How to test XAUrally properly in 15 minutes (and what the account guards actually do)
Two traders have downloaded the XAUrally demo so far. If you are one of them, or about to be, this post will save you from the most common testing mistakes on a multi-engine M1 gold EA.
XAUrally runs three engines on one XAUUSD M1 chart: BASKET, BREAKOUT and LEVEL. Two of them can switch themselves into shadow mode, so a short or badly configured test can show almost nothing, or the wrong thing. Here is the checklist I use myself.
STEP 1 — Use the right tester settings
- Symbol: your broker's XAUUSD (any name containing XAU or GOLD works).
- Timeframe: M1.
- Modelling: "Every tick based on real ticks". The guards and the basket react to intrabar moves, so cheaper modelling modes are not representative.
- Deposit: 1,000 USD. Leverage: 1:100. Inputs: defaults.
STEP 2 — Set the server time correctly
In the Strategy Tester, AUTO assumes an EET server. If your broker's server is not EET, select "Fixed GMT offset" and enter your server's offset. The Asian-range module and the daily guard both depend on the trading day, so the wrong offset shifts every session.
STEP 3 — Test long enough
The BREAKOUT swing modules start in shadow mode. They place virtual trades first and send real orders only while their recent shadow results are positive. In a one-month test you may see very few breakout trades. Use at least 6 to 12 months.
STEP 4 — Run a second, harder test
Results depend strongly on the starting month. Some 12-month periods ended with a loss, and calm, low-volatility months tend to lose. Pick a quiet period on purpose and judge the result by equity drawdown, not by net profit. If the drawdown on a bad period is something you can live with, the EA may suit you. If not, it does not, and that is a useful answer too.
STEP 5 — Read the report the right way
Look at these, in this order:
- Maximal equity drawdown (not balance drawdown, because the basket can hold floating losses).
- Profit factor.
- Total number of trades. Fewer than 100 trades is too small a sample to judge.
- The worst month.
WHAT THE ACCOUNT GUARDS DO
XAUrally has two guards that watch the whole account on every tick.
- Trade-idea guard: if the floating loss in one direction reaches 20% of balance (default), that side is closed and new entries in that direction pause for 60 minutes.
- Daily guard: if equity falls 20% (default) from the day's reference, all XAUrally positions and pending orders are closed and trading pauses until the next trading day. The state is saved, so a terminal restart does not reset it.
On top of that, the BASKET engine has its own brake (300 USD loss per 0.01 lot by default) and a maximum basket age.
Two honest limits: the guards cannot act while the market is closed, so a weekend gap can exceed them. And 20% is far too loose for prop-firm challenges, which is why XAUrally is not designed for them.
IF YOU DO NOT WANT AVERAGING
The BASKET engine adds equal-lot layers when price moves against it. No martingale, maximum 10 layers, but it is averaging. If that is not for you, switch BASKET off in the inputs and test BREAKOUT and LEVEL only. Run the same two tests and compare.
BEFORE GOING LIVE
Each release carries a built-in switch history valid for 45 days. For the most current history, run a backtest on your broker's XAUUSD M1 chart ending today with default inputs. The EA saves the file to the terminal's common folder, and the live EA loads it automatically if it is less than 14 days old. Then use a 24/5 VPS, one chart per account, and no other EAs on the same account.
TRY IT
The demo is free in the Strategy Tester. If the tests look acceptable to you, rent XAUrally for one month and run it on your own account before deciding on the unlimited licence.
XAUrally on the MQL5 Market: https://www.mql5.com/en/market/product/204969
Previous post: XAUrally: three gold engines on one chart - how it works and how to test it
Questions about your test results are welcome in the comments or by private message.
Backtest results do not guarantee future results. Leveraged gold trading is high risk.
