Gold Breakwater EA Free Download – MT5 Grid Recovery Review

Gold Breakwater EA is a free MetaTrader 5 (MT5) robot for XAUUSD (gold) that runs six entry strategies and a grid recovery system from one chart. For this 2026 review we ran 100% real-tick backtests over nine months at three lot settings, rebuilt every basket from the trade log, and set the result against a live account. Free download at the bottom.

The closed-trade record is close to spotless. The risk sits in the open trades, and one day in September shows all of it.

TL;DR – Free MT5 gold robot: six entry strategies plus a volatility-spaced recovery grid, with no stop loss on entries. Nine profitable months at fixed 0.01 lots, but one basket floated 14.7% of equity, and with auto lots the 35% guard closed it at a loss. Size small. 

What is the Gold Breakwater EA?

Gold Breakwater EA is an MT5 expert advisor for XAUUSD that opens one position at a time from six strategies and closes it with a trailing stop at 24 pips ($2.40 per 0.01 lot) of profit. If price moves against it, the EA adds recovery positions $20 to $49 apart until the whole basket can exit at a small profit.

Core Strategy and Logic

• Six strategies, one basket at a time – Support and resistance (S01), pullback (S02), trend breakout (S03), M5 bounce (S04), volume (S05) and momentum (S06) are on by default. A seventh, impulse, ships switched off. In 407 test trades, no two strategies ever had positions open together.
• No stop loss on entries – Every first position opens with no stop and no target. A broker stop only appears once the trade is in profit.
• Small target, trailing exit – At about 27.5 pips of profit the EA sets a stop that locks 24 pips and then trails price. The median single-position trade made $2.55 at 0.01 lots and lasted 7 minutes, and 80% closed inside 30 minutes.
• Recovery spaced by volatility – If price moves against the entry, the next position goes in further away, using four-hour volatility. In our test the gaps ran from $20 to $49 of gold price, with a median of $31. The cap is 10 positions per basket.
• Bigger lots on deeper legs – At a 0.01 start the fourth position was 0.02 lots. At larger sizes the lots climb from the second or third position (§4).
• The target shrinks as the basket deepens – 24 pips with one or two positions, 18 at three and 13.5 at four, measured from the basket’s breakeven price. Deep baskets aim to get out, not to earn.
• Calendar blocks – Entries run Monday to Thursday in three windows (01:15-14:20, 14:35-15:20 and 15:45-22:30 server time) plus Friday mornings. The EA still skips the first and last Friday of each month and every Friday in July and August, so Fridays produced only 14 of 383 baskets.
• One loss limit – A drawdown guard closes everything at 35% and stops entries for the rest of the day. The all-time-high filter and the fast-move exit both ship switched off.

This is grid trading with selective entries, the same family as the Gold Grid Strategy EA. The difference is the spacing. A fixed $5 or $10 grid fills fast in a falling market, while these wider gaps mean most baskets never need a second position. Our guide to grid and martingale risk covers the general trade-off.

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

Gold Breakwater EA Settings

These are the inputs worth knowing. Our main tests ran every one at its default unless the table says otherwise.

Recommendations for the Gold Breakwater EA

• Deposit – At least $1,000 for each 0.01 starting lot. The arithmetic behind that number is in §4.
• Leverage – Gold traded between $3,987 and $4,511 in September and October, so one 0.01-lot position needed $40-45 of margin at 1:100. The deepest basket held 0.05 lots, about $220. Margin is not the limit here; the floating loss is.
• Pair and timeframe – XAUUSD only. The strategies read their own timeframes, so the chart timeframe does not matter. We tested on M5.
• Broker – A hedging account with low gold spreads. A raw-spread ECN account keeps you under the 50-point spread cap.
• VPS – The EA manages every exit itself, so MT5 has to stay running (FXVM).
• Check your own numbers with the margin calculator, the drawdown calculator and the lot size optimizer (Find the Perfect Broker For You Here).

Performance Review: Live Account and Our Own Backtests

1. The live account

The live record is five weeks long. For a grid robot, the record that counts is how it handles its worst basket, and this account has not had one yet.

Note: The live record began four days after the deepest basket in our nine-month test (10-11 September). A record that starts after a drop like that can’t show how an account came through it. On our numbers, that basket is enough to stop out a €400 account at 0.02 lots (§4).

2. Our test over the same weeks

We ran the default settings from 15 September to 23 October 2026 on $1,000 and again on $400. Both runs placed the same trades at the same prices, since the lot is fixed at 0.01.

These are not the same trades. The live account took fewer than half as many and sold far more often. In those weeks gold fell from about $4,300 to $4,030, which suited short trades. Our test’s two deepest moments, on 13 and 20 October, were buy baskets the live account never had.

Lot size does not explain the gap. The live account’s gross pips and gross profit line up with 0.01-lot trades, the same size as ours. Treat it as a separate record that has made money since mid-September, not as proof of the defaults you download.

3. Nine months at fixed 0.01 lots

Our main run: Tickmill demo, XAUUSD M5, 1 February to 23 October 2026, $1,000 at 1:100, 100% real ticks, every input at its default.

The run made +$1,087.82, with every month between +$77 and +$144. It closed 383 baskets: 367 needed one position, 15 needed two or three, and one went to four. Only one basket closed at a loss, -$2.08 on a single M5 bounce buy opened on 30 March. The report counts 20 losing trades, but 19 of them were the first, higher-priced positions of baskets that finished in profit.

The report’s 2.90% balance drawdown and 14.69% equity drawdown (equity counts open positions at their current price) differ by a factor of five, and almost all of that gap is one basket.

March is the month that surprised us. Gold dropped from $5,419 to $4,099 inside the month, the biggest monthly range in our test window, and no basket went past two positions. March was also the best month, +$144.43.

4. One basket, three outcomes

On 10 September at 10:06 server time, S01 bought at $4,416.61, near that day’s high. Gold fell for the next 20 hours and the EA added three positions. At the low of $4,300.76, a little before 07:00 on 11 September, the basket was about $287 down on 0.05 lots.

Gold bounced to $4,360 by 09:14 and the basket closed at the shrunken four-position target, +$3.78. That includes a small glitch: in the same second the EA fired two more 0.02-lot recovery orders and closed them at once, for -$0.58.

We ran the same nine months twice more with auto lot sizing, at the High and Very High settings. Everything matches the fixed run until the balance grows enough to lift the lot.

At Very High, the basket’s three positions were 0.05, 0.06 and 0.08 lots. The guard fired at 15:37 on 10 September with gold at $4,334, closed the basket at -$904.85 and stopped trading for the day. That was 35% of a $2,561 balance, gone hours before the bounce.

At High, the guard fired at 01:00 the next morning, as the new trading day began. The basket was about $447 down, roughly 26% of the balance, so the guard does not simply measure a share of your balance. Neither auto run’s deepest equity dip was in September. Both came on 13 October, from a three-position buy basket, after the September loss had shrunk the balance.

The rule this gives you: the worst basket floated about $287 per 0.01 starting lot. To keep that under the 35% guard, you need well over $820 of balance per 0.01 lot. Fixed 0.01 on $1,000 passed. High and Very High did not, and auto sizing raises the lot as the balance grows, so the cushion never builds up.

A fresh account the size of the live one shows how tight this is. On about $465 (€400) at 0.01 lots, the same basket reached a 35% loss at 15:37 on 10 September, and the default guard would have closed it about $164 down. With the guard off it would have floated to -$288, 62% of the account, before recovering. At 1:500 that is not a stop-out, but it is most of the account. At 0.02 lots with the guard off, the same account reaches a 50% margin stop-out by 22:03 that evening at 1:500, and by 15:38 at 1:100.

5. Which strategies carry it

Momentum and M5 bounce made more than half of the profit. Trend breakout fired three times in nine months, so switching it off changes almost nothing.

Two thirds of the baskets were buys, and the deep ones were buys even more often. 8 of the 10 baskets with the largest floating loss were long, including the 10 September basket from S01. Three strategies cannot sell at all, so in a falling gold market they keep buying dips. That is the setup to watch.

6. News days, and what no backtest can show

The EA reads high-impact USD news from the economic calendar and blocks the full day and the next day for CPI, NFP and FOMC. The MT5 Strategy Tester has no economic calendar, so in any backtest that block does nothing. Against the official CPI and NFP release lists and the FOMC calendar, our nine-month run opened 59 baskets on days the EA is set to skip. They made +$178.31, about 16% of the profit, so a working news block would trim the result rather than save it.

The only news rule that ran in the tester treats the first Friday of each month as NFP day. That missed three of the nine 2026 releases: 11 February (a Wednesday), 8 May (the second Friday) and 2 July (a Thursday). With the live calendar connected, the full-day block should catch those. The US bank-holiday block also depends on the calendar, and our test traded 14 baskets on US holidays.

Gold Breakwater EA

An MT5 XAUUSD robot that picks entries from six strategies, exits at a small trailing profit, and handles losing trades with volatility-spaced recovery positions instead of a stop loss. Our nine-month real-tick test at fixed 0.01 lots closed 383 baskets with one small loss.

Profitability

★★★★☆

Risk control

★★☆☆☆

Ease of Use

★★★★☆☆

Pros

• +$1,087.82 on $1,000 at fixed 0.01 lots in nine months, every month positive
• 367 of 383 baskets closed as a single position, and only one closed at a loss (-$2.08)
• Recovery gaps widen with volatility, and only one basket can be open at a time
• Built-in blocks for month-end Fridays, summer Fridays, news days and a daily loss limit

Cons

• Entries carry no stop loss, and one basket floated about $287 per 0.01 starting lot
• With auto lots, the 35% guard closed that basket at -$905 hours before price came back
• Three of the six strategies only buy, and 8 of the 10 deepest baskets were buys

Summary

3.2

★★★☆☆

The entries are selective and the recovery gaps are wide, so most trades never need help. When one does, the basket can sit more than $100 against you for a day. At a fixed 0.01 lot per $1,000 that is survivable. With the lot growing as the balance grows, the drawdown guard closed the same basket at a loss.

MT5 Is this working?

FREE DOWNLOAD

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

How it compares to other gold grid robots we have tested

The three tests cover different windows and sizes, so read them as a rough guide. At a fixed lot it matches the Gold Grid Strategy EA on equity drawdown, with a far shorter live record.

Frequently Asked Questions (FAQ)

Conclusion

Gold Breakwater EA is one of the more careful grid robots we have tested, and at a small fixed lot its nine-month record holds up. However, that record depends on the lot staying small. The same basket that cost a fixed-lot account nothing cost a compounding account 35% of its balance in one afternoon. If you run it, use fixed 0.01 per $1,000, connect the news feed, and give it a month on demo. Then judge it by the lowest equity it reached, not the balance line, and cut the lot if that low passes 15%.