MT5 Expert Advisor: Pre-Live Risk Settings Checklist
· Updated
An MT5 Expert Advisor should be audited for risk inputs before it is judged by backtest or live performance. The audit should cover symbol mapping, lot sizing, spread limits, stop-loss behavior, magic numbers and account-level safeguards. Without that record, live results cannot be fairly compared with the Strategy Tester or any historical report.
Why an MT5 Expert Advisor audit comes before performance
An MT5 Expert Advisor should be audited for operational risk inputs before it is judged by profit, loss or an equity curve. If the symbol is mapped incorrectly, the lot size does not match the broker’s contract, the spread filter is disabled or the magic number conflicts with another EA, live trading is no longer a clean test of the robot’s logic.
MetaTrader’s official help describes the Strategy Tester as the place to test and optimize trading robots before using them in live trading, including a run of the EA on historical data with initial parameters. The practical shift is to treat the tester not only as a performance report, but as a controlled environment for checking whether risk inputs behave as expected. metatrader5.com
A pre-live audit is not a market forecast. It is a quality-control process: it verifies that the EA does what you think it does, on the intended symbols, with the intended size, cost limits and response to adverse conditions.
Symbol mapping, contract specifications and multi-market testing
Many live execution problems do not start with the strategy; they start with the symbol. An MT5 Expert Advisor designed for several markets must know the broker’s exact symbol name, minimum tradeable size and the cash effect of each price movement.
MetaTrader’s help explains that the tester can process data for symbols used by a multi-currency Expert Advisor, but the required symbols must be enabled in Market Watch. A symbol audit therefore includes symbol activation, name matching, history availability and contract-spec comparison, not just selecting a chart. metatrader5.com
- Compare the broker’s exact symbol name with the EA’s expected symbol; XAUUSD, XAUUSD.m and GOLD can represent different contracts.
- Read the minimum volume, volume step and maximum volume from the symbol specification, then compare them with fixed-lot or percentage-risk inputs.
- For multi-currency EAs, make sure every required market is visible in Market Watch, not only the chart where the EA is attached.
- If the demo and live accounts are hosted on different servers, check contract specifications separately on both.
| Market group | What to check in MT5 |
|---|---|
| XAUUSD | Broker suffix or prefix, contract size, minimum and step volume, tick value, typical spread |
| JPY pairs such as GBPJPY and USDJPY | Decimal format, pip value, minimum stop distance, trading sessions |
| Cross pairs such as EURGBP and AUDCAD | Profit currency, conversion to account currency, swaps and commissions |
| Multi-symbol operation | Market Watch visibility, no duplicate names, server-time consistency and historical data availability |
Lot sizing, drawdown limits and account-level safeguards
Lot sizing is one of the most important risk inputs because it translates trade outcomes into account impact. Two EAs with the same entry logic can produce very different drawdowns if the position size is different; therefore size calculation must be understood before performance is evaluated.
MT5 advanced testing settings allow the user to specify account parameters such as trading limits, margin settings and commissions to simulate different broker conditions. That makes them useful for checking how an EA reacts to margin constraints, commissions and position limits, not just for reading net profit. metatrader5.com
Recent MQL5 community guides also highlight account-protection inputs such as maximum daily loss, maximum total loss, safety buffers and account-wide close rules, while warning that such settings can affect manual trades and other EAs on the same account. mql5.com
- If the EA uses fixed lots, compare that size with the worst drawdown period the account could tolerate, not only with average historical profit.
- If the EA uses dynamic sizing, test the formula against balance, equity, stop distance and the tick value at the same broker.
- Define a cap for each trade, a cap for total open positions and a cap for simultaneous exposure across symbols.
- For accounts with daily rules, align the protection with the account’s actual calculation method: balance, equity, start-of-day value or another rule.
- If manual trades or other EAs share the account, understand whether account-level controls will affect all positions.
Spread, slippage, stop-loss behavior and trading pauses
Spread and slippage are not invisible costs; they are part of the EA’s execution logic. A robot that behaves acceptably in normal spreads may behave very differently near the market open, during news or in thin liquidity if no filter or rejection rule is active.
The MetaTrader Strategy Tester can emulate execution delays so the effect of time between trade request and execution can be studied. Under instant execution, the EA’s response to a requote can also be tested when the difference between requested and execution prices exceeds the allowed deviation. metatrader5.com
A stop loss is not only a distance value; the submission method matters. A stop sent after entry, a virtual stop managed by the EA, or a stop that depends on continuous connectivity creates a different operational risk profile from a server-side stop placed immediately.
| Input | Audit question |
|---|---|
| Spread limit | Should the EA block entries or only warn when the limit is exceeded? |
| Slippage or deviation | If the fill price differs from the requested price, is the order rejected or accepted? |
| Stop loss | Is it sent with the entry, modified afterward or managed virtually by the EA? |
| Restricted sessions | Are new entries stopped, trade management stopped or both? |
| News and events | Are only new entries blocked, or can existing positions also be closed before the event? |
Magic numbers, EA conflicts and account controls
The magic number is a small control with a large effect. It tells the EA which positions belong to it; without clean identification, a robot may manage a trade it should ignore or ignore a trade it should manage.
When several instances of one EA run on multiple symbols, or several EAs run on the same account, magic-number auditing should be treated like identifier management in software. Each identifier should be unique, documented and reviewable, especially when the robot also uses account-level protections.
MT5 visual testing lets traders observe EA trade operations on a chart and inspect orders, deals and modification requests in the toolbox. This is useful for finding failures such as wrong position handling, rejected requests or unsuccessful stop-loss changes. metatrader5.com
- Assign a unique magic number to each EA instance on each symbol and keep a written record.
- If the EA should manage only its own positions, make sure manual trades are not mixed into the same identifier logic.
- If account-wide protection is enabled, accept in advance that it may affect manual trades and other robots.
- Use any close-all or delete-pending-orders setting only when its scope is fully understood.
- Read the MT5 journal after visual testing; rejected orders, error codes and stop modifications often do not appear in the final equity curve.
From Strategy Tester to live account: build an audit file
The checklist is not designed to predict the future. It is designed to make later differences explainable: if performance changes, you can investigate whether the cause was market behavior, broker execution or a settings mismatch.
MT5 allows parameter sets to be saved and reused, which helps keep EA inputs portable and repeatable. That makes it easier to compare the version used in the tester, the demo account and the live account. metatrader5.com
For LFP, reading a backtest without understanding risk inputs is not enough. A trader should know which markets the robot trades, what historical drawdown it reported in the backtest and how the intended live account size and risk tolerance will map onto that behavior.
- Create a baseline settings file and record the date, broker, account type, server, initial balance and EA version next to it.
- Run the same settings in the Strategy Tester with historical data, execution delay and costs that resemble the intended account.
- Do not judge the result only by net profit; review drawdown, underwater periods, trade count, losing months and stop behavior.
- After the tester, run the same file on demo and compare journal messages with the backtest report.
- Before scaling up, consider a short live observation period at the smallest permitted size or another low-risk environment to see broker execution.
Frequently asked
- What should I check before running an MT5 Expert Advisor live?
- Check symbol names, broker suffixes, lot sizing, spread limits, stop-loss behavior, magic numbers, daily loss limits and account-level protections before live use. The same settings should be tested in the Strategy Tester and then observed in demo or very small live size.
- What is a magic number in an EA?
- A magic number is an identifier that helps an EA distinguish its own trades from manual trades or trades opened by other robots. If multiple EA instances share the same number, position management can become ambiguous.
- Is the MT5 Strategy Tester enough to approve risk settings?
- No. The Strategy Tester is essential for checking historical behavior and input logic, but broker execution, spreads, slippage and server conditions still need demo or low-risk live observation.
- Why do spread and slippage limits matter for an MT5 Expert Advisor?
- A spread limit can stop entries when trading costs widen beyond the intended range. Slippage or deviation settings define how the EA reacts when the execution price differs from the requested price.
- Why is lot sizing more important than the entry signal?
- The entry signal decides when the EA trades, but lot sizing decides how much a wrong trade or losing period can affect the account. A reasonable signal can still create excessive risk if the position size is wrong.
This is educational material, not investment advice. All performance figures are backtest results, not live trading, and are no guarantee of future results.