if it doesn't. When you look for a replacement, our guide on how to evaluate an expert advisor lists the numbers and red flags to check first.