Due Diligence File — 8 Required Documents
Click each document card to expand. Mark off what your vendor provides.
Insufficient documentation — high risk
Click the document cards and mark items you have received from your vendor
What Documentation Should You
Demand Before Buying a Gold EA?
The 8 documents every legitimate XAUUSD EA vendor should provide upfront — and the red flags that signal you should walk away.
Published 15 June 2026 · Updated as best practices evolve
Before buying any XAUUSD EA, demand: (1) named strategy description, (2) 3-year backtest report on tick data, (3) live or forward-test results on a verified account, (4) parameter documentation, (5) confirmation of hard SL on every trade, (6) a contactable support channel, (7) update history/changelog, and (8) clear licence terms. If any of these 8 are refused or unavailable, treat it as a red flag proportional to which ones are missing.
Why Documentation Matters — The Problem It Solves
The EA market has a significant transparency problem. Because MT5 Expert Advisors are compiled — meaning you receive a binary .ex5 file, not readable code — you cannot inspect what the EA actually does. You are buying on faith unless the vendor provides documentation.
This creates three distinct risks that documentation directly addresses:
Over-optimised backtests
A developer runs 10,000 parameter combinations on historical data and presents the one that performed best. The resulting backtest looks extraordinary — 90% win rate, smooth equity curve, minimal drawdown. In live trading, it performs at random because the patterns found were specific to the past data, not repeatable market structure.
How documentation protects you: A full backtest report (not a screenshot) allows you to check modelling quality, test duration, and consistency across sub-periods. High modelling quality + consistent performance across 2020, 2022, and 2024 conditions suggests real robustness.
Hidden risk: no real SL
Some EAs advertise no stop loss losses but use grid or martingale logic — averaging down into losing positions. They show a long period of small wins, then a single catastrophic loss that wipes all accumulated profit. No SL is not a feature; it is an unquantified risk.
How documentation protects you: Risk settings documentation confirms whether SL is present at the broker level on every trade. If it is not explicitly stated, assume it is not there.
Disappearing vendors
EA vendors disappear more often than they should. A vendor who closes their website or MQL5 account takes their support channel with them. If your EA needs an update after a major MT5 version change, there is no one to call.
How documentation protects you: Update history and licence terms reveal vendor commitment. Recent updates indicate active engagement; a clear licence policy (MetaQuotes ID-based, not account-locked) protects you if the vendor closes.
How to Request Documentation — Scripts That Work
Most traders do not ask because they feel awkward or do not know what to say. Use these pre-sales email templates.
Template 1: Pre-Sales: Backtest and live results request
Hello, I'm considering purchasing [EA name] for XAUUSD trading on MT5. Before buying, I would like to review: (1) the full MT5 Strategy Tester report (.htm file, not screenshot) for at least the last 3 years on XAUUSD, (2) a link to your live or forward-test results on a verified platform (MyFXBook, FXBlue, etc.). Could you share these? Thank you.
Template 2: Pre-Sales: Risk management question
I have a specific question about risk management: Does [EA name] place a hard stop loss at the broker level on every trade? I want to confirm this is not a virtual stop managed only by the EA. Please confirm the SL mechanism and the default SL size in pips.
The Licence Terms Trap — What to Check
Licence terms are the most frequently overlooked document — and the one that causes the most friction after purchase.
As an example of clear documentation: Goldie Razor V2.8.4 provides upfront documentation of its strategy (M15 range breakout with H4 200 EMA filter), confirms hard stop loss on every trade, includes a spread filter in the standard package, and provides a complete setup guide. This is the documentation standard to use as a baseline comparison for any EA you evaluate.
Further Reading
Related pages to complete your EA evaluation process.
Best Expert Advisor for XAUUSD in 2026
Which EAs meet the documentation standard outlined on this page.
Gold EA Development Cost: How Much Should You Spend?
Understanding vendor costs helps you assess whether their pricing reflects real investment.
What Is the Best EA for XAUUSD?
EA comparison once you have completed your due diligence process.
Who Makes the Best EA?
Evaluating the vendor behind the EA, not just the product itself.
Gold Trading Signals: Buy or Build?
Applying the same due diligence framework to signal services as an alternative to EAs.
Frequently Asked Questions
Red flags in a backtest: (1) Balance curve is perfectly smooth with no significant drawdown periods — real markets have drawdown. (2) Test period is only 6–12 months on a recent bull run. (3) Modelling quality is below 90% — this means the test used incomplete data. (4) The report is a screenshot, not the original .htm file — screenshots can be edited. (5) Every trade wins, or win rate is above 90% with large winners — this almost always indicates curve-fitting. Ask for the .htm file and open it yourself in a browser.
A forward test runs the EA on a demo account in real-time on live market data but without real money. A live account uses real money with a real broker. Both show how the EA performs on data it has never seen (unlike a backtest). A live account with real money carries more credibility because the vendor has financial skin in the game, but a properly conducted demo forward test for 6+ months is still valuable evidence.
If a vendor will not share backtest data, forward test results, or strategy documentation — do not buy. The phrase "we cannot share our algorithm for IP protection reasons" is legitimate for the source code, but not for: what the strategy type is, what the backtest statistics show, or whether there is a stop loss on every trade. These require no code disclosure. Refusal to share them indicates either that the results are poor, or that the strategy violates basic risk management principles.
Yes — and you should. Send a pre-sales email asking for: (1) backtest report for the last 3 years on XAUUSD, (2) confirmation that every trade has a hard stop loss, (3) a link to their forward test or live account results, (4) their parameter documentation. A legitimate vendor will provide all of this readily. The quality and speed of their response is itself a form of due diligence on the vendor's professionalism.
Yes — this is the most common failure mode for EAs that show great backtests but poor live performance. Over-optimisation (also called curve-fitting) occurs when the developer runs thousands of parameter combinations on historical data and selects the one that performed best — finding patterns in past data that do not represent repeatable market structure. Signs include: suspiciously high backtest performance, very tight parameter ranges that only work in a specific window, and dramatically different live versus backtest results.
Goldie Razor V2.8.4
M15 breakout + H4 EMA filter — built for XAUUSD on MT5