Kingfisher MT5
- 专家
-
Joshua Dunnink
关于我
这些 EA 的开发者。
我为 MetaTrader 5 开发自动交易系统，专注于黄金 和主要外汇货币对。一切都围绕一个原则：
账户安全第一。
每一个系统在交付之前，都会经过多年历史真实 Tick 数据的充分测试。
为抵御爆仓而设计，而非追求最大数字
我用两件事来衡量一个系统：它能赚多少，以及它能承受多少。
如果某个配置无法通过我自己设定的回撤上限，无论利润看起来多好，它都不会发布。
我宁愿给你一个稳健、能活下来的系统，也不要一个在第一个糟糕的月份就让账户归零的花哨系统。
我如何构建 - 版本: 1.17
- 激活: 20
Kingfisher is a gold (XAUUSD) level-break Expert Advisor with a stop loss on every trade.
Five hunters rest a stop order just inside the nearest unbroken swing high and swing low, so the position opens at the moment the level gives way.
Most breakout EAs are sold on their equity curve.
This one is sold on its protection layers - every value measured on real ticks, every one an input, and a quiet market is a reason to stand aside, not to trade smaller.
Six protection layers, each one measured
Backtest, MetaTrader 5 Strategy Tester, real ticks, XAUUSD, Kingfisher 1.17 and 1.18 (the same trades), 0.01 lot unless stated.
| layer | what it does | how the value was set |
|---|---|---|
| 1. Stop loss and target | Every position opens with a fixed stop and target in gold price: 2.00 and 15.00 USD per 0.01 lot, on any account currency. | Measured on four raw-spread feeds; the stop is fixed, so the spread takes a fixed share of it - use a raw-spread account. |
| 2. One bet per level | Risk is sized per level, not per order: the four four-hour hunters share it, the one-hour hunter carries a quarter, the prior-day hunter three quarters. | Four hunters fill on the same break; sizing per order would have counted one bet four times. |
| 3. Quiet-market gate | No order while the stop would be more than 20 percent of the hourly range. | In the four-year test it placed 18 trades across two quiet years: flat years, not losing ones. |
| 4. Grip exit | Once 1.00 USD ahead per 0.01 lot, a slow fill is closed if it gives back half of its best profit, or more than a share of the 15-minute range. A fast fill rides to the target. | Read on the first tick after the fill: how fast price arrived decides which exit it gets. |
| 5. News and spread guards | No orders ten minutes either side of a tier-1 US release, and none while the spread is more than four times its normal and wider than 1.00 USD. Both on, both inputs. | Priced in dollars, not points: the same guard on two- and three-digit feeds. |
| 6. Weekend and margin | Resting orders are removed before the weekend. A margin cap keeps the orders that can be open at once inside your free margin. | Up to seven orders can rest at once; the cap is why 100 or more at 1:500 is the recommended start. |
Plus the rule guard for prop-firm accounts: daily loss, maximum loss, profit target, best-day cap and a Friday close hour, all off on a Personal account.
No grid, no martingale, no averaging down, no position without a stop loss, no news-trading logic, no DLL, no files, no external resources.
It never changes a setting you chose.
Results
Backtest, MetaTrader 5 Strategy Tester, real ticks, 100 percent history quality, XAUUSD, 10,000 starting balance, Medium risk level, shipped defaults of build 1.15.
Nothing below is from a live account.
| feed | window | growth | profit factor | equity drawdown |
|---|---|---|---|---|
| BlackBull (raw spread) | 2024-01 to 2026-09 | x21.7 | 2.38 | 10.5% |
| Fusion | 2024-01 to 2026-09 | x19.4 | 2.20 | 10.6% |
| Vantage | 2025-07 to 2026-09 | x27.7 | 3.01 | 8.3% |
The headline is the weakest feed.
The first trades fall in 2024-04: gold was too quiet before that.
Results differ by broker: on an account whose gold spread was about 0.24, 8.5 percent of the entries were stopped out within a second of the fill, and they earned about half as much per trade.
Test on yours.
Backtest, MetaTrader 5 Strategy Tester, real ticks, Kingfisher 1.17 with the news, spread and fill guards on, Medium risk, BlackBull, 2024-01 to 2026-09: growth x25.8, profit factor 2.56, worst equity drawdown 10.1 percent.
Vantage and Fusion are left out of this line: their 2026 tick history is thin.
Five hunters, one level
Four hunters watch the four-hour swing high and low with slightly different distances and patience (15 to 60 hours).
One watches the one-hour level and is called the Skim.
A sixth, Daybreak, watches the prior day's high and low.
They fill on the same break, which is why risk is per level.
The system earned its result in one kind of market, gold moving in large daily bursts, and stands aside in quiet ones.
What to expect
- Flat stretches. In quiet markets it stands aside, sometimes for weeks; one test month had no trade at all.
- Slow bleeds. Between active periods small breakouts keep failing: the balance can drift down for weeks, now and then for a few months.
- Profit comes in bursts. Most of the tested result came from periods of large daily gold moves.
A year without them is a small year. - Small accounts. No minimum balance of its own: the broker's margin for 0.01 lot of gold sets it (about 5-9 at 1:500, 25-45 at 1:100, 120-225 on EU-regulated accounts).
Backtest, real ticks, Kingfisher 1.15, three live brokers, an account started on every day of 2024-2026 and followed for twelve months: none got stuck from 100; from 50 about one start in five did at 1:500; from 25 or 10 half or more.
Quick facts
| Symbol | XAUUSD; on any other chart it loads, says why and stays flat |
| Account | Hedging recommended; a raw-spread (ECN) gold account |
| Chart | Any timeframe - it reads H4, H1 and M15 itself (M15 in every test) |
| Sizing | Risk level Low, Medium, High, Hot or Custom; the Journal states the lot at start-up |
| Minimum deposit | None of its own; 100 or more at 1:500 recommended |
| Setup | One section, "Start here"; no .set file needed |
| Prop-firm rules | Account types Prop 2-step and Prop 1-step; the rule inputs are off on Personal |
| Panel | A glance panel: status word, one dot per hunter, a guard strip that lights when a layer acts |
| VPS | Recommended - orders rest on the server but the exit logic needs the EA running |
Manual and research
The complete manual (every input, the sizing arithmetic, the prop-firm chapter, the exact Strategy Tester settings for every figure above), the research note (how each mechanism was measured and what was rejected) and a monthly gold market-state page are published on the seller's website - the address is on the seller profile under Contacts.
Market rules do not allow links inside this description.
Risk disclosure
Trading leveraged instruments carries a substantial risk of loss.
Every position has a stop loss, but a stop can fill past its price in a fast market, and a run of failed breakouts is a normal part of this design.
The results above come from a period of large gold moves; a long quiet period produces few trades and slow losses.
Past backtest performance does not predict future outcomes.
Test on a demo account first.
