MQL5 verified seller Instant licence delivery Lifetime updates & support
New: Every Apex Drawdown Zero licence now includes 12 months of Forex VPS hosting, free. Plus 10 MT5 account seats instead of 5 until 30 Nov. See the offer →

Best Gold EA for MT5: 7 Checks Before You Buy an XAUUSD Robot

Apex Drawdown ZeroAuthor Tshivhidzo MbedziRisk Management
Published Nov 14, 2026, 07:14 PM · 6 min read
Best Gold EA for MT5: 7 Checks Before You Buy an XAUUSD Robot
© Apex Drawdown Zero
In this article: XAU/USD +0.03% EUR/USD -0.61%

The best gold EA for MT5 is not the one with the highest backtest. It is the one that is still trading your account in six months. XAUUSD is the most popular market for Expert Advisors and also the one that destroys them fastest, so any shortlist of XAUUSD robots has to start with how they lose, not how they win.

This guide gives you seven checks to run on any gold EA before you pay for it. Each takes a few minutes, needs no coding, and filters out most of what is for sale. At the end, we show our own XAUUSD robot measured against the same checks, including the numbers that do not flatter it.

Why gold is different from every other EA market

Gold moves several times more than a major currency pair in a typical session. Spreads widen sharply around US data releases, weekend gaps are normal, and a single news candle can travel further than EURUSD does in a week. That has three consequences for an Expert Advisor:

  • Fixed-pip stops break. A 20-pip stop that suits EURUSD is noise on XAUUSD. Stops and targets need to scale with volatility, usually through ATR.
  • Grid and martingale systems blow up faster. They rely on price coming back. On gold, the move that does not come back arrives sooner.
  • Backtests lie more. Variable spreads and slippage on gold are hard to model, so a tester report on XAUUSD flatters an EA more than on any major pair.

The 7 checks for any XAUUSD EA

1. Is there a stop-loss on every trade, from the moment it opens?

This is the single most important filter. An EA with no hard stop-loss can turn one bad day into a blown account. Ask the seller directly, or run the EA on a demo account and look at the open positions in MetaTrader: the SL column should never be empty. Our article on why grid and martingale robots fail shows what happens without one.

2. No grid, no martingale, no averaging down

Look at the trade history for several positions opened in the same direction at worse and worse prices, or lot sizes that double after a loss. Either one means the EA is hiding losses rather than taking them. These systems produce beautiful equity curves right up to the day they do not.

3. Equity drawdown, not balance drawdown

Balance drawdown only counts closed trades, so an EA that holds losers open can report almost none. Equity drawdown includes floating losses and is the number that actually blows accounts and breaks prop firm rules. If a seller only shows balance drawdown, or claims "zero drawdown", read does a zero drawdown EA exist? before going further.

4. A verified live account, not just a backtest

A backtest is the seller's best-case story. A live account tracked by a third party, such as Myfxbook with "track record verified", shows what happens with real spreads and slippage. Check that the account has run for at least a few months, that deposits and withdrawals are visible, and that the trading privileges are verified, not just the balance. Our guide to five ways a backtest flatters an Expert Advisor explains what to look for in the gap between the two.

5. Risk per trade you can actually survive

A gold EA should size positions as a fixed percentage of the account, not a fixed lot. Then check what that percentage means for you: at 2% risk per trade, five losses in a row already cost close to 10%, which is the whole drawdown limit on many prop firm challenges. Put the EA's win rate and reward:risk into our risk of ruin calculator to see the probability of hitting your limit, and use the lot size calculator to check the position sizes.

6. A spread and news filter

Gold spreads can multiply during news releases and around the daily rollover. A good XAUUSD EA checks the spread before entering and stays out when it is abnormal, and either avoids high-impact news or has been tested through it. If the settings have no spread limit at all, assume the EA will eventually enter at the worst possible price.

7. Realistic claims and a seller you can reach

Monthly returns of 30% or more, "no losing months" and "guaranteed profit" are warning signs, not features. A real gold strategy has flat and losing stretches, and an honest seller shows them. Check that there is a real support channel, a licence you can move between accounts, and updates when brokers or market conditions change. Our post on losing 96% of a live account covers the checks we learned the hard way.

Quick checklist

CheckPassFail
Stop-lossOn every trade from entryNone, or "virtual" only
Position managementOne position, fixed riskGrid, martingale, averaging
Drawdown shownEquity drawdown, publishedBalance only, or "zero"
ProofVerified live account, months oldBacktest screenshots only
Risk sizingPercentage of accountFixed lots
Market conditionsSpread and news filterTrades any spread
ClaimsLosing months disclosed"Guaranteed", "no losses"

Our gold EA against the same checks

We build and sell Apex Drawdown Zero, an MT5 Expert Advisor for XAUUSD, EURUSD and EURJPY on H1. Here is how it scores on the list above:

  • Stop-loss: a structural stop-loss on every trade from the moment of entry, with break-even management.
  • Position management: no grid, no martingale, no averaging and no recovery mode. At most one qualified trade a day.
  • Volatility: breakout distance, range filtering and stop placement scale with ATR, so the EA adapts as gold's volatility changes.
  • Risk sizing: automatic percentage-based position sizing.
  • Proof: one $500 real-money deposit on 17 July 2026 at RoboForex, with no withdrawals, reached $854.36 by 12 October 2026: +70.9% over 87 trades, profit factor 1.69. The account is public on Myfxbook.
  • Drawdown: maximum drawdown on that account was 13.69%. Not zero, and we say so. August did most of the work, September added about $29, and early October was negative.

Past results do not guarantee future results, and a strategy that took 13.69% drawdown once can take more. Run it on a demo account first, size risk with the calculators above, and only trade money you can afford to lose.

Common questions

What is the best timeframe for a gold EA?

Most robust XAUUSD robots trade H1 or higher. On M1 and M5, spread and slippage are a large share of every trade, so results depend heavily on the broker and rarely survive the move from backtest to live account.

Can a gold EA pass a prop firm challenge?

Yes, if it has a stop-loss on every trade, risks a small fixed percentage and respects the daily loss limit. Check the firm's rules on EAs first: our guide to whether EAs are allowed at prop firms covers the common restrictions, and the prop firm calculator gives a safe risk per trade for the limits.

Do I need a VPS for a gold EA?

Yes, if you want it to trade while your computer is off. An EA only runs while its MetaTrader terminal is open and connected. A forex VPS keeps it running 24/5 with low latency to your broker.

How much money do I need to run an XAUUSD EA?

It depends on your broker's minimum lot size and the EA's stop distance. With 0.01 lot minimums and a 1% risk per trade, several hundred dollars is usually the practical minimum on gold. Use the lot size calculator with your EA's typical stop to check.

Tools for this setup

From our catalogue, matched to what this article covers.

Automate the discipline part.

Apex Drawdown Zero sizes every trade from a fixed percentage and attaches a stop-loss, so you don't have to rely on willpower. It has no account-level drawdown halt in V9; V12 adds one.

Learn more
Keep reading

Related articles

All articles →