Gold Spread Reversion EA Free Download – Precise Pair Trading MT5 Review

Gold Spread Reversion EA is an automated MT5 robot that trades EURUSD and XAUUSD (gold) as a single two-leg spread instead of picking a direction on either one. Its public live signal shows 396% growth since September 2025. We ran a 100% real-tick backtest over the identical window and hit the same profit factor to two decimals, which is rare. It also showed where every dollar of that profit came from, and it is not where the name suggests. Free download at the bottom.

TL;DR – Free MT5 pair-trading robot. Two legs, EURUSD against XAUUSD, opened and closed in the same second, fixed lots, no grid and no martingale. Also no stop loss on either leg. Our matched backtest agrees with the live signal on profit factor (1.38 against 1.38) but shows the gold leg earning 100% of the profit while the euro leg loses money. Rating: 3.5 / 5.

What is the Gold Spread Reversion EA?

Gold Spread Reversion EA is a free MetaTrader 5 expert advisor that opens a buy on one symbol and a sell on another at the same moment, then closes both when the price gap between them narrows. On its shipped settings it pairs EURUSD with XAUUSD on an hourly chart. It holds one position per symbol, never adds to a loser, and exits on its own signal rather than on a stop loss.

That is pair trading, sometimes called statistical arbitragess="fc-gloss-tip" data-slug="arbitrage" tabindex="0">arbitrage: betting that two instruments which normally move together will snap back after they drift apart. The bet is on the gap, not on the market going up or down.

Core Strategy and Logic

We ran the EA on a Tickmill demo with 100% real-tick history over 20,138,080 ticks, then read the order and deal log rather than the summary page.

• Always two legs, always opposite – Of 303 entry timestamps, 298 opened exactly two positions in the same second, and every one of those 298 paired a buy on one symbol with a sell on the other. 164 were sell EURUSD plus buy XAUUSD, 134 the reverse.
• The hedge never breaks – All 298 baskets also closed both legs in the same second, so measured across 827 hours of open positions the time spent holding only one leg came to zero. The only exceptions were 5 lone EURUSD orders fired at the 01:00 session boundary and closed one second later, costing $5.96 in total.
• Fixed lots, no recovery – Every one of the 1,202 orders was 0.04 lots on EURUSD or 0.02 on XAUUSD. No doubling, no averaging, no lot ladder. The “not martingale” claim on the listing is true and we can show it.
• No stop loss and no take profit – Every order in the log carries an empty S/L and T/P field. Nothing is parked at the broker. Positions close when the EA decides the gap has closed, or when its profit target trips. This is the single biggest thing the sales page does not mention.
• A correlation filter that really switches it off – The CorrThreshold input ships at 0.8. In November 2025 the EA took 4 baskets in the entire month and in December 8, against 52 in April 2026. When EURUSD and gold stop tracking each other, it mostly stops trading.
• Short holds, but not always – Median holding time was 46 minutes and MetaTrader reports the average as 1 hour 38 minutes. Seven baskets were carried across a weekend, the longest running 62 calendar hours from a Friday to a Monday.

The legs are not the same size, and this is where the strategy stops being neutral. A 0.04 EURUSD position is about $4,660 of notional exposure. A 0.02 XAUUSD position averaged about $8,980 across our test window. The gold side carries roughly two dollars of exposure for every one on the euro side, and gold moved from $3,445 to $5,534 during the test while EURUSD stayed inside a 6% band. The spread this EA trades is, in practice, a gold trade with a small euro brake on it.

That makes it a different animal to the MT4 correlation hedge from cmillion, which sizes both legs for equal notional and tops up the basket when the gap widens. It is closer in spirit to the Mean Reversion Trading EA, applied to a synthetic instrument rather than a single pair.

Please test in a demo account first for at least a week. Also, please familiarize yourself and understand how this MT5 Gold Robot works, then only use it in a real account.

Recommendations

• Minimum deposit – the vendor says 300 EUR at 1:33 leverage, and the margin arithmetic backs that up with almost nothing spare. At 1:30 with gold near $5,500, a single 0.01 lot gold position needs about $183 of margin and the euro leg another $39. The live account’s maximum deposit load of 98.93% is that number in practice.
• Pair and timeframe – EURUSD with XAUUSD on H1, which is what both the live account and our test ran. The symbol inputs accept any pair, but nothing else is proven.
• Account type – low spread ECN. Commission on the euro leg alone came to $72.72 in our test, against a $24.79 loss on that leg, so a wide-spread account turns the hedge into pure cost.
• VPS is effectively required. It trades around the clock inside a 22-hour window (FXVM).
• Check the margin per basket on the margin calculator before funding, then size with the lot size optimizer (Find the Perfect Broker For You Here).

FXVM Recommended VPS · From $19.99/mo Get a VPS Deriv Binary Trading Open Account

Recommended Brokers Compare All

XM Editor's Pick

Best for low-deposit beginners

$5 min deposit1:1000 leverageMicro accountsMT4/MT5

Open Account

Plexytrade Allow USA Traders

Best for high-leverage trading

$50 min deposit1:2000 leverageSmall accountsMT4/MT5

Open Account

Prop Trading Firms Compare All

Moneta Funded Editor's Pick

Instant funding · 1-step evaluation

Instant Funding1 Step EvaluationMT5 SupportUSA Traders

FLASH20 20% off challenge fee Join Now

Performance Review: Live Signal and Our Own Backtest of Precise Pair Trading EA

Two sources. The Precise Pair Trading Pro EA run on a real IC Markets EU account at 1:30 leverage since September 2025. Our backtest ran on a Tickmill demo, MT5 build 6061, 100% real-tick quality, over the identical window of 1 September 2025 to 31 July 2026, on the shipped defaults, with a $250 deposit at 1:100.

Before the numbers, the honest limits. The live account is in EUR at 1:30 and ours is in USD at 1:100, so the two run different position sizes. The live account also traded a little XAUEUR and one AUDUSD deal alongside the main pair, meaning it is running more than one instance. And a tester models slippage and swap imperfectly. Treat our run as a clean fixed-lot reference for how the strategy behaves, not a prediction of the live curve

1. The live record

One detail deserves credit first. Deposits total 2.42 EUR on top of the original 224.79 EUR, and withdrawals are zero. On most signals the headline growth figure is inflated by cash moving in and out mid-period, and the honest number has to be rebuilt from net profit against total capital. Here the two agree: 899.67 EUR of profit on 224.79 EUR deposited is 400%, and the page says 396%. The number on the page is real.

2. Our backtest reproduces the profit factor exactly

Two independent runs, on different brokers, in different currencies, at different leverage, arrived at the same 1.38 profit factor and a recovery factor within 0.23 of each other. In both records the average win is slightly smaller than the average loss, the same signature. That agreement is the most useful line in this review: the published record is the strategy doing its normal work, not a lucky account or a private setting.

The columns diverge where you would expect. Our run made more (552% against 396%) and hurt more (56% drawdown against 24%) because it carried more size per basket. That dial belongs to whoever funds the account, not to the EA.

3. Where the profit actually comes from

Split the result by symbol and the pair-trading story collapses into a gold story. Our backtest attributes +$1,405.13 to the XAUUSD leg and -$24.79 to the EURUSD leg, against a net of $1,380.34. The gold leg supplied more than 100% of the profit. The live signal’s own distribution panel says the same thing from a completely separate account: gold up about 1.1K, EURUSD down about 64, XAUEUR down about 24. In pips it is starker still, roughly 98,000 gained on gold against 2,800 given back on EURUSD.

This is not a flaw in the code, it is the arithmetic of the lot pair. The gold leg carries about twice the notional and sits on an instrument that moved 61% over the test window, while the euro leg sits on one that stayed inside 6%. The euro side was never going to be a profit centre. It is a brake that costs commission, and in our run it cost $72.72 of commission to lose $24.79.

So judge this as a gold system with a hedge attached, and size it against your tolerance for gold rather than the comfort the phrase “pair trading” suggests. The Quantum Queen gold grid engine and the Gold Pivot Breakout EA are the honest comparison set, not other correlation robots.

4. Ten of eleven months agree

Every month agrees in direction except October 2025. Both records lost money in November 2025 and both had their worst month in July 2026, which is stronger verification than any single summary statistic.

November shows the correlation filter working and hurting at the same time. Our test took only 4 baskets that entire month and still lost $175.02, one of them costing $144.48. The filter had correctly spotted that EURUSD and gold were no longer tracking each other and cut trading to a trickle, but the few baskets that passed were the worst of the run. That month dropped our balance from $566 back to $247, a 56.37% fall that erased three months of gains and left the account roughly where it started.

5. What is the real risk on the Gold Spread Reversion EA?

No floor under a losing basket. All 1,202 orders shipped with an empty stop loss field and closeTargetLoss at 0. Drawdown (the largest peak-to-trough drop in account value) is bounded only by the EA’s own exit logic and, past that, by your broker’s margin call. Our worst basket lost $227.94 against an average winning basket of $20.29.

The profit is concentrated. Across 298 baskets the win rate was 69.8%, but the best 30 baskets produced 76% of the net profit and the best 10 alone produced 31%. A smooth curve built on a handful of large wins is more fragile than the same curve built evenly, because removing a few trades removes most of the result. Check it on the drawdown calculator before deciding the curve looks safe.

Margin, not the market, is the practical limit. The live account reports trading activity of 8.19% and a maximum deposit load of 98.93%. Read those two together: it is out of the market roughly 92% of the time, and when it is in, one basket at gold’s highs uses essentially the whole account. Our test bottomed out at a 178.32% margin level. On 300 EUR at 1:30 leverage there is room for one basket and none for a second, so the failure mode to plan for is a gap against you while fully loaded, not a slow bleed.

One point in the EA’s favour. Its Z-Score came out at +5.42 at 99.74% confidence, meaning trades alternate more than chance rather than clustering, which is the opposite of the strongly negative reading a grid or recovery basket produces. The live record makes the same point from another angle: its longest losing run was 13 trades but cost only 5.69 EUR, while its most expensive losing run was 2 trades for -174.43 EUR. The damage here does not come from long streaks. It comes from single baskets that nothing stops.

Gold Spread Reversion EA

Free MT5 pair-trading robot that trades EURUSD against XAUUSD as a two-leg spread. Our real-tick backtest over the same 11 months as its live signal reproduced the profit factor exactly and traced the profit to one leg.

Verified Live Record

★★★★★☆

Risk Design

★★★★☆☆

Profitability

★★★★☆

Pros

• Live profit factor of 1.38 reproduced exactly by our independent real-tick test
• Fixed 0.04 and 0.02 lots across all 1,202 orders, no averaging and no recovery ladder
• Both legs opened and closed in the same second in all 298 baskets, so exposure is never one-sided
• 734 live trades over 11 months on an account with no top-up deposits

Cons

• No stop loss and no take profit on any of the 1,202 orders
• Our worst single basket lost $227.94, more than two average months of profit
• The XAUUSD leg earned 100% of the net profit; the EURUSD leg lost money in both records
• Maximum deposit load hit 98.93% live, so one basket can consume the whole account's margin
• 76% of our net profit came from the best 30 baskets out of 298

Summary

4.0

★★★★☆

Mechanically one of the more honest builds we have tested. Lots never change, both legs open and close in the same second, and 11 months of live history line up with a clean tester run. The problem is the risk shape: no broker-side stop, one trade in our test cost more than two average months, and the account runs at almost full margin.

MT5 Is this working?

FREE DOWNLOAD

Last tested by @Silent in Jul 2026 for ForexCracked Editorial.

EA-vs-EA Comparison

Sources: each competitor’s published review on this site. Our 12-point EA scam checklist covers why a matched backtest matters more than a high headline number.

Frequently Asked Questions (FAQ)

Conclusion

Gold Spread Reversion EA is mechanically one of the cleaner builds in this category: fixed lots across 1,202 orders, both legs always opening and closing together, no averaging, and a live record that our own real-tick test reproduced to the second decimal of its profit factor. However, the strategy is not the balanced pair trade the name implies. Gold earned all of it, the euro leg lost money in both records, nothing caps a losing basket, and one bad month in November 2025 took our account from $566 back to $247. Download it, demo it, and size it as the leveraged gold system it actually is.