Golden Liquid MT4 Review: Backtests, Settings & DCA Risk
Golden Liquid MT4 is a dedicated XAUUSD scalper by Nguyen Hang Hai Ha that combines entries with DCA recovery. Its public controls cover lot sizing, spread, trailing exits, stop loss, drawdown and trading hours.
Our assessment: the straightforward input list will appeal to traders who want an adjustable gold EA. The key consideration is its averaging exposure: the available backtests are simulations, the drawdown limit is shown disabled, and there is very little customer feedback to assess.
View Golden Liquid MT4 — $299.95 →
Scalping entries and DCA recovery
The developer’s description identifies a proprietary scalping model but does not publish the exact entry formula. It explicitly combines those entries with dollar-cost averaging to manage floating positions. We therefore would not describe Golden Liquid as a single-entry or no-averaging system.
A recovery approach changes the exposure of an open sequence. A price retracement may help it exit, but continued adverse movement can increase losses and margin pressure. The public overview does not specify the full recovery spacing, lot progression or maximum depth, so those details need checking in the actual software rather than guessing from a smooth balance curve.
Two original developer backtests
The fixed-lot report uses 0.01 lots and a $1,000 starting deposit. It covers January to July 2026 on XAUUSD M5, showing 20,879 trades, profit factor 3.19 and 3.21% relative drawdown. The displayed net profit is $33,882.71.
Its 91.68% win rate needs context: the average winning trade is $2.58, while the average losing trade is $8.90 in this simulation. Frequent small wins do not make losses unimportant. Changes in costs or the frequency of larger losses can materially change a scalper’s result.
The automatic-lot example begins with $500 and displays roughly $24.4 million in net profit, with 20,872 trades, profit factor 3.52 and 17.87% relative drawdown. This is an extreme compounded simulation, not a realistic earnings forecast or a result we have achieved.
Both reports show 90% modelling quality, and their fixed displayed spreads differ: 35 in the automatic-lot test and 45 in the fixed-lot test. They are therefore not a clean comparison of sizing alone. Screenshots cannot establish how changing spreads, slippage, commission, rejected orders or broker volume limits were represented. We have not reproduced either test or found a linked product-specific forward account.
The protection settings need an active choice
The description says trades have stop loss protection, but its settings table allows Stop Loss to be set to zero, disabling it. The displayed stop is 3,000 points. Maximum Drawdown is also disabled at zero, and both supplied reports show MaxDrawdownPercent=0.
That is a meaningful distinction: a drawdown feature being available does not mean it constrained these tests. Configure and verify the actual control, and do not assume the historical drawdown percentage is a future maximum loss.
Lots, points and trading conditions
- Fixed Lot: a positive value selects manual lots. With zero, the Auto Lot input can determine sizing.
- Auto Lot: the developer’s example maps a value of 1 to 0.01 lots per $1,000 balance. This is a sizing rule, not a promise of 1% risk at the stop.
- Symbol digits: the author instructs users of three-decimal XAUUSD quotes to scale point-based exit distances by ten. Check the actual price distance and contract size before trading; do not blindly copy illustrative numbers.
- Trading hours: the listed 01:30–22:30 window uses broker server time, so it is not a universal local-time session.
- Execution: raw/ECN spreads, hedging permission and low latency are the author’s recommendations. Broker conditions still need testing.
The stated $300 minimum balance is the developer’s starting reference, not an assurance that a particular DCA sequence is affordable. Position size, contract value and recovery behaviour matter more than that number alone.
Customer reviews and live-signal evidence
The single entry is from Aleksandr Kiselev, dated 22 July 2026. It contains no written comment, so there is nothing substantive to quote about performance, reliability or support. We do not infer a star score from that entry or import it as a CheaperForex purchase review.
The comments page contained no discussion. The seller profile’s visible signal was for Power EURUSD on MT5, a different product; we have not used its returns as Golden Liquid evidence. No product-specific live signal or separate public manual was found in the inspected materials.
The linked broker article is a general recommendation with an introducing-broker rebate offer, not a Golden Liquid operating manual or independent performance study.
Buying decision and comparison
At $299.95, Golden Liquid offers a compact set of controls for gold scalping with recovery. It is most relevant to someone comfortable evaluating DCA exposure and broker-specific execution. The next useful evidence would be a monitored product-only account and reproducible tests with realistic costs.
For another MT4 gold system, see Quantum Dark Gold EA MT4. Compare its documented approach separately; neither settings nor results should be transferred between products. This review is based on published material and original captures, not hands-on live testing.
Golden Liquid review FAQs
Does Golden Liquid use averaging?
Yes. The developer explicitly describes DCA position recovery. The public description does not specify the full spacing, lot progression or maximum recovery depth.
Is there a public Golden Liquid live signal?
No product-specific live signal was found in the inspected listing, comments or seller profile at our 12 September 2026 check. The displayed seller signal was for a different EURUSD MT5 product.
What does the customer review say?
The single visible entry, dated 22 July 2026, contains no written feedback. It does not provide evidence about returns, broker compatibility or support.
Do the backtests prove the EA will reproduce those returns?
No. They are developer-provided historical simulations that we have not reproduced. Fixed spreads, modelled ticks, costs and execution conditions limit what can be concluded from the screenshots.