Ghost XAU EA is Leonid Arkhipov’s M15 gold robot for MT4 and MT5. It combines separate buy, sell and Guardian models with a virtual-trading mode, entry diagnostics and completed-trade replay. Its most useful distinction is the ability to inspect why it accepts or rejects a trade. Its main unresolved question is live performance: the developer’s linked MT5 signal had no trades when we checked it on 28 September 2026.
Our assessment: Ghost XAU is a candidate for traders who want a gold-specific system with visible decision controls and are willing to test it on their own broker. We would begin with virtual observation and conservative demo settings. The optional Recovery x2 feature increases exposure after losses, and the impressive historical report is not a substitute for an established live record.
This review examines the public manual, original tester report, both MQL5 listings and their shared signal. We have not independently replicated the backtest or traded the EA. CheaperForex sells the two editions.
How Ghost XAU chooses gold trades
The developer describes an ensemble of three models: 220 trees for buy decisions, 220 for sell decisions and 180 for the Guardian check. The inputs cover market state, momentum, volatility and other price features. Those model counts describe the advertised architecture; they do not independently establish predictive accuracy.
The practical workflow is more informative than the AI label. A potential entry must pass market and direction checks, confidence thresholds, Guardian checks and the operational gates. A rejected signal can therefore be a deliberate decision rather than a malfunction. The X-Ray panel is intended to show the relevant scores and restrictions, which helps distinguish a quiet strategy from a spread, session or news problem.
Ghost XAU is designed for the broker’s gold symbol on M15, including suffix variants. The developer describes no grid and no averaging, with an initial target-to-stop distance of 1:1. The manual also describes time-based exits and internal ATR-based levels, so actual closed-trade gains and losses need not have that exact ratio after costs and execution.
Ghost Mode, X-Ray and Decision Replay
Observe before placing orders
Ghost Mode uses the decision engine and the broker’s price feed with virtual positions and balance. It lets you watch which opportunities pass the filters without sending broker orders. It cannot demonstrate actual fills, slippage or the account’s response to live trading costs.
Inspect accepted and rejected decisions
X-Ray explains entry gates and model scores. Decision Replay examines completed decisions and their entry/exit context; it does not reopen a past trade. Together these tools can make setup checks more concrete than relying on a chart arrow alone.
Recovery x2 changes the risk after a loss
The public manual documents two base sizing methods: fixed lots and a percentage of equity. Optional Recovery x2 then doubles that base lot or risk after a losing close and holds the elevated setting until the accumulated deficit has been recovered. It does not continue doubling through an unlimited ladder, but repeated losses at the doubled level still deepen the deficit.
This matters when comparing settings. A 1% base risk with Recovery enabled can use 2% during recovery, subject to the EA’s sizing and broker constraints. Calling the system “no grid” does not mean it maintains constant exposure. We would inspect this switch explicitly before testing and assess the recovery sequence separately from the base setting. The supplied historical report has Recovery enabled.
The linked live signal has not established a track record
On 28 September 2026, the developer’s Ghost XAU signal showed a USD 150 initial deposit, balance and equity, 0.00% growth and zero trades. It displayed a start date of 23 September 2026 and an age of one week. MQL5 identified it as a real MT5 account on ICMarketsSC-MT5-3 with leverage of 1:1000.
The real-account label identifies the account type; it does not turn an empty trading history into evidence of profitability. At retrieval, MQL5 prohibited copying and flagged the absence of trading activity and leverage above its 1:500 copying limit. There are no winning streaks, drawdown cycles or execution records to evaluate yet. Both product listings link this same account, so it is not separate MT4 evidence.
What the historical stress report does—and does not—show
The developer’s original report archive contains an MT5 Strategy Tester report for XAUUSD_TICK on M15, covering 5 May 2003 through 24 September 2026. It starts from USD 10,000 with leverage of 1:100. The report identifies 1% risk, Recovery enabled and Friday blocking disabled. The developer attributes the tick data to Dukascopy; the report labels the test as using real ticks.
The reported 3,385 trades, 99.08% profitable trades and profit factor of 328.51 are exceptionally strong simulation outputs. They deserve scrutiny, not treatment as an expected outcome. We cannot establish from this package that the modeling avoided selection bias, that the full period was unseen during development or that the result can be reproduced independently. The original report also separates relative equity drawdown from maximum money drawdown; those are different measures and should not be substituted for each other.
The test settings include news protection, but the manual says live WebRequest is not used in Strategy Tester. That limits what the input switch alone proves about historical calendar filtering. A fresh broker-specific test, followed by forward observation with the intended settings, would be more useful than projecting the historical equity curve into future returns. Neither this report nor the shared MT5 gallery establishes equivalent MT4 results.
MT4 or MT5: choose the platform you actually use
Ghost XAU EA MT4
The MT4 listing advertises the same core model names, Ghost Mode and diagnostic features. Choose it for an existing MetaTrader 4 gold setup. Check the actual MT4 input labels and terminal permissions rather than copying the MT5 manual’s installation screens word for word.
Ghost XAU EA MT5
The MT5 edition matches the platform identified in the shared public guide, linked signal and historical tester report. This makes the reference material easier to follow directly, but does not prove that the MT5 engine earns more or has extra trading strategies.
Both current MQL5 listings displayed version 5.87 and a 24 September 2026 publication date when checked. Older interface captures show 5.84. These are dated source observations, not verification of the version in a customer’s standalone package. No separate public release-history entry was available to explain that interface difference.
Broker and setup checks that matter
- Gold symbol and M15 history: check the actual contract, tick value, minimum lot and lot step. The preflight panel checks symbol/history readiness and broker trading constraints.
- Spreads and execution: the developer recommends low-spread conditions. The manual’s spread limit is expressed in gold-price units per ounce, so do not mistake it for a universal points value. Ghost Mode cannot verify live execution quality.
- News and broker time: the guide uses the FairEconomy calendar WebRequest domain and requires the broker GMT/DST settings to match the server. The default news windows are 60 minutes before and after; inspect the configuration you actually use.
- Friday and session gates: Friday blocking prevents new entries; it is not a promise that every existing position is closed before the weekend.
- Terminal operation: keep the chosen MT4 or MT5 terminal and connection running. A VPS can help continuity. The manual says DLL imports are not required.
The shared public guide comes in 11 languages and uses an MT5 installation flow. Its MQL5 purchase and activation steps describe that marketplace’s offer. Follow the separate delivery instructions supplied with the CheaperForex order for this standalone offer.
Customer feedback and other EAs by Leonid Arkhipov
Both Ghost XAU listings showed no customer reviews when checked on 28 September 2026. The developer’s broader profile rating is not a Ghost XAU rating, and we have not imported it into this product’s score. The absence of buyer feedback is another reason to treat this as a new release rather than an established recommendation.
For another gold-focused product from the same developer, compare Aurum AI MT4 and Aurum AI MT5, with the Aurum AI review providing its own strategy context. If you are comparing the developer’s currency-market products, see EUR Machine Pro and GBP Machine Pro. Their histories and trading approaches should be assessed separately; sharing a developer does not transfer results to Ghost XAU.
Buying decision: the diagnostic controls and virtual mode are the clearest reasons to consider Ghost XAU today. The missing live trading history and the optional increase in risk after losses are the main reasons to test it carefully before relying on it. Use the edition matching your terminal and compare its actual settings with the documented examples.
Sources checked 28 September 2026: official MT4 listing, official MT5 listing, the linked guide, tester archive and signal. Trading involves risk and historical simulations do not guarantee future results.
Ready to Buy Ghost XAU EA MT5?
$239.99 with crypto · $299.99 by card
Official version · Instant download · Free lifetime updates · 7-day money-back guarantee
