Boreas MT4
- エキスパート
- バージョン: 1.42
- アクティベーション: 15
Boreas is a breakout Expert Advisor for XAUUSD (gold) on MetaTrader 4. Backtest on real ticks, May 1, 2024 to Oct 10, 2026: $200 grew to $687 at a fixed 0.01 lot, 925 trades, 82.1% win rate, profit factor 1.70. Stop loss on every trade. No martingale, no grid. One position at a time.
Live signal (MetaTrader 5 version, same logic, real account, $200, fixed 0.01 lot, started October 15, 2026): https://www.mql5.com/en/signals/2395623
Gold doesn't move randomly around certain prices. It stalls there, or it breaks through hard. Boreas waits for the hard break.
It sits flat until price comes close to a level that matters, puts a stop order on the server, and lets the market pull it in. If the move is real, the trade is in profit within seconds and the stop is already at breakeven. If it isn't, the loss is small and fixed.
Median time in a trade: 59 seconds.
At a glance
| Symbol | XAUUSD (attach to any chart timeframe, M5 recommended) |
| Backtest, May 2024 - Oct 2026 | +$487 at 0.01 lot from $200, profit factor 1.70 |
| Win rate | 82.1% (759 wins, 166 losses) |
| Trades | about 7 per week |
| Risk control | Stop loss on every trade, no martingale, no grid, one position at a time |
| Minimum deposit | $200 for 0.01 lot |
| Account | Raw spread / Zero / ECN, fast execution |
Key features
- Real stop orders on the server. When price gets within $5 of a level, Boreas places a buy stop or sell stop. The order fills on the broker side, not after a round trip from your VPS. (You can switch to virtual orders if your broker doesn't like pending orders.)
- Breakeven fast, then trail. The stop moves to entry as soon as the trade is about $1 in profit, then follows price tick by tick and only ever tightens.
- Several level types, several timeframes. Each level is used once. Price already crossed it? It's spent, Boreas moves on.
- Buys and sells. 484 buys and 441 sells in the backtest.
- Skips NFP and FOMC days. No new trades on those days, and everything is closed before the weekend.
- Built-in slippage log. Every fill is written to a CSV file: the price Boreas wanted, the price you got, spread, delay. You'll know exactly how good your broker is after one week.
- Restart-proof. After a terminal restart or VPS reboot it picks up its open trade and remembers which levels it already used today.
Backtest results
The MT4 tester has no real tick data, so the backtest was run on the MetaTrader 5 version of Boreas (same logic) in the MT5 Strategy Tester, every tick based on real ticks, Exness XAUUSD, real spread, fixed 0.01 lot. Each period starts from $200.
| Period | Gold that year | Result (0.01 lot) |
|---|---|---|
| May - Dec 2024 | Slow, grinding rally | -$16 (PF 0.93) |
| 2025 | Strong trend | +$199 (PF 1.74) |
| 2026 (to Oct 10) | Big swings | +$304 (PF 2.57) |
The pattern is simple. Boreas makes money when gold moves fast and breaks levels cleanly. In 2024 gold crawled upward and kept faking breaks, and Boreas went roughly flat (a small loss, 8 months). In 2025 and 2026 gold ran, and so did the equity curve.
Average win is $1.56, average loss $4.19 at 0.01 lot. So it wins often and small, and the losses are fixed in size. Worst single trade was -$9.73, when gold dropped straight through the stop during the Jan 29, 2026 sell-off. Max drawdown was $72, between Oct 2024 and mid-2025, before the strong part of the curve.
Costs matter here, and I tested them. Screenshot 5 adds extra cost to every single trade. With +$0.30 per trade (slippage, wider spread or commission) profit factor is still 1.28. At +$0.40 it's 1.15. That's why I built the slippage log: run it, read the file, and you'll see where your broker sits on that chart.
How it works
Boreas trades breakouts of reference levels that gold watches, on more than one timeframe. It doesn't use any indicator that repaints and it doesn't look into the future. Every decision is made on prices that are already final.
The exits do most of the work. The initial stop is set behind the level when the order is placed, so the risk is known before the trade opens. After that it's breakeven first, trailing second. A trade that doesn't run gets cut small. A trade that runs gets followed.
It doesn't trade every day. Some days nothing breaks, and that's fine.
Requirements
- Symbol: XAUUSD. If your broker uses a suffix (XAUUSDm, XAUUSD.r and so on), attach to that symbol.
- Deposit: $200 per 0.01 lot. A losing trade costs about $4-5 at 0.01 lot, roughly 2% of $200.
- Leverage: 1:100 or more.
- Account: Raw spread, Zero or ECN. Fast execution matters more than anything else for this EA.
- VPS: strongly recommended, close to your broker's server.
- Server time: the levels follow your broker's day. Built and tested on a GMT+0 server (Exness).
Quick setup
- Open an XAUUSD chart (M5 is fine) and attach Boreas.
- Leave the defaults. The live signal runs these same settings, with one difference: it uses a fixed 0.01 lot.
- Choose your lot: fixed 0.01 for $200, or "lots per 1,000" to let it grow with the balance.
- Turn on AutoTrading. Check the Experts tab: Boreas prints whether trading is allowed.
Main parameters
| Parameter | Default | What it does |
|---|---|---|
| Lot mode | Lots per 1,000 | Fixed lot, lots per 1,000 of balance, or % risk at the stop |
| Lots per 1,000 | 0.0333 | 0.01 lot per $300; below that it uses the minimum lot |
| Order type | Stop orders | Real stop orders on the server, or virtual (market at trigger) |
| Place within | $5 | How close price must be before the stop order goes on |
| Avoid news | true | Skip NFP and FOMC days |
| Friday close | 19:58 server | Close everything before the weekend |
| Max spread / overshoot | off | Optional filters if your broker's spread jumps |
| Slippage log | true | Writes every fill to MQL4\Files\boreas_slippage_<magic>.csv |
| Magic number / comment | 20261014 / bo | Change if you run it next to other EAs |
The level settings are advanced. Leave them as they are.
FAQ
Why no trades today? Price didn't come near a level, it was an NFP or FOMC day, or the level was already used. Normal.
Can I run it on other symbols? No. All distances are tuned for gold.
Does it work on a Standard account? It can, but you're paying a wider spread on every trade. Check the cost chart first. Raw or Zero is the better choice.
Prop firm? It uses a fixed stop on every trade and keeps one position, which fits most rules. I haven't run it on a prop account myself, so check your firm's rules on short holding times.
Will my results match the signal? Close, if your broker's spread and execution are similar. That's exactly what the slippage log tells you.
Support
Message me after purchase and I'll help with setup, broker choice and reading the slippage file. Updates are free.
Trading gold with leverage carries a high level of risk and you can lose money. The backtest uses historical data and the live signal is new. Past performance does not guarantee future results.
