Gold Nine Layer
- Experts
- Version: 1.0
- Activations: 5
Gold Nine Layer runs seven pivot-breakout layers on one XAUUSD H1 chart. Fixed 0.01 lot, a stop loss on every order, no martingale, no averaging down. 2022.02-2026.09 real-tick backtest with default inputs: PF 1.62, max equity drawdown 11.8% on $3,000. Reproducible: same inputs, same period, your own tester.
What it does
The EA has nine layer slots, each with its own swing window (20 to 48 hours), level offset, stop loss, take profit and trailing distance. In the shipped tables 7 slots are active and 2 are empty (#2 and #5), so 7 layers place orders. Each active layer puts a buy stop and a sell stop at levels derived from confirmed swing pivots and the day's opening price, and re-places them only when the pivot changes. Every order carries its stop loss and take profit; after entry a per-layer trailing stop follows the price. The lot is fixed at 0.01 per layer: no averaging down, no martingale.
Why you can trust the numbers
Every figure on this page comes from one routine you can repeat:
- Tester and data: MetaTrader 5 Strategy Tester, "Every tick based on real ticks", history quality 100%.
- Broker, symbol, period: Exness Technologies Ltd, XAUUSD H1, 2022.02.01 - 2026.09.25.
- Account: initial deposit $1,000, leverage 1:500, default inputs (nothing optimised on this window).
- Three runs, median shown: three runs with random execution delay; the run with the median profit factor is shown. Profit factor of the three runs: 1.62 - 1.64; maximum equity drawdown of the three runs: $343 - $354.
If your run differs, post your broker and report in the Comments tab.
Backtest (reproducible with default inputs)
| Net profit | +$4,458 |
| Profit factor | 1.62 |
| Maximum equity drawdown | $354 (35.4% of the $1,000 test deposit; 11.8% of $3,000) |
| Trades | 2,732 (win rate 44.2%) |
| Average win / average loss | $9.64 / $4.72 |
| Longest losing streak | 27 trades |
| Most positions open at once | 9 (median when a new position opens: 3) |
| Average holding time | 4.2 hours |
| Trades per weekday | 2.25 |
By year (net profit / profit factor / trades):
| 2022 | +$110 | 1.10 | 591 |
| 2023 | +$420 | 1.35 | 632 |
| 2024 | +$635 | 1.45 | 560 |
| 2025 | +$1,721 | 2.07 | 543 |
| 2026 | +$1,573 | 1.83 | 406 |
2022 is partial (from 2022.02.01); 2026 is partial (to 2026.09.25).
Long and short trades (same report, split from its deal history):
| Trades | Net profit | Profit factor | |
| Long | 1,483 | +$3,676 | 1.93 |
| Short | 1,249 | +$782 | 1.24 |
Short trades lost in 2022 (-$4), 2023 (-$59) and 2024 (-$135) and gained in 2025 (+$313) and 2026 (+$667).
Re-tested at the recommended $3,000 (2026-09-30, Exness server Exness-MT5Real5, otherwise the same settings, three random-delay runs): profit factor 1.62 - 1.64, net profit +$4,462 to +$4,578, maximum equity drawdown $348 - $359 (11.6% - 12.0% of $3,000), 2,732 - 2,736 trades. With a fixed 0.01 lot the drawdown in money is the same as in the $1,000 test.
Requirements
- XAUUSD H1, one chart. Hedging account (several layers can hold opposite positions).
- Recommended deposit: $3,000 per 0.01 lot. The main test above started from $1,000; with a fixed 0.01 lot the drawdown in money does not depend on the deposit, and the re-test at $3,000 below confirms it.
- A VPS running 24/5. The spread filter is off by default (Inp_MaxSpreadPts = 0).
Risks
- Drawdown: the maximum equity drawdown in the test is $354. That was 35.4% of the $1,000 test deposit, and is 11.8% of the recommended $3,000 (11.6% - 12.0% in the $3,000 re-test). Size accordingly; we consider 10% the comfortable ceiling.
- Losing runs: win rate 44.2%. Most trades are small losses cut by the stop, and the profit comes from fewer, larger winners. The longest losing streak in the test was 27 trades.
- Several layers can hold positions at once: up to 9 in the test (median 3 when a new one opens). The global cap is 22 positions.
- Profit is uneven over time: 2025-2026 account for 73.9% of the net profit in the test. Years with a loss: none. Long trades produced $3,676 of the $4,458; short trades were negative from 2022 to 2024 (see the table above).
- Tick data note: the report shows history quality 100%, but the tester log shows that for 5.1% of minute bars the real ticks did not match the bar prices; the tester generated ticks for those minutes.
Inputs
Inp_Lots = 0.01, Inp_MaxSpreadPts = 0, Inp_BaseMagic = 20263500 (layers use +1 to +9), Inp_Strategies = "" (all), Inp_AllowBuy / Inp_AllowSell = true, Inp_MaxPosTotal = 22, Inp_UseTrail = true, Inp_TrailScale = 1.0, Inp_RepivotOnly = true, Inp_ConsumeOnFill = true. Keep the defaults to reproduce the test.
Where it sits in the lineup
Gold (XAUUSD), from one behaviour to the full set:
- Gold Swing Break MT5: one H4 swing breakout, one position at a time, low frequency.
- Gold Flash Scalper MT5: 32-hour channel breakout, one position at a time, fast trailing stop. The entry point for small accounts.
- Triple Trigger MT5: breakouts of the 24, 48 and 72-hour rolling high/low, trailing exits.
- Gold Nine Layer: a nine-slot pivot-breakout group on one chart (7 slots active in the default tables).
- Gold Grand Line MT5: 6 groups on one chart, 29 active facets. It contains the Gold Nine Layer, Gold Swing Break and Gold Flash Scalper engines (same rules and tables) plus three groups that are not sold separately on MQL5. Triple Trigger is not part of it.
Other markets:
- OIL Surge Break MT5: WTI crude oil (USOIL) H1 swing breakouts.
- Relay Grid MT5: AUDNZD / AUDCAD / NZDCAD mean-reversion grid whose lot grows with each level, no stop loss by default. The only grid in the lineup; every other EA above trades 0.01 lot per order with no averaging down and no martingale.
Utility:
- Adaptive Closer MT5 (free): manage all open positions on a symbol as one book from a chart panel.
The same nine-layer engine, with the same tables, is one of the six groups inside Gold Grand Line MT5.
Support
- Comments tab of this page: questions, setup problems, and "my tester shows something different" reports. Post your broker, symbol and the first lines of your report; we compare it with ours.
- MQL5 private message to the author (Kazuki Kamio) for anything account-specific.
- Version notes are posted in the Comments tab when a version is published.
We answer in English and Japanese.
Price
Price rises by $50 after every 10 purchases, up to $499. Current: $99.
